Monday, July 5, 2010

Regenerative Braking - Entirely a Myth?

Rather interesting week. It started with a longish drive in the new Speedster design. We reached about 107 miles when it seemed politic to bring it into the barn and check some batteries.

One notable thing about this was that 107 miles was the max drive we ever actually did with the original Speedster. Over the winter, we took out all the toe-in on the front end, went to a Redline MTF fluid in the transaxle, and made a couple of other minor tweaks that might perhaps get us another mile or so. But I never did get around to actually doing the drive.

But the new Speedster design features an AC induction motor and 3 phase Curtis controller that has more than adequate provisions for regenerative braking. As we are using a very similar battery pack between the two cars, we should have gotten SOME increase in range from the regenerative braking.







Regenerative braking, of course, is the concept of using the motor as a generator to slow down and stop a vehicle, instead of using the "friction" brakes that basically convert forward motion into heat in order to stop the vehicle.

It is fairly obvious to anyone that this must offer some efficiencies for an electric car. And indeed, the idea is not only not new, but has been actively discussed and deployed for some thirty years.

I have for some time suspected all is not as it seems in this. The thing that bothered me was the law of conservation of energy. Broadly applied this indicates that a mass in motion will tend to stay in motion, and the energy of interest is usually referred to as kinetic energy or "stored" energy held in the forward inertia of the mass. But its best and highest use is of course to continue in motion. If we convert it to electricity and try to put it back in a battery, we naturally fall prey to all the inefficiencies inherent in that.

In a real drive of a real car, how much of this regenerative braking would actually be useful? While the concept of driving a car down the street doesn't sound like rocket science, there are really a LOT of different things going on more or less simultaneously and in various sequential time frames that is in reality quite complex. Yes, we do accelerate, and yes we do brake, but we do a lot of other things as well, and so does the car.

So I had always said it is probably a little overhyped, and my sense was regenerative braking could provide 6-8% efficiencty gain, not the 15-25% I've heard bandied about.

More recently I fell under the spell of a fascinating speaker, Dr. Illah R. Nourbakhsh. Dr. Nourbakhsh is Associate Professor of Robotics, The Robotics Institute, Carnegie Mellon University. He was the former Robotics Group Lead,
Ames Research Center, National Aeronautics and Space Administration.

Dr. Nourbakhsh drives a RAV-4 EV, and has noted some startling efficiencies in real world driving. Unfortunately, on reviewing his data, it becomes apparent that he actually measures AH consumed going forward, AH produced during regen, and assigns efficiency values based on those measurements. Bu the hasn't actually done anything with actually measuring how much energy is required to move the car. This tends to ignore driving dynamics.

He has started a web site and is aware of the concept that regen will do various things on various driving patterns, and indeed his web site at http://www.chargecar.org purports to collect actual drive data from various people across the country. But he appears to be focused on variations in the two measured quantities - AH out and AH in.
They are actually working on a design for an ultracapacitor implementation that will make capture of this regenerative braking more efficient.

Unable to resolve all this, we set up three tests.

In the first test, we drove the car 76.6 kilometers with full maximum current regenerative braking tied to pressure applied on the brake pedal. We drove the route almost entirely in 2nd gear. And we ignored how much was coming out and going back in, and looked for the TOTAL AH consumed on the 76.6 mile drive. We drove through very urban area mostly with dozens of traffic stop lights and signs, lots of vehicular traffic, and plenty of hills and curves that should have made regenerative braking quite attractive.

In test two, we drove the same route at the same temperature and SOC and again used 2nd gear all the way. But this time we added neutral braking. Neutral braking is braking applied by using the motor as generator but driving that with the accelerator pedal. The car accelerates of course when you press the brake pedal. But when you start to release the throttle pressure, regenerative braking kicks in. I missed a turn on this lap and so we totaled 75.2 km.

In our third test, we drove the same lap under the same conditions but with ALL regenerative braking disabled.

Obviously, we expected maximum efficiency from the more aggressive regen derived from brake and throttle both.

We would expect the 2nd greatest efficiency from the car with regen deployed only on the brake signal.

And NO regen should bring up the rear.

Surprisingly, here's what we found.

Drive 1 76.6km 78.2 AH or 1.02088 AH per kilometer
Drive 2 75.2km 77.7 AH or 1.03324 AH per kilometer
Drive 3 76.6 km 78.0 AH or 1.01827 AH per kilometer


This shows the LEAST amount of energy used to accomplish the 76.6 km drive to be when using NO regenerative braking at all. The differences were pretty close - probably in the noise level. But what IS apparent immediately is that regenerative braking in ANY form would appear to have no efficiency additions AT ALL.

Yes there were plenty of AH going out and coming back in during regenerative braking, but over the course of the drive, it appears to have nulled out to zero as a benefit.

I found this a bit astounding. Chevrolet's Volt, the Tesla Roadster, almost ANY serious attempt at an electric drive train pays homage to regerative braking. Has NO one ever done an end to end test to see if it actually worked?

I'd repeat the test, but I haven't a clue what I would do different.

Jack Rickard

120 comments:

  1. That is very interesting, and surprising. My assumption has been that unless you have long steep hills where you would normally need brakes to hold your speed, or a lot of stop and go city driving, regen wouldn't do much, but didn't think it would do nothing at all. When you think about it, a few seconds of 100-200 amps coming to a stop really isn't that much energy. I often drive down a mile long hill with a 45 mph speed limit that requires braking to hold speed. With regen I average about 50 amps going into the pack over that mile. That certainly won't get me back up the hill but I should get around a mile of range on the flat at about 40 mph. Still not a significant addition to my 50 mile range, but a number of those hills on a trip must add up to a few miles. You just need longer, steeper hills.

    Another take on the throttle/brake pedal regen issue:
    http://www.saxton.org/tom_saxton/2010/07/3-ev-lessons-for-nissan.html
    "I can control speeding up, maintaining speed and slowing down all with one pedal. With just a little bit of time behind the wheel, it quickly becomes a more natural and comfortable way to drive."
    This equals my own experience, I feel a bit disappointed when I have to use the brakes.

    JRP3

    ReplyDelete
  2. If it is a more comfortable way for you to drive, then I would note that it doesn't cost much to do it, and indeed the small differences we saw could be noise and it costs nothing to do it. But it also doesn't gain anything as far as range or efficiency. You would be doing it for the feel of it.

    I myself prefer to free roll although I do like the fell of the brakes from regen on braking and that's probably how we'll set up the car. I have to believe there is some added stress on the cells and on the drivetrain to do all that regenerative braking, and with no efficiency gains I'm tempted to remove it all. But we'll probably configure it for modest assist when actually putting on the brakes.

    On the other hand, I feel a bit foolish spending $104 on the pressure transducer to improve what were already perfectly good 4 wheel disk brakes.

    You are uniquely positioned to refute or confirm this claim JRP and I wish you would. Do a lap - even 20 miles the way you have it and another with it turned off and let us know what YOU get. As long as its the same lap done the same way, I think you'll come up with the same result.

    Jack Rickard

    ReplyDelete
  3. Unfortunately the programmer was a rental and I no longer have it, so I can't turn regen on and off. However, once I get my replacement throttle from Kelly I think I can do a lap watching the ammeter using a neutral throttle position and left foot braking to approximate no regen. I'll give it a shot and let you know.

    JRP3

    ReplyDelete
  4. That sounds difficult. I assumed you had a 1311 programmer.

    Jack Rickard

    ReplyDelete
  5. Again tempting us with capacitors: I remember you talking about them in another video. If I understand how they work, they could go a long way in increasing the longevity of already tired lead/acid batteries. Also, it's a shame you never got too much into the A123's or similar batteries. They seem a little more affordable/available than prismatics. I plan on playing with some 18650's that is rumored to power the Tesla. Ironically, these are the only "lithium" available to the poor man's EV. I sure love your videos, I hope you never stop making them....

    ReplyDelete
  6. Jack,

    Are you going to do a test and compare the same drive using your other Speedster with the DC drive system? I'd be interested in the same AH used that way too. My opinion is that regen is mostly for helping stop the vehicle rather than depending upon the brakes to fully take the brunt of stopping the vehicle. It has a side benefit of relieving stress on your brakes and you do get that little bit back in. If your in a stop and go situation for lets say 10 miles and then a straight drive no stops at 10 miles you may actually see some difference but I agree that regen is quite over rated for what it can actually do for you in the end.

    Pete

    ReplyDelete
  7. I'm toying with a concept for a project I have backed away from three times before. The demise of regenerative braking rather inspires another look. I just didn't have an AC drive train available to me large enough at any reasonable price. But a 13 inch Netgain or a pair of 11's would be pretty easy. And we have several "thousand amp" controllers to pick from now. And yes, I have another batch of Maxwell's in hand at the moment.

    I do not find the A123's viable. Tesla is now using a cell from Panasonic and never did use A123 cells. I don't find either more available or more affordable than the prismatics. If I was on a budget and doing small projects, I would probably be playing with Headway's right now. About the size of a beer can, they can do 8 or 10 AH, and they have screw terminals allowing pretty easy assembly. But I haven't tested them....

    I will undoubtedly stop making the videos, but thank you. It probably will not be soon as we are having too much fun and I believe true electric cars are going to be "just on the horizon" for a number of years. I will do it while it appears to be a frontier technology and when the townies move in, I move on...


    Jack Rickard


    More on that later...

    ReplyDelete
  8. Love your videos

    JMS

    ReplyDelete
  9. About the only other thing you might do is to run the test again at least once more with any one of the setups and see what the "noise" level actually is. I'm assuming you are going to leave the car in the mode used in the first test (conveniently the middle value), so try running again with that setup and see how close you are. If you are not as high or low as the other test points, you would further prove the point.

    ReplyDelete
  10. I was too intrigued by this issue to wait for a new throttle, so I took the trigger throttle off my AMPhibian 6 wheeler and strapped it to the shifter. I've done it before and as long as you don't need to shift it works well, and allows very fine throttle control. Here's how I tested. All driving was done in second gear.
    Did a 5 mile drive to warm everything up a bit and to see if I could reliably hold zero regen, and I could. A few brief moments of low amp regen but nothing that would affect anything. Charged the car, did a 16 mile trip with regen on fairly empty roads with only one traffic light to reduce variables. Trip includes about a mile long downhill that takes about 50 amps of regen to hold speed at 45mph, or brakes if not using regen. Came back, charged up, did it again without regen. Charging was done through a kill a watt meter since I don't have an ah counter hooked up.
    Trip with regen: 4240 watt hours from the wall
    Trip without regen: 4680 watt hours from the wall. Since it was about 90F outside everything was much hotter after the second trip so the batteries may have taken a little extra charge. I'll need to repeat both runs again to eliminate any noise, including driver input. I should also have an ah counter hooked up this week.
    Some things I noticed when driving without regen, I tended to carry my speed longer into speed zones by coasting, probably because I'm not used to using the brakes, and I also carried more speed before coming to stop signs or lights, for the same reason. I think I should try to coast more even when using regen since as you point out conserving forward momentum is better than trying to recover some through regen. Watching your trip in the video I saw very mild hills and very little opportunity for what I would consider significant regen. Not surprisingly long steep hills are needed to provide that, maybe block to block city driving in traffic as well. I'll continue testing and report results, but I'd say unless someone regularly drives long hills where braking is needed to hold speed, you won't get much of anything from regen. I'd like to see tests of stop and go city driving, as well. The higher percentage of time spent braking should make regen more significant, especially for jackrabbit drivers. Even with no benefit from regen I think there are still significant advantages to AC systems and hope to see more affordable higher power units in the future. Hopefully we won't have to wait for wrecked Leaf's and such.

    JRP3

    ReplyDelete
  11. JRP:

    Our drive was definitely stop and go city driving for at least half the drive with LOTS of lights and cross traffic. And about a third was very rolling hills and curves. We did not observe the difference you did at all.

    We had several hills of about 1/2 mile - not terribly steep, but steady. Using regen we did recover energy during these, but on the other hand , not using regen we coasted freely for the whole distance without any throttle - both are gains.

    The mapped drive was designed to MAKE regen significant. We were of course aware that simply driving down the freeway doesn't accomplish much here.

    Jack

    ReplyDelete
  12. It appears to me that you have discovered that using brakes wastes energy by heating the brakes and using regen wastes energy by heating the controller, motor, and battery. Who knew that it would all come out even?If you want to use a capacitor to increase range, I figure a one farad cap would increase range by 3 meters, and then only if it is charged with house power. When you drive, you should use the gear which gives the minimum current for the speed you want to drive.

    Bob

    ReplyDelete
  13. "We had several hills of about 1/2 mile - not terribly steep, but steady. Using regen we did recover energy during these, but on the other hand , not using regen we coasted freely for the whole distance without any throttle - both are gains."

    I think that says it all right there. You coasted freely, which means the hills were not steep enough to force you to use the brakes. Coasting is better than regen, sure, but regen, over time, is better than braking over time. If the time interval is too small there won't be anything to notice. A quick look at the topography of Cape Girardeau shows maybe 100ft elevation change for the whole area and none of it abrupt. I drop 400ft over two miles on my way to work every day. If I drove slowly on the few flat sections I might actually arrive with more charge than when I started. Certainly the topography of New York's Finger Lakes region is unusual, but those of us living near the lakes can probably expect some actual return from regen, and much longer brake life as well. Most places, probably not so much.

    JRP3

    ReplyDelete
  14. Mr. Rickard,
    I would certainly love to get my hands on some Headways, but they are a little steep in price still, and no "used" ones yet that I have found. The 18650 form factor has been around for a while though in laptops, and "bad" laptop batteries are giveaway priced with usually only one or two bad cells in a pack. This can potentially offer a large number of cells for "poor man EV" prices. Of course, I am told that attempting a pack with these is foolish without a good BMS. Sound familiar? :P
    Yes, I admit it, this is my feeble attempt at getting you interested in 18650 cells as a potential energy source where you can once again blast the myth that a BMS is necessary. Though these type of cells do seem less stable than prismatics. That would explain all the liquid cooling and safety redundancies in Tesla packs. But I like living on the edge anyway. Not like it could be much more dangerous than driving around with 10 gallons or more of highly flammable liquid around anyway, right?
    As far as making videos, if you keep making them as long as EV's are fun, that's pretty reassuring, because that should take a very long time indeed before they stop being fun. I took mine for a little spin today just for kicks. It just never gets old....

    ReplyDelete
  15. JRP;

    Still typing. So regen, demonstrated to have NO demonstrable effect at 100 feet variation, but you firmly believe it will work GREAT at 400 feet elevation. And you base this on what data point? Test question, 4 X 0 = WHAT????

    Why do you guys do this? You argue the number of angels on the head of a needle but you don't believe in angels and you don't have a needle?

    Buy or borrow a 1311 and configure your vehicle both ways, drive it both ways, and let us know what YOU get with 400 feet of topography. It might indeed make a difference. Let's try to explain it AFTER you have a different result.

    You've already got an explanation for a different result, but you don't HAVE a different result. You don't have any results. You have a guess.

    That's what got me in trouble here....

    Jack Rickard

    ReplyDelete
  16. Hello jack, Thanks for the great videos and you are always entertaining.
    With regards to the regenerative breaking i would like to see an analysis done that takes into account the laws of thermodynamics.

    As I recall these laws govern everything that happens.

    These I believe are the first and second laws please correct me as needed.

    1 Energy can be neither created nor destroyed. It can only change forms.
    In any process in an isolated system, the total energy remains the same.
    2 total entropy of a system will increase over time{meaning that all processes "leak" energy} All engines are less than 100 percent efficient.

    I don't believe going up or down hills matters as long as you plan on returning to your point of origin. (All the downhill coasting is offset by uphill power needs}.

    As said above, maybe heating of the electric components when using regen is equal to heating of the breaks when breaking.

    I would love to see a demonstration illustrating what is happening, maybe you could break out the redwood sticks again.

    regards, JMS

    ReplyDelete
  17. Hi Jack, I think it is not necessary to say that I love your programs.
    My background: I am a mechanical engineering student somewhere in the universe and also recently we had an electric car project here.

    As far as I conceived the story...
    I think you've measured the wrong parameter for regenerative thing. I can't run the mathematics here but the AH you have measured is the energy that is consumed to move the car with that mass and drag coefficient for ~78km with an overall drive train efficiency and it has nothing to do with regenerated energy you put back into the battery pack. I mean if you push the car with the same conditions (speed,distance,mass,...) by your own hands, the total energy consumed will be the same (it is a function of dynamical properties of the object you are trying to move mainly aerodynamics).
    Actually, The regenerative brake delivers you a battery pack with higher state of charge at the end of your trip for regen on, in compare with regen off.
    If you want to measure the regen efficiency you should measure the full range of your car starting with fully charged battery pack and regen off, then do the same with regen on.
    What do you think of that?

    ReplyDelete
  18. This is an example of thermodynamics in action.
    We must talk about the system consisting of:

    1 The car and battery

    2 The route you take with friction and other variables
    [wind and heat}

    3 processes you are running {such as radio heater a/c/ and regen}

    This can be considered as a closed system with 3 basic parts as listed.
    It seems the only one we can control is number 3 so we have to limit as best we can the things that are moving energy into or out of the battery.

    The hills don't matter, the thing that matters is how many processes you are running. the regen process only gets its energy from the forward motion of the car and the forward motion of the car only gets its energy from the battery. Furthermore all processes are less than 100 percent efficient so the regen process has to put less energy back than it takes out. think about it

    ReplyDelete
  19. I forgot to say I love these videos.
    The above comment regarding thermodynamics is mine.

    Regards, JMS

    ReplyDelete
  20. Navid - I think you missed what Jack did on the test drives. The Xantrex measures the enery put back into the battery during regen and hence the figure read at the end of the drive is NETT AH used - inclusive of drivetrain, battery and controller efficiency. If regen was actually putting energy back into the pack the nett AH used should be less and Jack would be able to report a higher "efficiency" - or more range for the same AH used, in other words.

    Your suggestion of driving to the full range of the pack (apart from being extremely time consuming and running the risk of expending the pack in the middle of nowhere) is no different than measuring the energy used for a given trip.

    ReplyDelete
  21. Jack, did you miss the part where I drove the same route twice, and used about 400whs more without regen? Yes it is possible to hold a neutral throttle position and not use regen, I don't need a programmer, (but I'll take one if you're giving them away). Sure there might have been some variables with my methodology, and I plan to repeat the test. At some point you have to realize that a long enough and steep enough hill will put real energy into your battery pack. Are you going to do that in Cape Girardeau? Doesn't look like it. Is it possible that somewhere else with steeper and longer hills you might? I guess you don't think so. I'll keep testing and see what happens.

    JRP3

    ReplyDelete
  22. JRP.

    I didn't precisely MISS the part where you drove the same route twice, and thought you did one of them without regen. I didn't comment on it as an obvious kindness. You do actually need a programmer, and you do actually need to verify that you don't generate any current back into the pack.

    Longer steeper hills means more regen. It also means more energy to climb back up the same hills. A series of small hills or a series of 14000 foot mountains makes no difference per se. If the frequency and complexity of accels and decels and lights and pedestrians, etc is notably different, you might get a different result. But not persuasively by holding your throttle real carefully to "simulate" no regen.

    You are so focused on the regen being there, you would not be a good candidate to test anything. You already know the answer before performing the test. There is nothing I can do with that.

    Jack Rickard

    ReplyDelete
  23. JMS:

    Yes, there will be less going back in than we took out. But the theory, which made perfect sense, is that you have to brake sometimes anyway, and that is a waste of energy as heat. If we could recapture SOME of that, it would offer SOME efficiency gain overall. Many have estimated this at 20-25%, mostly by doing a very simple comparison of the actual amp hours out and the amp hours in.

    I personally thought it would be 6-8% because I did recognize that the dynamics of an actual drive are more complicated than that.

    A zero percent gain, and indeed technically a small penalty for regen was not on my radar screen. But that's what we saw. It's an hour and 45 minute drive guys. Not a lap around the block. That they came out to within 0.2 AH per lap is amazing in itself.

    So why the result? I've actually got a bit of grip on this. But it is proving aggravatingly difficult to pose succinctly.

    We all see it as accelerate/decelerate. An actual drive is somewhat more complex than that.

    As a thought experiment. You start at a stoplight, drive about 3 blocks, and stop for a second light.

    1. You could accelerate for 1 block, glide one block, decelerate 1 block. You're speed would be about 50 mph, but you could do that.

    2. You could accelerate for 100 feet, glide for 2.5 blocks, and stop suddenly in 100 feet.

    3. You can move the end of acceleration point, anywhere along the line, and indeed you can change the rate of acceleration dramatically in doing so and still average the same elapsed time for the 3 blocks. IN other words, at some low level rate of acceleration, you could acclerate from a standing start the entire 3 blocks and brake the final 50 feet and never exceed the speed limit.

    4. You could accelerate the entire 3 blocks at the fastest rate the car can achieve.

    5. You could miss the light and blow through at the end.

    6. The light could be green at the end.

    7. The light could be red at the end.

    8. There could be a rise over the first block, and a downhill for the last block.

    9. It could be all uphill.

    10. It could be all downhill.

    The regen is not just going to put energy back in the pack. It is also going to dramatically effect the ratio of glide to accel to decel, the feel and performance of the car, and the way you drive it. Driving those three blocks actually comprises a CONTINUOUS stream of decisions and changes and modifications to acceleration, speed, and decelerations and that continous stream of corrections will be signficantly altered by regen, the degree of regen, or teh lack of regen. The one thing we know it will NOT be is an identical drive.

    Expand 3 blocks to 76.6 kilometers. And then put a meter on only the end result. That's what we've done.

    But that's what IS real. The goal is not to count amp hours. It's to drive a car a certain distance and use the minimum energy. We employ techniques to improve that, but most of them require additional hardware or software. Are the additions worth thte gain? Well you have to first measure the gain.

    Aside from theoretical comparisons of what happens to current flow. What if we drive the car and measure the net? That's what we did. And teh results indicate NO gain from regen.

    The easiest explanation for why we get identical net results with three DIFFERENT levels of regenerative current IN to the batteries is that we actually had three different levels of current OUT to drive the car.

    And the only thing that I can conclude is that the additional current to restore a given speed after braking from regen increases at least sufficiently to negate the gain FROM the regen.

    If that's a little circular, as I say, I'm struggling to come up with a redwood stick analogy that will make it clear. BUt I'm convinced the heart of it is in the dynamics of the drive - NOT the setup of the controller, the batteries, or how well we measured the current in each direction.

    Jack Rickard

    ReplyDelete
  24. Interesting.

    It seems both simpler and more complicated than the current explanation...

    Simpler, because regen is ultimately just another form of parasitic drag. Sure, its a little less parasitic than it initially seems, because some of the loss is returned to the pack- but only some, and in average conditions like yours, its a slight net loser. Its probably a bigger net loser on flatter terrain. That's the complicated part, but before we get there...

    The curious footnote here is that under certain circumstances, like JRP's hilly terrain, regen can indeed offer a significant net gain over not using it, despite the drag. That certainly would happen where travel was essentially uphill and downhill, and staying under the speed limit on the downhill portions required lots of drag (service brakes,) some of which would be returned to the pack if you did the drag with regen instead of brake friction. Of course, the magnitude of the hills is going to determine how much brake energy must be dissipated, and therefore how much regen energy could be collected to offset the basic drag of having regen. So JRP (or better, Roland Wiench) might indeed have a net regen gain over not using it where he would otherwise be frequently riding the brakes or engine braking an ICE, but it probably isn't as much as JRP thinks, and it certainly does need proper investigation.

    But the complicated part is far more interesting: I think you are about to say that the way you drive the two regen schemes you programmed is the most reason you used more energy using them. I agree with you, and I don't find that surprising at all.

    I've always thought that part of the reason manual transmission cars get better mileage than automatics is that people simply drive them smoother. Modern automatics are so efficient that there should really be less of a difference in mileage than there is with them- I explain this as a human factors issue. People simply throttle manual cars more gently. They buck and complain if you don't. I would love to see some data comparing throttle position of automatic vs. manual transmission cars over the same routes. The torque converter encourages constant throttle adjustment, lugging, excessive tip-in, and other wasteful accelerator techniques that are actually jerky and unpleasant in a car with a gearbox that feeds that information back to the driver instead of masking it, as the TC does.

    I think the same thing is happening on your test loops, but it is both more subtle and more obvious. Coasting freely is a very satisfying feeling, particularly in a very quiet car, and if you're at all interested in playing with that, it encourages you to put in less power and see how far you can get on it. That's the subtle part. If you are also motivated to save energy because the vehicle has limited range, we can start to put together a more obvious human factors case about what is happening.

    I notice in your description that you've been careful to say you weren't hypermiling the car without regen, it was just easier to coast it. That's a clear indication that there is a significant operational (what happens when you drive it) component to the results in your test, and in fact you are talking about the feel explicitly.

    I would say it this way: a coaster is more rewarding and intuitive to drive efficiently, and that operational fact overwhelms the energy regen can capture in average use.

    Just a thought...

    TomA

    ReplyDelete
  25. This comment has been removed by a blog administrator.

    ReplyDelete
  26. This comment has been removed by a blog administrator.

    ReplyDelete
  27. This comment has been removed by a blog administrator.

    ReplyDelete
  28. This comment has been removed by a blog administrator.

    ReplyDelete
  29. Of course, there probably isn't any regen scheme you could come up with that would significantly improve regen's test performance in your loop. Even if you just substituted regen for the service brakes- you don't use them very much, so there's apparently very little braking energy available in typical EV driving.

    Good data, Jack, and useful, too...

    TomA

    ReplyDelete
  30. So sorry about the multiple posts! The interface was throwing errors at me, but obviously publishing posts nonetheless...

    ReplyDelete
  31. In spite of the multi posts I think Tom summed it up very well. I think Jack is driving differently during his testing, I know I was during mine and why in intend to repeat the tests. As I mentioned when coasting my speed would often creep higher than when I used regen. I know Jack still thinks you can't coast with regen, I guess I'll have to make a video showing "Zero" on the 840 display gauge while I'm driving. A few seconds of mild regen on a 16 mile trip means nothing, as you have clearly proven. Additionally, there is a difference with longer steeper hills and many shorter flatter ones, the shorter ones don't require you to brake to hold your speed. It's a simple and obvious difference. We have hills around here that coasting would get you 100+mph at the bottom if you didn't use your brakes. Jack I guess you think if you start at the top of Pikes Peak you'll have the same energy in the pack if you coast and use brakes or coast and use regen? Again, it is quite possible to coast with regen on the accelerator pedal, they are not mutually exclusive.

    JRP3

    ReplyDelete
  32. JRP:

    Why would you comment on what I might think? It would appear to be a full time job to manage what you might think and express it cogently.

    It is not that I think there is no gain in coming DOWN Pikes Peak. The problem is that after you've come down, you have to go back UP Pikes Peak.

    There are some very significant losses in converting the down energy iinto batttery energy end to end. But these are NOT accounted for in thte energy taken out to go back UP. So indeed you could generate much MORE coming down, (hypothetically, you can't really) and still lose the efficiency game because there are no such modifiers in taking current out to go UP.


    Tom has summed it pretty well. It gives the impression of "driving different" as accounting for the difference. That sounds damning, but it's actually quite true and unavoidably true. Cars with different weights and acceleration rates and so forth are naturally driven differently and indeed the manual/automatic discussion is pertinent to this. It's not about hypermiling, it's just about a car that drives differently. And you are GOING to drive it differently when it does. If I'm going downhill and rolling freely, and assuming it isn't a hill that makes me go 100 mph ( a dubious proposition by the way JRP) why wouldn't I let it roll on. And if I got to the bottom and flattens out and the car continues to roll for some distance with only minor decrease in speed, why would I accelerate?

    Same hill with regen, if I back off the accelerator or ride the brakes, it is going to go in regen, but that is not as efficient a use of the kinetic energy as simply rolling forward, and after the benefit of the hill, I still have to continue on with NO boost from kinetic, and as a result, I have to take from the pack energy to continue.

    My point is that the driving cycle is MUCH more variable than depicted in the little standard urban driving cycle graph, and your NATURAL reaction to the way the vehicle feels does indeed come into play - inarguably.

    Which brings up the obvious point, driving technique versus additional hardware? No brainer. Learn to drive the car. As I said, with 10,000 km on essentially the same car without regen, I can't NOT drive it that way. Although I would note that it's still for all intents and purposes a Porsche Speedster, and for much the same reasons as above, I also can't resist "taking off" in it as a sports car instead of a GEM.

    Net/net/net. No gain from regenerative braking. And no JRP I don't need a video. You perhaps do. If you can watch that 840 every second of your drive, your drive will likely be very efficient, and very short. It's a dangerous thing to even attempt. You shouldn't be on the road watching that gage. We had TWO in teh car, and Brian primarily monitored the gage. I was watching for trucks, cars, and pedestrians - more common here than on Pikes' Peak apparently. I seriously question that you can eliminate regen by constantly monitoring this gage and drive the car at the same time. I thought it impolitic to mention it when you first brought it up and thought it rather obvious. You either don't know if you were using regen, or you didn't get where you were going without a wreck. I do assume and hope there were no injuries.

    Jack Rickard

    ReplyDelete
  33. Sorry, It seems that I misinterpreted the AH you measured. if it actually measures the net Ah of the battery pack it is really an exclamation sign...

    "...Yes there were plenty of AH going out and coming back in during regenerative braking, but over the course of the drive, it appears to have nulled out to zero as a benefit..."

    I have a question!
    What if you pull the car with another car at constant speed and set regen to max, what are you going to gain? is battery pack going to charge up as it expected?! if so, what is the amount of energy that the whole motor-controller system generates as a generator. you can compare it with the total energy used to pull the car.
    This would eliminate driving dynamics, terrain, hills, etc. just to make sure that you have a considerable amount of regen Ah.

    ReplyDelete
  34. Jack, I have a test scenario for you. Or anyone else.
    I don't own an electric car.

    The Tesla long distance driving record from australia:

    http://green.autoblog.com/2009/10/27/tesla-roadster-runs-313-miles-on-a-charge-in-global-green-challe

    313 miles was the range.
    This was pretty flat ground. They didn't do the long distance test on a mountain, if they did they would have had to use regen breaking or mechanical breaking and would not have gone as far as they did.

    For a new test to see if regen breaking versus breaking breaking is viable you would need a long steepish hill that requires constant breaking to limit speed on the downhill.
    Then at the bottom of the hill you turn around and start back up to the top of the hill making 10 or so round trips.
    Doing this with regen on for scenario one and doing it again with regen off for scenario two I would expect to see that that when you check your batteries you would have saved more energy with the regen on.

    Jack, what do you think about that for a different way to run your test?

    This is a lot different than real world driving.

    ReplyDelete
  35. Jack, I actually mounted my 840 where it can easily be seen, not under the dash. Maybe you didn't want to cut into the dash of the Speedster or use a gauge pod on top, but that location is terrible, if you want to talk about safety, for any gauge.
    As for Pikes Peak, or any hill, yes you use energy getting up it, the same energy by the way if your car has regen or not. So you've used that energy on the way up in both vehicles, now you're at the top. If you have to use your brakes on the way down in the non regen vehicle, and you use regen in the same places on the regen vehicle, which will have more energy at the bottom? I think your problem was that you used regen where you would normally coast instead of using regen only where you would normally brake. We know coasting is more efficient so of course using regen instead of coasting is going to be a loss. However, using regen instead of braking should give you something measurable. I think you are very used to driving a vehicle without regen and using coasting very efficiently, and you are not used to applying regen efficiently. Both are learned behaviors, and you've had much more practice coasting. I'm not sure if you feel the case is closed but I'd urge you to keep practicing with regen and see if you can improve your numbers. I'm going to keep practicing using coasting more than I have since it is more efficient than regen, but regen has to be more efficient than braking. If sending amps into a battery doesn't do anything then how are we charging our packs?

    JRP3

    ReplyDelete
  36. JRP3,

    Isn't that the point of the entire original post? There is a paradox in regard to regen. Apparently if you use regen on "brakes only" or "decel and brakes" there is no benefit over a vehicle without any regen.

    I too would love it if you would perform the test with the same protocol as outlined by Jack. It seems as though you are in a good position to repeat the test. Not a similar test, but a test using the same protocol, and see what results.

    ReplyDelete
  37. JRP:

    Your logic is sound. I agree it makes sense. We can type and type and type and break into VIOLENT agreement. But it doesn't come out that way.

    I STILL struggle with why the mode with regen ONLY on the brake pedal doesn't provide a gain. The answer is fairly obvious. it doesn't provide a gain for EXACTLY the same reason it doesn't with regen on the accelerator. But it just more viscerally seems like it SHOULD with regen on brake only.

    But it still changes the driving dynamics of the car.

    Potentially, if you are so wed to this you want to prove that it does in fact work, I suppose you could find some form of terrain or lap where it does, all downhill for example. But if you lay out a representative lap of real world driving conditions in an honest fashion, even in your more spectacular terrain, you should get the same results. If you do, we'd like to know. If you don't, we'd still like to know. But you do need to be able to turn the regen OFF.

    Jack Rickard

    ReplyDelete
  38. This goes to JMS comment on a specific terrain selection to show regen working. We're not trying to show it working. We're not trying to show it NOT working.

    This started with a bit of a casual observation on the mini that it didn't seem to exhibit much difference regen on or off.

    We went to some trouble to redesign the Speedster to incorporate regen, and we got the same range with the same size battery pack as the original speedster.

    So this starts with the question, "What happened to the efficiency gains we were supposed to get from regen?"

    We ran the test. There were no regen efficiency gains. We're not THEORIZING as to WHY that was the result, not as to WHETHER it was the result.

    Of course we'd like to hear of different results using the same methodology. But what does it tell me if you get different results with different methodology (even if it makes sense to you)?

    We'd also like to hear of identical results - again with the same methodology. It's not very rigorous or exacting guys. Layout a circuit of sufficient length to cancel variations in lights and traffic interaction. Turn regen on and tune it as you like it. Measure total AH across course. Turn regen off, repeat lap as identically as possible. Measure total AH across course.

    Did I say kWh? Err.. no. And I have many good reasons for NOT specifying kWh. Did I say "close your eyes and pretend you turned regen off, because that's easier for you to do?" No. Not really.

    kWh varies with your voltage. And the voltage varies with both load and SOC. It is quite possible to totalize it as with AH, but not with any of the equipment we have installed. So kWh to AH is like comparing the speed of a train to the smell of frying bacon. I'm sure everyone has a preference, but I can't quantify which is better or even quantify what either is given both.

    As voltage varies under current load and as that effect varies under SOC, we would have to have the same size pack, the same course length, and a half a dozen other things to control this. If we are counting AH and you are counting AH, it doesn't matter how MANY AH you take per mile or for your drive, the ratio of AH to course length compared in the two modes should be either similar or different. Information either way.

    To JMS concept of a different test, if it is a lot different than real world driving, why do we want to do that? I would be interested in a suggested alternate test that was a lot MORE like real world driving, but I can't picture what it would be.

    What I think might be more productive is to do it with a different car. Today I did the Mini Cooper with regen on. All the regen settings are different on the Mini and it has a different feel. The brakes are almost nonexistent with regen off, which is going to make tomorrow mornings drive a BIT interesting. But if I can pull it off without wrecking the car, it will be telling.

    The thing that continues to astound me is that this could have been missed. What I keep coming back to is that I can't find anyone who's done it. The BMW Mini-Es have strong regen and a lot of people very interested in range. But they have no switch to turn it off to see. They get what the manufacturer gives them.

    JRP and Tom with the Suzuki Swift are running the same basic motor controller configuration, and could run this test if it at all interested them, and on different terrain at that.

    Jack Rickard

    ReplyDelete
  39. It seems to me that if you have to go through all this rigamarole just to see even a slightly NOTICEABLE difference between regen and non-regen, at the very least it seems as though the gains would be too minimal to be worth the trouble and expense. Of course, it's a moot point if regen continues to show a loss in the end....

    ReplyDelete
  40. I agree. As I said, the very BEST we can say of regen at this point is that it does no harm and offers no gain in efficiency. But it does come at a cost and indeed does include quite a bit of stress and strain on all the drives system components.

    In some ways, I'd prefer to concentrate all that stress and strain on a $18 set of brake pads.

    Jack Rickard

    ReplyDelete
  41. Consider a fully charged battery pack with x AH. start consuming its power for example by 100 Amps. meanwhile try to reverse the current of 200 amps for 3 seconds in every one minute. how much do you think you are going to gain.
    I think all regenerated power is going to only heat up the battery since charging a battery is like packing your suitcase; Substantially it can discharge in 20 minutes but taking hours to charge up.
    You do can regenerate some current but you cant charge your battery pack with that pulsatile flow efficiently. also I think regen is going to lower your battery-life.

    ReplyDelete
  42. Over thirty years ago when I was growing up on the flat prairie in Illinois, I worked in a shop where we occasionally did some repair work on an old farmer's equipment. My boss once pointed out: "You see the brake shoes on this trailer- like new and its at least 15 years old."

    In those days, organic brake linings were good for about 10,000 miles, 20k tops by looking after them. On a working trailer, it was an incredible sight. "I asked John about it, because I've never seen that before and I thought they might not be hooked up properly," my boss continued, "and John told me he hasn't replaced the linings on the brakes of anything he owns, from his pickup to the combine, for as long as he can remember."

    "How is that possible?" my 16-year-old freshly minted driving mind wanted to know.

    "He doesn't really use the brakes."

    YMMV

    ReplyDelete
  43. Jack, was the pack at the same SOC at the beginning of each run? If not, amp-hours would vary slightly. For example, if the average pack voltage during one run was 3.25V, but was only 3.15V during another run, AHs would differ by about 3%.

    Bill

    ReplyDelete
  44. Dimevision:

    I'm not sure I agree with all that. These cells can charge at up to 3C. They are NOT lead acid cells and do not share these behaviors. I've also observed much angst about regen with a fully charged pack. It isn't a problem.

    The batteries can do fine. And yes, we are generating regenerative current. But across the overall drive, it does not appear to matter.

    Jack Rickard

    ReplyDelete
  45. The pack was fully charged for all the tests.

    Jack Rickard

    ReplyDelete
  46. To add fuel to the fire, and not to give away this weeks video or anything, but we did repeat the test with the Mini Cooper.

    The results were indeed MORE damning of regenerative braking. We did the first run of 48 miles in 47.1AH and one hour and 28 minutes. This had a mild regen on the accelerator and a quite strong regen on application of the brakes.

    On the second drive, this morning, we cut out regen entirely. We completed the drive in one hour 25 minutes - 3 minutes faster and averaged a little over 33 mph. We also did the drive in 43.8AH. That's 7% BETTER economy WITHOUT regen.

    The evidence is piling up on us here pretty quickly. This is not good. This is very very very not good.

    Jack Rickard

    ReplyDelete
  47. Jack, Have you talked to anyone at Tesla or Such? They should be up on this. It is simply amazing to see this play out. I agree if it's not real world driving it doesn't really matter.

    ReplyDelete
  48. JMS

    Not really. I've been conferring a bit with Dr. Illah Nourbaksh. He's been doing a great deal of work on regen efficiency and maximizing it with Ultracapacitors.

    Instead of dismissing it, he said the testing was "spot-on" and to his knowledge no one had published anything on an end-to-end drive circuit testing various configurations of regenerative braking or comparing it with no regenerative braking.

    They are setting up a test track to work further on the Ultracap concept and he indicated he was further interested in this and might make some test modifications to duplicate the concept.

    To tell you the truth, I don't know WHAT to do with this information. It just seems surreal. But the meter is pretty much the meter.

    We've done rear wheel drive and front wheel drive, different weights, different motors, different controllers and different regenerative configurations. So far, it has come out that no regen is the best regen, every time. And I can't get the needle to budge on that.

    Jack Rickard

    ReplyDelete
  49. Jack,

    after the Mini displays a similar trend as the speedster, would it make sense to let someone else drive the 3 test cycles to eliminate some kind of experimenter bias?

    I can hardly imagine a physical reason for regen costing energy (not discussing wether it makes sense or not) exept for braking more often or harder with regen than without it. So this unnecessary braking has to be compensated for. Without regen you coast more.

    Regen is dependent on braking. The less you actively brake but coast down, the less energy is left for the regen to use.

    So from an easy or hypermiling kind of driving styles perspective -supported by a low mass (low energy storage) car - I think your results actually start to make sense.

    Thanks for dipping my nose into this.

    regards
    Philip

    ReplyDelete
  50. Philip:

    Perhaps.

    We took the transaxle out of the Speedster today. We'll send it off for rework Monday so its out of commission for a week or so.

    i'm hesitant with the Mini. The steering is a bit stiff and with the regen off you really have to put some leg into the brakes. They brake, but they're very stiff.

    I'll have to think about that...
    Turning someone at random loose with one of these cars is a little frightening.

    Jack Rickard

    ReplyDelete
  51. Jack,

    Clearly what you have proven here is that if you don't use your brakes you waste less energy. I hardly find that startling or surreal! As has already been commented, the only energy you can possibly recapture with regen is energy that you would have otherwise wasted in braking.
    Let me pose a scenario. I have 2 cars of the same make parked at the Eisenhower tunnel in Colorado. The blue one is stock and has an empty fuel tank. The red one has been fitted with electric drive with regen braking. A million dollar prize goes to the driver of the car that reaches 1 mile past the bottom of the hill first. You must not exceed 60 MPH. Which car do you choose, and why?

    ReplyDelete
  52. I think you confuse people when you imply regen is a myth. You leave the impression that from a physics standpoint it doesn't work which obviously you know is not true. I think it is much more useful to argue whether or not it is beneficial in real world driving scenarios with real cars and real people. Although I have great respect for your scientific methods, I would not agree that your car and your method of driving is close to the real world of cars and driving in the USA today. I would argue that anyone who drives an electric car that they "built", drives it completey differently than the typical driver and probably quite differently than their own gas powered car. It's part of the game.

    ReplyDelete
  53. That last post was mine but it picked up my google ID. Sorry, rookie poster!

    ReplyDelete
  54. I will setup a test with a 30A constant current source, a couple of my LFP90 and a 0 to 60A constant current load, where I can set duty cycle and simulate different cycles. If there is some delay in the system where the cells are not charging, it should show up.

    I have not the time right now to set up the test, but it has been written on the "todo list".

    Martin.

    ReplyDelete
  55. Jack:

    Clearly any regen setup that adds drag is a net energy loser. The "single pedal" configuration adds drag when you would otherwise be coasting, even if you have a soft touch and don't think you are using it that way. Seems pretty obvious how the mileage comparison to no regen would come out there.

    The more interesting scenario to me is regen on braking only. I don't understand how, other than driving style, this scenario could cost more energy than putting THE SAME amount of drag on the car using the regen as you would with the service brakes alone, or a combination of both.

    Any measured energy loss from braking-only regen to regen disabled is either in the noise, an inability to really entirely turn regen off if enabled (an equivalent of the slight drag in the service brakes that is impossible to entirely eliminate,) or there is some peculiar thing going on with these short recharge bursts that is resulting, not in a very small net charge as expected, but in a slight discharge. All possible, but its most likely to be testing variability (noise.)

    At this point, I think only two regen modes are worth continued comparison: brake-only regen (true regenerative BRAKING where the car is otherwise free rolling,) and regen disabled.

    There's another factor in here we haven't gotten to yet, because you're still testing regen with the same drive system: Powertrain efficiency.

    What happened to all that talk about AC systems being 90+% efficient, with motor ratings in the high 90s, while lowly DC systems are usually considered much lower? How does that square with the enormous amount of heat your AC controller and even your AC motor are making, while your heavier DC powered car seems to do the same job and make less heat? At the end of the day, this is the real regen question- not whether it is more efficient turned on or turned off, but whether a system that has it is
    more efficient than one that doesn't, and can be cheaper and maybe even a little smaller for the same torque delivered.

    One more AC mantra I want to ask you about- high rpm operation. Every one of these AC motors is rated very high, up to 11,000 rpm. In an ICE, you get more power with more RPM, which is great because you need it to overcome skyrocketing aero drag. AC motors do just the opposite, and it seems very clear that these motors just aren't going to spin to 10k at highway speeds because they run out of torque well below that, like 3-4k tops. I'm a gear head; I've learned a ton from racers who spend more time with the gearbox apart than the motor, and don't get how these AC motor torque curves are going to provide useful high rpm in the car. What are you learning about that in the Mini?

    To a rodder like myself,a DC motor with higher torque and better gearing would seem to be a much more effective way of moving the car. Probably more efficient, too. That's the fair comparison regen should be up against...

    TomA

    ReplyDelete
  56. Well, we're mixing a lot of apples and oranges here.

    Dan and Karen, your ability to see and comprehend has no impact on what the meter shows. It is not in the noise. The Mini test pretty much closes it out. As to whether my driving, or my drive are "real world" or not I won't comment on . THe answer is already pretty obvious.

    Martin Zacho. This has nothing to do with the batteries. I happen to know the cells are more than capable of absorbing this energy and the losses are actually fairly minimal compared to previous battery technologies. But it misses the point. The batteries have NOTHING to do with it. You understand we are measuring CURRENT and we are measuring it in two directions - OUT and IN. The meter totalizes this for a NET number. What the batteries do or do not do with it is entirely immaterial and we're not measuring it at all.

    Tom and AC RPM. AC Motors can feature higher RPM limits because they do not have brushes riding on a commutator. There are two bearings and a shaft and if they are good quality bearings and a true shaft 10,000 or 12,000 rpm are quite easy to accommodate. This is not like a gasoline engine with parts moving in different directions. This is just rotary motion. On DC Series motors, the brushes riding on a commutator provide a point of failure and the motor simply flys apart at a lower RPM.

    As to power, it is limited to the power you put IN to the motor, voltage times current. This can be distributed over 1 RPM or 10,000 RPM but it is the same power. Motors have an efficiency curve and so we'll see torque maintained out to a certain level but then it begins to fall off as RPM increases - ALL electric motors.

    The difference in efficiency bettween series DC and AC induction is pretty complex. Series DC actually has better magnetic coupling but experiences losses in the brushes and commutator again. On balance the AC induction motor generally is considered more efficient, but it is by a percent or two in practice.

    ReplyDelete
  57. Apples to oranges on heat. We did have a blower for the motor in our first Speedster and we've come up with a slightly better way of doing that in Speedster Part Duh. The controller cooling is a different kettle of fish. We had the Kelly controller in the first speedster mounted on a largish aluminum plate and we hung it vertically with the end of the plate in the air stream below the car. This was not viable with the CUrtis controller which we really wanted on the shelf above the motor - under the hump. There was not room there for an air cooling system, which would have been simpler and less expensive. So we had to devise a way to MOVE the heat to somewhere we could get rid of it -- liquid cooling. Actually, we're seeing more cars go to liquid cooling of both motor and controller and in fact we do this in the Mini Cooper.

    The regenerative braking is very much a myth busted. It IS about driving style, just not mine. This is a classic of taking a lab phenomenon and ASSUMING that it extends to a much more complex real world application. In this case, it makes a lot of sense that ti would directly apply if you consider a drive simply acceleration and deceleration in alternate processes.

    Unfortunately, or fortunately, a real drive is much more complex than that and every decision made modifies every subsequent decision made with regards to acceleration and deceleration rates, duration, and results. Throw in random ilght changes and pedestrian lunges and it becomes really quite the little model. By using a drive of sufficient length, the differences are largely cancelled out.

    Is the glass half empty of the glass half full? We did three drives in the Speedster that were an hour and a half long but resulted in a difference of less than 0.5AH. In the Mini cooper, we timed the drives more closely and did one at one hour 25 minutes and the other at one hour 28 minutes, again within 3.3 AH of each other. I find that remarkably consistent. The problem is, the regenerative braking should have made a difference - that's what we were trying to measure.

    It is no longer debatable WHETHER this occurred. It occurred, and was repeatable. As to WHY it occurred, asssuming I was somehow MAKING it occur is childish. Why would I do that? Why would I even want that?

    When you change the drive characteristics of an automobile, you WILL drive it differently. But the two Mini drives were net/net within 3 minutes of each other on an hour and a half. Had I done the Mini drive with regen in 1 hour 28 minutes, and the drive without regen in 1 hour and 57 minutes, I could get with this theory. As it is, I cannot.

    We drove the car. Then we drove the car again. The meter reads what it reads. That it doesn't match what you or I expected it to read doesn't budge the needle.

    I think it rather obvious that this matches real world driving conditions better than simply counting AH out and comparing them to AH in to determine the efficiency gain you OUGHT to be getting.

    Jack Rickard

    ReplyDelete
  58. Jack,

    Is there no way the AH meter is diluting us? That it goes astray when the current direction changes? Have you checked the SOC as a second independent measurement?

    Regards, Jan

    ReplyDelete
  59. I kind of agreeing with Jan. There must be something we are not getting here.

    Assuming your measurements are correct it raises the question of why the car is using more energy when regenerative breaking is enabled?

    You said it yourself and I agree that the battery can take the charge at almost any rate - so if you are putting charge into the battery and getting the same result, then you must be using more charge on the drive.

    On a test run I did with my citystromer and the old batteries, I used 28Ah and the regenerative breaking added a single Ah the other way (the charge efficientcy is much poorer with LA). The run was not made to prove anything about RB, but just to serve as a reference drive with the old technology. In august I can do the same trip with the LFP's installed.

    Martin.

    ReplyDelete
  60. Jack,I have followed your work for some time and I admire your methods and most of all the fact that you don't just talk about stuff, you actually get out there and do it! I especially like your work on battery balancing and I agree that this is a myth busted. In the case of regen however, it seems a bit arrogant to run 3 tests with 1 car and 1 driver and announce regen absolutely does not work. As I see it there are four possible explanations for your test results.
    1) Your test did not include significant opportunity for regen power to be recovered (ie use of braking)
    2) Your regen system is working at an efficiency level that is negligible (the current heading back to the battery gets lost in heat somewhere)
    3) Your measurement system did not collect data properly
    4) You debunked the first law of thermodynamics (might want a few more tests to confirm this one)

    I once had someone proposing a perpetual motion device to me and I told him it couldn't work. He asked me if I had tested it and my reply was "I don't have to because it defies the first law of thermodynamics". The same applies here. I don't dispute your results, just your conclusion. BTW - did you want the blue or red car in the million dollar challenge?

    I really do admire what you are doing and hope to be able to come by your place someday and see it in person. I fly a Cirrus SR22 and my buddy and fellow pilot Tom who has (unlike me) actually built an electric car (as well as a couple of airplanes) and I are talking about flying up from south Florida to see your operation if that would be acceptable. KCGI flight plans at about 5 hours from West Palm Beach.

    Warm Regards,

    Dan

    ReplyDelete
  61. Certainly Jack has shown a fairly small difference between regen and no regen. The disconnect comes when saying wasted brake heat is more efficient than regen, we know that can't be possible. So the question is can we drive differently, using regen, to effectively use that energy we know will otherwise be wasted? Jack seems to be saying that a regen car will always be driven less efficiently than a non regen car. Does that have to be true?
    From Tom with the SwiftE:
    "Yesterday I drove at average speed of 38 mph 6.3 miles up a 1600 ft change in elevation (low speed due to lots of switchbacks), about 4.8% grade, with outside temperature of 91 F. Motor/controller start temperatures were 55/53 C (drove about 7 miles to get there), and at the top were 61/56 C (max motor is 120 C, max contr is 85 C). Cell temp was 94 to 96 F at bottom and 95-97 F at top (measured with IR sensor on Extech EX830). These are in insulated boxes, but I had a 160 cfm bilge blower circulating air through the boxes (too noisy, need to get a different fan with lower tip speed). Current draw going up was about 135 to 210 A. Pack V going up was 114V at 200A draw, and 115V at 170A (about 119.7V at rest at this point). Current draw ranged from about 135A to about 210A, about 160 ave. Regen going back down was 21% of Ah used going up - used 25.6 Ah going up, gained 5.5 Ah going down. Did not use the mechanical brakes at all going down."

    JRP3

    ReplyDelete
  62. Jan:

    We have checked the meter and checked the meter. We are now on 2 cars and 2 identical meters. I can discharge 80 AH driving the car. When I recharge the car with a charger, it is within an amp hour of back to zero.

    The meter is working. It is working quite welll in fact.

    Jack Rickard

    ReplyDelete
  63. Martin:

    I agree. If we take out 47 AH driving without regen. ANd then we do regen and put 5 AH back in, then we must be using 52 AH to get the same result.

    You guys keep focusing on the batteries. The SHUNT is where we are measuring. We don't care a wit what the batteries do. We have current going through our wires and we are measuring it.

    Let me propose another thought experiment. Let's say we are on flat ground and driving 5 miles.

    In case 1, we set the throttle to drive the five miles at a precise speed of 45 miles per hour. We use 5 AH.

    In case 2, we speed up to 50 miles an hour, then we use regenerative braking to decelerate to 40 miles per hour. Then we accelerate to 50 miles per hour. Then we regen down to 40 miles per hour.

    How many AMP Hours net do you imagine we would use in case 2, if we used precisely 5 amp hours in case 1?

    It takes a seriously greater amount of current to accelerate from 40 to 50 than it does to maintain 45. When we regen, we also have signficant losses in converting the kinetic energy of the car into current. It is NOT 1 for 1. Forget the batteries. COnverting the kinetic energy of the car into current is what I'm talking about.

    You would be very lucky to get back to a net of 5 AH. ANd you would certainly use more than 5AH not counting the regen.

    This is where its' going guys. The meter works.

    Jack Rickard

    ReplyDelete
  64. DAW:

    Now let me get this straight. You are calling me ARROGANT because I draw an absolute conclusion from an absolute test, you compare it to perpetual motion myths, and you have never built and do not own an electric car? ZBut yuze read about them thar thermodynamics and so you're pretty sure I'm full of shit?

    I wish I could get smart like you with that level of effort. Typing I mean.

    WHat DATA do you offer to refute my claim. Your opinion, and a VOTE don't move a needle a whit. If I had to count all the times I've learned something that a whole bunch of people knew differently, it would be a full time job. All that thinking and voting won't move me any more than the needle.

    If you have a better thought experiment to EXPLAIN the results I'm keen. If you have any DATA or experimental results to REFUTE the claim, I'm keen.

    Actually I'm completely disinterested in "opinions" on the measured results. They're already not only real, but confirmed with a second car.. WHY we get these results is a bit interesting and I do not have more than a vague theory myself.

    If your clumsy attempt to grossly misapply the laws of thermodynamics to this case puts me in your mind in league with perpetual motion goofies, I am truly and I mean very truly offended.


    As to your four scenarios:

    1. There was not only adequate braking, it was continously intermittent. This was urban driving with stop signs, trucks, traffic, stop lights, hills, and all manner of it. IT contained HUNDREDS of braking events over an hour and a half.

    2. The current and its relationship to batteries is entirely moot. We could easily measure current at the shunt in both directions and we were getting it. The regen was quite aggressive. We could stop the car entirely in the case of the Speedster by removing our foot from the accelerator. Accelerator regen was MILD on the Mini but very strong on the Mini. In both cases we repeatedly observed regenerative current levels in excess of 100 amperes.

    3. My measurement system collected data presicesly as I wanted it to, and in an accurate manner. How do you, without ever being in the room with it, pronounce it improper?

    4. You clearly have no concept of the first law of thermodynamics. I'll not attempt to debate it with you.

    I'm the large hangar on the East end of KCGI. We'd love to have you.
    I have two DC-3s, a Beech Staggerwing, and an assortment of others you might find of interrest.

    Jack Rickard

    ReplyDelete
  65. Jack,

    I think most agree on this blog that regen on the throttle isn't the most brilliant idea. Most are wondering why regen on the brakes doesn’t work. Or worse: gives a negative effect.

    I’m still not convinced the AH meter is perfect. Or nearly perfect. That 1Ah off, if that is cause by turning the current direction, it could be a lot more during a trip with regen. I don’t say that is the case, just speculating.

    You did observe during regen braking a decreasing of the AH’s on the meter?

    Regards, Jan

    ReplyDelete
  66. JRP:

    You're fighting to establish the wrong thing. We get amp hours out and we get amp hours in as well. Now go drive an hour route with ups and downs and not. Write down the AH net. Now turn regen OFF. Go do it again. Look at the two numbers.

    This is WHY this has gone on so long? You cannot even CONCEIVE that it could be otherwise. You can measure current usage. You can measure current gained. And so it MUST be a gain. You can measure it and you can see it. But you never compared it to not doing regen at all.

    Until you leap this mental hurdle, I can't have much of a conversation with you. If you already KNOW how it will come out and so needn't bother, as your fellow voter Dan does, there's nothing I can do with you.

    I started with a problem. No apparent range extension between two identical cars - one with regen and one without. That is an ironclad observation without a conclusion.

    I moved to perform a defined test on ONE of the cars, which I could do WITH regen and WITHOUT regen. I got a result. I know something you don't. Apparently I know something you can't even conceive of.

    Your actually in a position to duplicate this. So far, instead of doing so, you've performed TWO completely different experiments that had no hope of even relating to this one, and concluded what? Neither held constant for anything. I already know regen produces current and can observe it. It is childs play.

    Now I've extended it to an entirely different car, with a different drive train, different weights, front wheel drive instead of rear, on the same route and gotten the SAME results. Not only no benefit from regen, but a PENALTY. Which DID show up in the first test as well.

    Now you drive UP a hill and back DOWN the hill and report the AH used and the AH recovered. It's not even under question. In your mind, it is OBVIOUS that this extends to a beneficial efficiency gain from regen. I already know it does not.

    Now the only way to convince you is to get you to run the test. And you just are NOT going to do it. You already know better. And you want a DIFFERENT test to prove it.

    What am I supposed to do with that?

    The temperatures are interesting. I was doing some AH/speed tests on the Speedster Part Duh. I was concerned that we had overengineered a little on the water cooling system. At 70 mph going up a long uphill on the Interestate, we were doing 215 amps to hold speed for about a minute. Temp reported on the 840 reached 82C on the controller. As soon as we backed off it plummeted to 63C. The cooling system is working very well, and was very necessary. In fact, we probably will look for a slightly larger heat exchanger and fan. In the Rockies, we could easily get into an overtemp condition with current limiting on a long uphill.

    Jack

    ReplyDelete
  67. "I’m still not convinced the AH meter is perfect. Or nearly perfect. That 1Ah off, if that is cause by turning the current direction, it could be a lot more during a trip with regen. I don’t say that is the case, just speculating. "

    We're within an amp hour on an 80AH discharge charge cycle, indicating the meter is totalizing in both directions accurately. Hitting EXACTLY the same SOC on the cells during charge is squishy to within the 1AH. I attribute the 1AH difference to this, not to the meter.
    "
    You did observe during regen braking a decreasing of the AH’s on the meter?"

    We did so observe. But only on a long downhill. We observed positive current flow any time regen deployed. But the meter actually ticking backwards on the AH counter takes a long regen.

    Jack

    ReplyDelete
  68. The meter is not perfect. It's not nearly perfect. BUt I don't know what we are saying here. What does "perfect" mean to you?

    It measures current. At the bench, it matches the bench instrumentation to within 1 amp at 300 amps.

    IT totalizes current into AH. It does so both directions. As it is the same shunt both directions, and the same A/D converter, it would have to be in software for there to be any difference. I find that hard to believe.

    How accurate does it need to be?

    But all that doesn't matter a whit. It was the SAME meter on all three runs. It was the same TYPE of meter on a different car on two runs.

    If it was off, why would it be off differently on one run than another?

    There's no scenario here to type ourselves smart about. If it had a 300 amp disparity 100% of the time, it's a relative measurement between the modes. It could read in potatoes per fortnight and drop every fourth potato. The drive was the same length, peformed in very nearly the same time, and in the same car. SHOW ME THE REGEN. There was no gain. There APPEARS to be a regen PENALTY.

    Jack Rickard

    ReplyDelete
  69. I really wish I had a car with regen right now to run this experiment. Unfortunately my car is series dc. Darn! I really hope we can get somebody to repeat this on another car...please, somebody step up to the plate.

    ReplyDelete
  70. Jack,

    I am not talking about the batteries ! I am talking about where you put the energy (!) you are not putting into your disc brakes !

    If you dump it into the brakes as heat, then it is lost, if you put into the batteries (or supercaps) then you get some of it back.

    I thought you did to equal drives, but you just did to similar drives, which used the same amount of energy.

    It is your test that does not shows what you thought it did. With out the data I can not tell you where you do it wrong, but you do.

    I will do a couple of runs and let people see the data. I can log when the brakes are applied and have a GPS so I can show the route.

    I agree on the point that you do not recapture any energy if you just use up and down the hills, but only where you use the brakes in order to slow down - you will always loose about 10% in the batteries (in terms of Wh) and something similar in the drive train. It might be because we are focusing on the Ah and not the Wh used. I think you should be looking at the Wh that you put back into your battery pack after each run and don't care about the Ah on the meter. They only tell you when the battery is fully charged, but nothing about the energy you have used. At the end of the day, the only thing that counts is how much energy you will have to supply to your car from the plug in the wall.

    Anyway, I do not think this is going anywhere without more precise data - and I will try to supply those. GPS log and when the brakes are applied etc.

    Thanks for your thoughts.

    Martin.

    ReplyDelete
  71. And by the way.

    I do not think that regen matters (in my case) since I do 95% of my driving intercity, where it is insignificant amount of energy regenerated in any event.

    Martin.

    ReplyDelete
  72. Palmer:

    Your circling the whole point. If REGEN provides no efficiency gain, then the need for the expensive 3 phase controller and AC induction motor goes away, and we are left with the much less expensive series DC powerplants that would do the job just as well.

    There's some minor advantages in AC induction motors, but instant torque isn't one of them.

    Jack Rickard

    ReplyDelete
  73. "
    If you dump it into the brakes as heat, then it is lost, if you put into the batteries (or supercaps) then you get some of it back.

    I thought you did to equal drives, but you just did to similar drives, which used the same amount of energy.

    It is your test that does not shows what you thought it did. With out the data I can not tell you where you do it wrong, but you do.

    I will do a couple of runs and let people see the data. I can log when the brakes are applied and have a GPS so I can show the route."

    If I have one more person offer to show me how regenerative braking DOES in fact generate current, Im going to scream. Clearly you guys have neither watched the video or read the description above.

    We DO generate positive current into the cell pack when braking. We don't care. We don't care how much it is.

    Whatever you take out, and whatever you put back in, you WILL be left with a net amount out of the pack implying a net remainder in the pack. In theory, regenerative braking should provide a LESSER net out and by extension a larger amount in the pack for any given drive. The two drives WERE identical. Ok, I missed a turn on one drive and it's about a kilometer less. The AH per kilometer remained true.

    We've now duplicated it on a second car. Same result. No gain from REGEN.

    Yes, it would be nice if some others, who perhaps do drive differently, or on different terrain did confirm this. Duane Ball reports no difference on his Spyder on a 10 mile run. Beyond that, a lot of angst about how I must be missing something. But no help. I don't need GPS data.

    What might be helpful is if you go for a long drive with your normal regen. Note the net AH removed. Now recharge the car and go on the SAME drive wiht the REGEN TURNED OFF. Note the net AH removed. What were the two numbers?

    It's really not any more compllicated than that.

    Jack Rickard

    ReplyDelete
  74. Hey Jack, My apologies for offending you! Never intended that. As I said I admire you and respect your work. What I hope we are having here is a friendly albeit spirited debate on what I believe to be an interesting and quite important subject.

    I think this has been danced around a time or two but let me say again, mechanical brakes are by definition 0% efficient in recapturing usable energy. Can we agree on that point? So if you go off and do a test where you get a lower efficiency than mechanical brakes(a "penalty" as you say) it can't be true period! What you are saying in effect is I have negative efficiency which of course is a nonsensical concept. There is by necessity a different explanation.

    Martin makes a very good point that I think may be the key to this conundrum. Amps is NOT what we need to be concerned with. It is in fact about ENERGY. The first law (which for some reason you don't think I can handle) is about conservation of energy, not power. As you well know there are 3 components to electrical energy; Current, voltage, and Time. Current multiplied by voltage is power, and power for a period of time is energy. It seems to me that you are missing the voltage ingredient in the recipe since you are just talking about Amp Hours. Perhaps the voltage changes during RB and recaptures more energy than you are accounting for when you look at AH alone. I realize that if voltage is constant through the entire driving routine it can be factored out for comparison purposes, but is it really always the same even during hard accel and hard braking? Seems unlikely to me since clearly voltage must increase if we want to drive more current into a given load, and voltage will decrease if we pull more current out of a given load. Do you think it is possible that this explains the apparent impossibility you are reading? Please direct me to a resource where I can understand the particular electrical setup you are using.

    In the pursuit of truth, wherever it leads,

    Dan

    PS - Red car or Blue car, I need to know!

    ReplyDelete
  75. Jack, I cant justify spending $500 on a programmer to turn off regen to do this test. I might be able to justify renting one again. Problem is I'm not sure why I need to. You say regen is a penalty. OK. So if I do a run trying for zero regen, even if it's not perfect, it's still going to be a lot less regen than my normal driving, so it should be fewer AH's used according to you, yes? If not, then why?

    JRP3

    ReplyDelete
  76. Jack,

    I think you misread some post above an misquoted me. I think you were referring to a post from Martin Zacho. I'm on your side in this argument. I see no fault in your test or the results. I am curious as to why such a paradox exists. There may be a simple explanation and we are both overlooking something. The best way to get to a resolution is for some 3rd party to try and repeat your test. I would do it, but I don't have a car with regen. In my mind, the AH vs wH may have something to do with it, but I can't resolve it in my mind just yet. More testers may help, but everyone keeps arguing that "regen exists therefore it must reduce consumption".

    ReplyDelete
  77. I think what has not been alluded to is the overall velocity.
    Jack put me on the this track when he described accelerating / decelerating for a block.
    Let us concentrate on that and ignore uphill/downhill.
    You set off from zero velocity and after a block (500m length) you are travelling at 50km/h. whether you increased speed quickly and then glided or if increased speed gradually the energy used will be about the same.

    So you are travelling at 50 km/h and you now want to stop the car in a blocks length.
    For method I let us stipulate that the car will, if energy is no longer applied, stop exactly in the block's length and also this deceleration will be linear. Method II is to keep driving for 400 m at a constant 50km/h when energy will be applied and then in the last 100 m the speed will drop from 50 km/h to zero km/h linear when energy will be regenerated.
    I)dropping from 50-0 km/h in 500m.
    no energy used or regained.
    Time is 0.5km/25km/h = 0.02h (72sec.)
    IIa)constant speed for 400m
    Time is 0.4km/50km/h = 0.008h(28.8sec.)
    IIb)dropping from 50-0 km/h in 100m
    Time is 0.1km/25km/h = 0.004h(14.4sec.)
    So with I (cruising - no energy used) takes
    72sec. or 25 km/h
    now II (gas pedal and regen) takes
    0.008+0.004 which is
    43.2sec or ~42 km/h
    The energy required for I is zero. Energy for IIb is energy regenerated which would be that required to increase speed from 0 to 50 km/h let us call that one unit.
    Energy for IIa is energy used to increase from 400m/500m of the 50 km/h which is 80%.
    thus we have used up about 80% of the energy regenerated.

    This is of course just one scenario but fits with Jack's data. It seems to show that when time is of importance then regeneration is of great importance. It seems as if, without regen, the driving style has a significant effect on energy consumption whereas with regen the style of driving has less impact on energy use. And thus regeneration is for this reason not to be ignored.
    CyrilMaurice

    ReplyDelete
  78. Holy Crap!! they did fill your mail box with hate mail!

    I think these guys need to stay on the DIY forums, typing themselves smart.....

    I don't listen to anyone arguing the point if they themselves have no data, nor EV

    Go JACK!!! love your show!!

    Simon

    ReplyDelete
  79. A few observations...

    Only one test of each type with the car doesn't give any data on the "noise." I never let my physics students get away with so few tests of each type. Especially with so many variables as Jack pointed out in the beginning.

    Energy (Wh) should be counted and compared not Ah. Probably the simplest would be to measure what comes out of the wall since we are interested in a difference in energy consumption.

    Only three regenerative braking methods were tested. The problem with the regenerative braking tied to a pressure transducer on the brake line is that regen doesn't get applied without the friction brakes also applied. This is a likely reason for the lack of an apparent benefit to this mode of regen. The regen on my SepEx based Gizmo is controlled independently from the brakes. The previous owner abandoned the pressure transducer idea. I should have just sent my transducer to Jack to use and save him the cost of his. Sorry Jack, I should have emailed you and asked you if you wanted it. If regen is hooked up so that full regen can be applied without the friction brakes your results will be different but not by much.

    On the AC system, does regen current remain positive, into the battery, all the way down to nearly stopped? If not, regen becomes a loss below a certain speed. On my Gizmo the current stays positive only down to around 10-12mph and then it goes negative down to about 3-4mph before dropping out. Just before my controller drops the regen the current out of the battery is about 25A. For this vehicle, like Jacks Mini, has undersized brakes and they wear out really fast. Getting variable regen set up has made the brake pads last over 3 times as long so far. Furthermore, I can stop much faster with regen and friction brakes.

    On ProEV.com they talk about how the use of regenerative braking allows them to make an entire race without running out of battery energy. Early on in their testing it meant finishing vs not finishing a race. More recently they have installed a larger capacity battery pack so having regen hasn't been as much of a concern but it definitely extended their range.

    Finally, Jack your statement that, "But what IS apparent immediately is that regenerative braking in ANY form would appear to have no efficiency additions AT ALL." is simply not true since you haven't tested all forms of regen. I provided another form of regen to test above.

    David D. Nelson
    www.evalbum.com/1328

    ReplyDelete
  80. "I think this has been danced around a time or two but let me say again, mechanical brakes are by definition 0% efficient in recapturing usable energy. Can we agree on that point? So if you go off and do a test where you get a lower efficiency than mechanical brakes(a "penalty" as you say) it can't be true period! What you are saying in effect is I have negative efficiency which of course is a nonsensical concept. There is by necessity a different explanation."

    I DO understand what you are saying. ANd no, I cannot agree. Since I've explained it in the video and the text, I'm a little at a loss. I assume you haven't watched or read them.

    But this may be the heart of the thing that has confused me about the industry missing this. You're kind of demonstrating HOW to miss it.

    Yes, we can agree that friction braking has a zero percent efficiency at recapturing energy. And I am assuring you in all seriousness that there is actually a PENALTY in efficiency if you employ regenerative braking. Now I understand both of those. But you are convinced that they cannot be true simultaneously.

    Here's the problem. And you HAVE helped me see how deeply embedded it is. There IS a way for them both to be true.

    Everyone looks at how many AMP hours are expended on the drive - that is the drive WITH the regenerative braking. And then they look at the energy captured by regenerative braking, that is energy restored to the pack. And indeed I can confirm both. We did expend energy. We did recapture energy.

    In the next step, they develop a ratio of the expense to the savings, and come up with an efficiency gain.

    Now READ this and REPEAT it until it sinks in. And I'll work on wording it better. The original premise of the Amp hours EXPENDED is FLAWED. There was ANOTHER drive not driven. And that is the drive WITH NO REGEN AT ALL CONFIGURED ON THE CAR FOR THE ENTIRE DRIVE. It was never conceived of. And it was never measured. So it was ignored entirely. It's off the radar screen because it never happened. An alternate universe. A path not taken.

    The assumption you are making is that with NO regen the expenditure of energy would be identical The simplest explanation for these results, and simplest is usually the best, is that if you simply measured EXPENDED energy and IGNORED regenerative energy for two separate drives, one using regen and one not using regen, they would be two very different numbers.

    I understand this is counterintuitive. And there is no simple and elegant explanation for its occurrence. But it is by far the most likely explanation for these results.

    So how could that be?

    Well, let's try another little thought experiment. I'm going to keep after these until I find one that works with you guys. And I assure you, if you don't get it, it's not you. I'm really struggling here to communicate this. I'm going to find a redwood stick analogy yet.

    ReplyDelete
  81. Let us say we have a drive of 5 miles down a perfectly straight, perfectly flat road. In the case of no regen, we are going to accelerate to 45 mph, and we are going to maintain 45 mph until the final hundred yards, at which point we simply brake to a stop.

    We actually will use quite a bit of energy to accelerate to 45 mph. But once there, the energy required to maintain 45 mph is miniscule. We have to overcome air resistence and rolling resistance to maintain that speed and air resistance is not too substantial at 45 mph.

    Let us say that to cover that 5 miles, we expended 5 AH.

    Now. We are going to go on a SECOND drive and this time we will have regenerative braking and indeed reap all the glorious rewards of having such a marvelous device. We're going to drive the same course. And we're going to do it in the same elapsed time.

    But we're going to first accelerate to 50mph. And as soon as we get to 50 mph, we are going to decelerate to 40 mph. ANd as soon as we get to 40 mph, we're again going to accelerate to 50 mph. And as soon as we get to 50 mph we are going to decelerate to 40 mph. Over teh course of the drive we are going to average very closely to the 45 mph speed of the first drive. And we're going to complete the drive in very nearly the same time.

    To accelerate to 50 mph, we do NOT simply have to replenish for rolling and wind resistance. We have to apply sufficient energy to accelerate the ENTIRE MASS of the car. This energy is stored kinetically in the forward motion of the car.

    But we're going to take it out. We're going to deploy our magic wind sail regenerative braking device and withdraw that kinetic energy to precisely the degree necessary to decelerate to 40 mph. But there are some efficiency problems. We have a VERY good system and it is fully 80% efficient at recovering that kinetic energy.

    Now we are at 40 mph. And we have to accelerate to 50 mph. We need 100% of the energy necessary to overcome rolling resistance and wind resistance, plus 100 percent of teh energy required to accelerate the mass of the car to 50 mph.

    Again we decel, and we reclaim 80%. Again we accel and expend the same energy the first drive was expending to overcome rolling resistance and air resistance, plus 100% of the energy required to accelerate the mass of the car to 50 mph.

    If we continue to use 100% and reclaim 80% and we do this 200 times in the course of the 5 miles, will we be able to complete the drive in the 5AH used by the first drive with no regen?

    ReplyDelete
  82. The answer is no. It won't even be close. The most efficient use of the kinetic energy stored in the forward motion of the mass is to continue the forward motion. No translation can improve upon this.

    And the difference between accelerating to a speed and maintaining a speed is dramatic. You can intuitively look at an amp meter while accelerating to any speed, and then watch the amp meter when you're maintaining that speed and readily observe this.

    So why the small difference in our drives? Pick a different number.

    Let's say in drive 1 we maintain a very steady 45 mph. And in drive two we accelerate to 47 mph. And we decelerate to 43 mph. And we accelerate to 47 mph. And we regen back down to 43 mph.

    NOW would they be comparable. Well, they would be closer. But the steady 45 would still win. It's really a very old story - the story of the Tortoise and the Hare.

    What you have failed to conceptualize is a drive not taken. That is where the missing energy is. The two drives were never the same energy expenditure. But because one wasn't driven, the comparison was never made. You instead focused on the regen part. The energy has to "go" somewhere. Actually, it never had to be anywhere at all.

    So yes. If no regen causes an expenditure of 78 AH. And with regen causes an expenditure of 90AH, but recaptures 11.8 AH, you are left with 78.2 AH, which is a 0.2 AH penalty for having regen, and precisely what we reported in the Speedster comparison of braking only regen against no regen whatsoever.

    Our thought experiment works BEST conceptualizing the regen of the accelerator, as if we never could quite find the sweet spot where none goes out and none goes in. The braking only is conceptually more difficult, but it is PRECISELY the same thing and the same cause and the same effect. In fact, apparently to pretty close the same degree. Acceleator regen PLUS brake regen were indeed less efficient than brake regen alone, which was in turn less efficient than NO REGEN AT ALL.

    And to put it further in scale, this thought experiment represents about a 7.5 minute drive across 5 miles. We had an hour and 25 minute drive across 48 miles that was HEROICALLY more complex than our thought experiment.

    The results did indeed blow me away. But another thing did as well. The random lights, traffic incursions and near accidents, etc. added up to quite a rigorous and variable drive. That we would complete it within just a few minutes of each other, and within tenths of AH of each other, just boggles my mind. The results were not just surprising, they were QUITE surprising in their consistency. In the analytical and statistical efforts to read messages from God in cloud formations, you don't usually get this tight a grouping. That was my first and main signal of validity. It was kind of an "aha" moment where you say "Hey, we've got something here... looks like a message from God...Times Roman font no less...."

    Jack Rickard

    ReplyDelete
  83. "Jack, I cant justify spending $500 on a programmer to turn off regen to do this test. I might be able to justify renting one again. Problem is I'm not sure why I need to. You say regen is a penalty. OK. So if I do a run trying for zero regen, even if it's not perfect, it's still going to be a lot less regen than my normal driving, so it should be fewer AH's used according to you, yes? If not, then why?"

    Ok, don't do it. I guess I'm a little puzzled how you built an electric car using a relatively newish 3 phase controller from Curtis, that has a LOT of variables in it, and don't have the programmer to modify it. It IS a little pricey at $500, but there are a lot of variables to play with and we just tune and tune and tune and tune. I've very nearly worn this one out. (Don't leave it in the sun for any time at all).

    I HAVE an 840 JRP. You could plant it in the middle of my face and I couldn't tell squat from it with regards to current. It samples slowly and the result is the little LCD characters dance all over the place seemingly at random. I would not be able to hold the car to ANYTHING watching this if all I had to do is watch it and mind the accelerator, ignoring entirely the road and the traffic. The idea of you trying to do this gives me a chill. I can just see the headline:

    " As a result of a forum discussion with Jack Rickard of EVTV on regenerative braking - local man whacks enormo garbage truck with electric car - Film at Eleven."

    To my mind, to experimentally confirm or refute, you have to be able to:

    1. Put together a persuasively similar drive in length. Does't have to be the same length, but same order of magnitude 30 miles ok. 3 miles not ok. Differing terrain, temperature, or falling rocks a PLUS actually.

    2. Count AH accurately and bidirectionally. wH introduces a whole series of things I DON'T want to get into.

    3. Turn off regen entirely once. Turn on regen once, but you can use whatever level of regen you think appropriate and comfortable. Shouldn't matter on ON. But OFF really means OFF, not ON but "wishing" it were off. Simulating it being off by not ever braking. Or whatever. It's best to just turn it off.

    4. Don't hurt yourself doing the experiment. This is just for fun. Not a single OEM will give up regen on my say so.

    How do you trigger regen now? I used the brake potentiometer input. If you did the same, you could put a switch on the wiper input and simply turn it off by disconnect. If you used the brake light signal somehow instead, you could use that. Cut a wire, put in a switch.

    I'd actually switch the 5v to the brake pot in my case. Leave teh wiper connected but just disable the power to it so it went to ground with the switch off.

    If you have it on the acceleator, you just have to have the programmer to turn it back off. I can't think of an electrical way to disable that on the Curtis. You have to have a throttle input.

    Jack Rickard

    ReplyDelete
  84. CyrilMaurice:

    I like this thought process. Not sure I agree with the conclusion. If I can get greater energy efficiency from driving style than I can from an additional $10,000 in hardware, teach me to drive.

    But along those lines, read my thought experiment on the 5 mile flat drive. We're basically on the same page. Acceleration of mass takes much more energy than maintaining steady state speed of mass already in motion. And if you regen down, you're at your 80%, or thereabouts somewhere. If you picture an oscillating cycle opposed to a more stately coast, it works out. In practice, we were doing the drive in the same amount of time, in fact 3 minutes quicker no regen on the MIni drive - using 7% less energy overall - apparently.

    Jack

    ReplyDelete
  85. Palmer:

    No I was responding to you and rereading it my response doesn't even make sense. I was coyly trying to communicate that you are already on the winning side, even if by accident. The advantages of AC drive appear to be not so pronounced as thought. And you having Series DC is ultimtely the smartest move of all. If you hold still a minute, the world will circle back to your way of thinking.

    Sometimes the simplest solution is the best one. Ouch. This one got me.

    Jack Rickard

    ReplyDelete
  86. "Holy Crap!! they did fill your mail box with hate mail!

    I think these guys need to stay on the DIY forums, typing themselves smart....."

    Actually no, I'm not like that. And I participated willingly on DIY until the moderator decided he needed to intervene. Everyone is welcome and they can say what they like more or less on topic. But I DO get to reply. And the "moderator" can't post messages purporting to be from me.... I do my own here...

    But yeah. I sat down with the results on the first drives and was initially quite elated of course - a new problem to noodle. But then the air just came out of me when I thought about the reaction...

    I'm doing what I like and having fun. There's a little housekeeping goes with anything you choose to do . Put a smile on your face and go do it. It's not like if you ignore it someone will sneak in and do it for you...

    Jack

    ReplyDelete
  87. David Nelson:

    A lot of very paternal admonitions there - none I can find any value in I'm afraid.

    We've been very successful with the brake pressure transducer on the Speedster. We got a little lucky early on in the mapping variables and learned to front load it mostly in the controller. We maintain positive current into the pack until very nearly stopped. We use very little friction brake. And it feels good.

    In any event, we use a completely different braking system on the Mini, using only the brake light signal to trigger regen. These are power brakes with no power. So you really have to put some leg in to get friction brakes. They are there underneath, but you can bring the car to a full stop, with current right down to jsut a few mph, on regen only. It is completely different from the Speedster, and we got the same results. SO your blind guess that we should try different modes doesn't wash.

    On the Speedster we had very strong regen on accelerator neutral, down to a complete stop rather quickly. On the Mini, it is very mild on accelerator neutral, we set it up that way. It barely simulates back pressure from an ICE engine.

    Your buddies are full of shit. Your theory is made up and contrived on information you don't have about our setup. It's just pure sniping with no purpose and wild guess work. I have no time for that. Try to contributte something positive to the discussion.

    AH and kW. Regen is current. AH are current. You can cut up all of electrical power a dozen different ways with ohms law, but in going to wH you are complicating not simplifying. There is an issue with voltage. Charge the car before you go, and do the same drive. That pretty much eliminates the effect and indeed with this close resulting numbers it is immaterial to the test.

    Jack Rickard

    ReplyDelete
  88. "What might be helpful is if you go for a long drive with your normal regen. Note the net AH removed. Now recharge the car and go on the SAME drive wiht the REGEN TURNED OFF. Note the net AH removed. What were the two numbers?

    It's really not any more compllicated than that."

    That drive will show zip about regen braking. First of all it will be most of all coasting, where no brakes are applied at all.

    The regen braking will ONLY help you if you HAVE to brake (ex. a red light or crossing etc.). It it TOTAL useless to anything else - it will WASTE energy in all other scenarios since there is a less than 100% efficiency in putting the energy into the batteries in the first place.
    But if HAVE to brake, then regenerative braking can supply SOME of the energy used to get you up to speed again (but not all).

    If your driving style are defensive and you understand the dynamics of the car, then there are little use for the regen braking, because you "keep the forward momentum" of the car. But if you drive car hard and use the brakes quite a lot, then you will not LOSE the amount of energy that you otherwise would with mechanical brakes. But the first style will always win from an energy conserving point of view.

    And then stop use the Ah as a measure for energy, it is wrong and only partly true due to a relatively flat discharge curve.
    Use teh Ah to tell you when the batteries are fully charged and note the energy used to fill the pack. And since the capacity of the batteries are decreasing (most of the time) even than has a built-in error.
    We are trying to measure energy savings below 5% and that is really difficult in a real world environment.

    I am not trying to type myself smart here and I am not selling regen braking to anybody - but I have 10.000 miles worth of driving in a electric car and have had little use for regen braking, but when I have to make a stop I will use it and compared to dumping heat into brake disc it is more efficient from a energy point of view. But not using the brakes at all will always be better. The same can be said about accelerating and deceleration compared to constant speed.

    And another observation not related to electric cars, but to petrol based cars. Two different drivers who drives the exact same trip in the exaact same car, can easilly come up with 20% different fuel economy - the same goes for electric cars (or steam cars for that matter).

    For what it is worth I have stored the drive with VRLA and regen braking here - http://www.citystromer.dk/Temp/ref%20drive%201%20-%20vrla.xls

    Please note that the drive was not done in order to prove anything else but lead sucks in electric cars ;-) At the end of this short drive it even went into current limiting mode...

    Martin.

    ReplyDelete
  89. A small PS:
    The CityStromer performes regen when the accelerator is off. And is only set for maks 0.25 x Nominal current (270Amp) and will decrease to 0 when the pack voltage is about 105V.

    Martin.

    ReplyDelete
  90. "So yes. If no regen causes an expenditure of 78 AH. And with regen causes an expenditure of 90AH, but recaptures 11.8 AH, you are left with 78.2 AH, which is a 0.2 AH penalty for having regen, and precisely what we reported in the Speedster comparison of braking only regen against no regen whatsoever."

    This more or less suggest that it is your driving (or lack of understanding of how to drive a specific electric car) that is the point here. The citation suggest that your are driving like a "bang-bang" controller. Full speed ahead and the coast. And I will totally agree that with that kind of (mis-)use of the accelerator there is not point what so ever for regen - you will loose the small gain when putting large amount of energy back and forth to the batteries.

    I am sorry for not realising this earlier, but in this blog entry the replies are comming more frequent for me to read it all - and during the nigth (CEST+1) there was the addition of 20+ entries ;-)

    Martin.

    ReplyDelete
  91. Jack, I built the car without the programmer, drove it, then rented a programmer for a couple of weeks to set it up as I liked it, very simple.

    Tom has done some testing, with regen turned off, and with regen on the A pedal, and has completely opposite results than you. His description follows:
    "I drove an about 5 mile loop with about 6 stop signs and a few small hills. Car charged to the same level before each trip. With regen I used about 12% less Ah. Results will depend on terrain, speed, stop/starts...I tried to accelerate as much the same as possible in both runs. With regen turned off I drove as if hypermiling, coasting as much as possible while not dropping too far below the speed limit - except to stop at the stop signs of course."

    Now, before you say the trip wasn't long enough, what does extra miles of steady state or acceleration give us in regards to the effects of regen? Nothing. If you aren't slowing down with brakes or regen you aren't comparing anything. Tom and I have a lot of practice with regen on the A pedal, you have a lot of practice with no regen at all. Your regen on the brake pedal means you're never getting max regen since the brakes are doing much of the work. It may feel like power brakes and that's fine but you are cutting out a lot of potential regen. With regen on the A pedal you may have your controller parameters set in such a way as to make it more difficult to use it smoothly, your throttle may be too touchy to control it easily to find the neutral spot for coasting. As for the 840, it's not hard to read at all, especially when it's showing a steady "0". Again, my normal driving gives me seconds of 100-200 amps and minutes of 50 amps, a few seconds of 1-10 amps isn't going to skew results. If you don't like my data look at Tom's.

    JRP3

    ReplyDelete
  92. Jack, your thought experiments aren't the right ones. Try this. You're doing 45mph for whatever distance you wish, then you have to stop within 200ft. Which battery pack has more energy in it when you stop, the one with regen or the one without? It's that simple.

    JRP3

    ReplyDelete
  93. Ok JRP. You win. You've busted my myth busted.

    Now I never could get you to duplicate the methodology. And you've got it all worked out how you don't need to to disprove it. So if you're happy I'm happy.

    I love your data and I love regen.

    I do have to point out that we made no attempt at coasting or hypermiling. Virtually ALL of our braking was with regen and indeed, in the SPeedster with 50% set on the accelerator neutral braking, we didn't really need the brakes at all. And on the MIni Coooper we very nearly don't HAVE any brakes without regen. And that as I've said numerous times, the regen current is clearly visible up to as much as 100 amps of regen braking current.

    The 840 is quite easy to read as to Mtemp and Ctemp. On my machine, the 840 current display jumps around wildly even at a steady speed to such a point I cannnot believe you even performed the test as you described.

    But if that's the data you got that's the data you got. We have two experimenters here that have data refuting my claim entirely - and using teh same motor and controller.

    And you are quite right, 50 feet and 50 miles are pretty much the same. All the same factors should apply at both ranges.

    So, help me understand then why I get almost identical results between braking on the accelerator lap and braking using regen braking on the brake pedal only? We rarely had to touch the brakes at all when I had it on the accelerator, and we braked hundreds of times when it was on the brake pedal only. But I used MORE AH per mile on the drive with the accelerator regen than I did on the drive with it on the brake only. SHouldn't the reverse have been true? And shouldn't it have been noticeably more?

    Jack Rickard

    ReplyDelete
  94. "Jack, your thought experiments aren't the right ones. Try this. You're doing 45mph for whatever distance you wish, then you have to stop within 200ft. Which battery pack has more energy in it when you stop, the one with regen or the one without? It's that simple."

    Oh now this is good. My thought experiments aren't the right ones? How should I reply to this JRP? You're just trying to get me to react badly to foolishness, because you know I do react badly to foolishness.

    In your thought experiment, I think regen would provide better efficiency.


    Now let's see here. A description of alternate accelerations and decelerations over 5 miles is what I propose as a simplification of a 50 mile drive through urban traffic with dozens of stop lights and stop signs etc. Both being representative of a real world drive.

    You feel that a MORE appropriate simplification would be to drive an infinite distance at 45 mph and then brake to a stop at the end in 200 feet? I take it THAT would be a thought experiment that would be the RIGHT ONE. That is a thought experiment that you would deem essentially "correct."

    Hmmmm. I'm going to have to work on that.... You give me much to consider here.... I think.....

    Jack RIckard

    ReplyDelete
  95. "This more or less suggest that it is your driving (or lack of understanding of how to drive a specific electric car) that is the point here. The citation suggest that your are driving like a "bang-bang" controller. Full speed ahead and the coast. And I will totally agree that with that kind of (mis-)use of the accelerator there is not point what so ever for regen - you will loose the small gain when putting large amount of energy back and forth to the batteries."

    Martin - then you are suggesting that I have derived some very erroneous data basically because I don't know how to drive a car properly or correctly. Well, that may be true. It's certainly a possibility.

    It might be worth pursuing the earlier suggestion of doing the test with another driver. Perhaps one that is better at operating these here electric vehicles.

    Jack Rickard

    ReplyDelete
  96. Jack


    I went back over the original post and wanted to know how you determined the amp hours used in the original tests?

    You said:

    "And we ignored how much was coming out and going back in, and looked for the TOTAL AH consumed on the 76.6 mile drive"

    Did you look for this using the zantrax?
    Did you look for this by recharging the batteries and reading KWHR in?

    I don't like using the word assume. It seems to me you are measuring the amount needed to "fill'er up" since you ignore AMP HRS in and AMP HRS out.

    Also,
    I read through the comments and I hear some people asking for you to take the readings in KWHs not AHRs.

    I also am wondering why.
    Although if you fill up at the wall I don't think it matters.

    Regards, JMS

    ReplyDelete
  97. My point is for a test to be valid the driving must be the same up until it's time to slow down. You seem to be saying that the driving cannot be the same with and without regen, and that you cannot drive a regen vehicle in such a way as to meaningfully recover any braking energy. According to you, simply driving a vehicle that has regen wastes energy even before you start to slow down with regen. This is because even when trying to coast you will probably be either using some current or some regen, which is less efficient than coasting, and that any braking energy that can be had will not overcome that inefficiency?

    JRP3

    ReplyDelete
  98. Jack,

    This is getting fun, isn't it? No comments on my last post?

    Let me pose this question to you. If I wrote in that I mounted a wind mill on top of my car and that I am now getting 10% better gas mileage and I have a lot of data to prove it, would you feel comfortable refuting this based on science or would you feel the need to duplicate my setup and gather data to refute it? I hope you would feel it to be legitimate to point out the fallacy of my invention and suggest that whatever test I used to gather my "convincing data" must have a flaw. Not because you ran the same experiment but that the fundamental laws as we know them (1st law) don't allow for my results.

    It looks to me that your thought experiment for the aggressive accel/decel drive has a problem on the comparison side. It sounds like you are comparing this aggressive drive using regen to a "normal" drive not using it. In other words, do you really believe that if you drove the aggressive drive with standard brakes you would be better off than with regen? You show how the inefficiencies of the regen don't recapture ALL of the energy of slowing down which is spot on. But you must realize that mechanical brakes don't capture ANY of the energy of slowing down. Which is better, "doesn't capture all", or "doesn't capture any"? This is precisely the type of drive that shows regen to be useful!

    I think you need to look carefully at your data collection and analysis to see if there could possibly be a fallacy. This is not a personal attack; its just an objective challenge. You have run the experiment with multiple passes with multiple vehicles, but with the same drive and the same instruments. My suspicion is that the instruments are not lying to you, they are misleading you.

    I continue to think the problem lies here;

    "wH introduces a whole series of things I DON'T want to get into"

    I'm not sure what you mean by that, but as I pointed out in my last post, energy is what we are talking about, and that is WH, NOT AH. Think about this. If your electric motor is drawing 50 amps at 100 volts, you are using 5000 watts of power. You do this for 1 hour and you have used 5000 watt/hours of energy. If you run the same 50 amps but this time at 95 volts, you use 4750 watt/hours of energy. This 5% difference resulted from a change in voltage that I would think can and does happen in the real world. In the case of regen, the voltage MUST go up to drive current into the battery (against the battery voltage) yet you only measure amps and assume the voltage is constant. The result in an understatement of the energy recaptured to the battery. I think you need to employ a WH meter connected between your motor and controller. I don't know if a conventional home WH meter would respond to the frequency and waveform of the AC motor you are using, but it is that type of meter that will give you true energy usage. In fact, I suppose the power companies use WH meters rather than current meters for this same reason.

    ReplyDelete
  99. Jack,

    This is getting fun, isn't it? No comments on my last post?

    Let me pose this question to you. If I wrote in that I mounted a wind mill on top of my car and that I am now getting 10% better gas mileage and I have a lot of data to prove it, would you feel comfortable refuting this based on science or would you feel the need to duplicate my setup and gather data to refute it? I hope you would feel it to be legitimate to point out the fallacy of my invention and suggest that whatever test I used to gather my "convincing data" must have a flaw. Not because you ran the same experiment but that the fundamental laws as we know them (1st law) don't allow for my results.

    It looks to me that your thought experiment for the aggressive accel/decel drive has a problem on the comparison side. It sounds like you are comparing this aggressive drive using regen to a "normal" drive not using it. In other words, do you really believe that if you drove the aggressive drive with standard brakes you would be better off than with regen? You show how the inefficiencies of the regen don't recapture ALL of the energy of slowing down which is spot on. But you must realize that mechanical brakes don't capture ANY of the energy of slowing down. Which is better, "doesn't capture all", or "doesn't capture any"? This is precisely the type of drive that shows regen to be useful!

    I think you need to look carefully at your data collection and analysis to see if there could possibly be a fallacy. This is not a personal attack; its just an objective challenge. You have run the experiment with multiple passes with multiple vehicles, but with the same drive and the same instruments. My suspicion is that the instruments are not lying to you, they are misleading you.

    I continue to think the problem lies here;

    "wH introduces a whole series of things I DON'T want to get into"

    I'm not sure what you mean by that, but as I pointed out in my last post, energy is what we are talking about, and that is WH, NOT AH. Think about this. If your electric motor is drawing 50 amps at 100 volts, you are using 5000 watts of power. You do this for 1 hour and you have used 5000 watt/hours of energy. If you run the same 50 amps but this time at 95 volts, you use 4750 watt/hours of energy. This 5% difference resulted from a change in voltage that I would think can and does happen in the real world. In the case of regen, the voltage MUST go up to drive current into the battery (against the battery voltage) yet you only measure amps and assume the voltage is constant. The result in an understatement of the energy recaptured to the battery. I think you need to employ a WH meter connected between your motor and controller. I don't know if a conventional home WH meter would respond to the frequency and waveform of the AC motor you are using, but it is that type of meter that will give you true energy usage. In fact, I suppose the power companies use WH meters rather than current meters for this same reason.

    ReplyDelete
  100. If we are to solve this debate, we must focus on the energy and where it goes. Draw a free body diagram around the car and account for all of the energy going in and out. It's easy to get lost in the weeds if we try to talk about complicated driving scenarios. Let's keep a focus on where does the energy go when we slow the vehicle down using brakes. We agree that if we use pads, all of it goes up in heat dissipated by the rotors. Where does it go if we use regen?

    Here is an interesting aside regarding power and energy. It is counterintuitive but nonetheless true that it takes a lot more POWER to accelerate a mass quickly (than slowly), but no more ENERGY (assuming no losses in the power delivery system). I know in the real world there would be some losses, but when we do calculations in physics using "frictionless planes" or "ignoring wind resistance" we don't imply the conditions exist. Rather, they serve to illustrate the principles.

    I look forward to your thoughts.

    Regards,

    Dan

    ReplyDelete
  101. Sorry for the duplicate post above... I got an error message and thought it didn't post the first time. Please delete one of them.

    ReplyDelete
  102. "Let me pose this question to you. If I wrote in that I mounted a wind mill on top of my car and that I am now getting 10% better gas mileage and I have a lot of data to prove it, would you feel comfortable refuting this based on science or would you feel the need to duplicate my setup and gather data to refute it? I hope you would feel it to be legitimate to point out the fallacy of my invention and suggest that whatever test I used to gather my "convincing data" must have a flaw. Not because you ran the same experiment but that the fundamental laws as we know them (1st law) don't allow for my results."

    Dan:

    First, what data? And second, how did you collect it. Assuming that it defies the fundamental laws of physics is the usual response to new reports.

    But this is comical. YOU are the one with the perpetual motion machine in presuming that you can put energy back with the regenerative braking. I'm reporting it didn't work. So you have your analogy badly out of kilter by about 180 degrees. Regen is your wind sail. And I've compared it and found it doesn't do what you report.

    Your stuck on "Where does it go?". It never was there in the first place. When you decel, you have to accel, and that takes more energy round trip than just maintaining a steady speed. The regen influences you to do this as you drive, and it is cummulative across a significant drive.

    If you want to go up a hill and down a hill you will show a gain. But if you do it across a longer drive, you wind up with lots of unnecessary accels and decels. They don't have to be extreme. As I said, you can reduce it to 47 and 43 around 45 and wind up with the same disparity, but to a lesser degree more commensurate with what we are measuring.

    I don't entirely agree that it takes the same energy to accelerate quickly as to accelerate slowly, but for the purposes of this discussion I'll allow it. It is essentially true. And it is probably at the heart of this.

    Maintaining a steady speed simply does not require the energy that it does to accelerate. And you don't get all of it back when you decelerate. I think the regen induces oscillations in the driving speed. Actually I know it does. I think that's where the disparity comes from.

    Again, and again, you are stuck on looking for the lost energy of the braking segment. But you are still looking at the drive you made, not the drive you didn't make (with no regen).

    Jack Rickard

    ReplyDelete
  103. "Martin - then you are suggesting that I have derived some very erroneous data basically because I don't know how to drive a car properly or correctly. Well, that may be true. It's certainly a possibility."

    I am unable to detect if you are being ironic, sarcastic or honest here... Please excuse if the finer aspects of the American language passes me by.

    But to my experience if there is something that seems to be incorrect, the data (or formulation of the test) is wrong in most cases. In some rare instances people might stumble upon new knowledge, but that is rare.

    "It might be worth pursuing the earlier suggestion of doing the test with another driver. Perhaps one that is better at operating these here electric vehicles."

    I am only trying to say that the driver normally have a very great influence on the efficiency of the drive - and to make a setup that will prove or reject your thesis, may be very hard in a real life environment.

    Martin.

    ReplyDelete
  104. "According to you, simply driving a vehicle that has regen wastes energy even before you start to slow down with regen. This is because even when trying to coast you will probably be either using some current or some regen, which is less efficient than coasting, and that any braking energy that can be had will not overcome that inefficiency?"

    I would suggest you mentally modify the concept of "coasting" and replace it with the concept of "maintaining a steady speed".

    As you go down the road at a steady 45 mph, you are NOT coasting if we stay focused on our flat ground 5 mile drive. You are driving and you are using energy. You must do this to overcome the effects of rolling and wind resistance. But you do not have to overcome the inertia of the entire vehicle mass which you DO have to do to accelerate. That same inertia is what you are "recapturing" with the regen.

    My "theory" to explain these results is that regen tends to induce these accel/decel oscillations around the steady state speed. I can readily OBSERVE this effect with the throttle based regen and in fact it is quite pronounced.

    Somewhat less obvious is that regen on the braking side does the same thing, in much the way power brakes would over manual brakes. It would be a lesser effect, but still there.

    This would account for the almost linear lineup of our results. Least efficient was with regen on both brake and throttle. Next most efficient was brake regen only. Most efficient was no regen. That lineup poses some problems in theory. This explanation would account for that.

    That doesn't mean it is correct. It means it would account for that.

    Jack Rickard

    ReplyDelete
  105. "My "theory" to explain these results is that regen tends to induce these accel/decel oscillations around the steady state speed. I can readily OBSERVE this effect with the throttle based regen and in fact it is quite pronounced."

    Hmmmm, to hold a steady speed you are still "accelerating" against friction and wind, so the throttle must be constantly in a positive position and you must constantly be drawing amperage from the pack. I've never seen regen appear while holding steady speed on the flat. At 45 mph on the flat I use about 50-60 amps, I've never seen that go negative in those situations, nor have I noticed oscillations.

    Are you saying that since even steady state driving has minuscule fluctuations in load that the motor is having microsecond events between power and regen as you travel, too small to show up in the gauge, but over time would show up in more energy used to travel?
    JRP3

    ReplyDelete
  106. "And I've compared it and found it doesn't do what you report."

    I don't doubt that your data as you gather it exists and is accurate for what it is. By replicating the test with multiple runs and vehicles with good repeatability you can have confidence in your numbers being accurate for what they are. I question if you are measuring the right parameters and interpreting correctly.

    "Your stuck on "Where does it go?". It never was there in the first place."

    Of course it was there, "it" being energy. There can't be any debate on this! When you are driving down the road the energy is there in the kinetic form. It can't be destroyed so it has to go somewhere.

    "I think the regen induces oscillations in the driving speed."

    If you have regen on the brake pedal and you are not touching the brake pedal, how can it possibly induce an oscillation? Steady state driving has no dog in the fight, nor does acceleration. Only deceleration at a rate that exceeds the decel rate provided by the other drag losses (wind, rolling)is even in the equation. I will agree to these points:
    1) There is no regen available during accel
    2) There is no regen available during steady state
    3) There is no regen available during coasting decel
    But not this:
    4) There is no regen available during braking at a rate that exceeds coasting decel
    I say "available" because being there doesn't mean it is captured. It might have been available but lost due to the design of the capturing circuit.

    Why do you completely ignore what I am saying is possibly the real culprit; namely AH vs WH? Do you have a technical argument that says WH is not relevant, or do you believe it is so trivial an argument that it isn't worth addressing directly? I gave you a cogent explanation of what might be happening with your AH measurement during regen and you ignore it completely. If it turns out that this is THE issue and you completely ignore it you will never find the truth. If its not the issue then give me an argument that proves it or at least provides some support for your assumption that it isn't important. I am genuinely interested!

    Dan

    ReplyDelete
  107. "Hmmmm, to hold a steady speed you are still "accelerating" against friction and wind, so the throttle must be constantly in a positive position"
    Nope. Your are applying work in order to MAINTAIN a constant speed.

    Let us please keep the term as correct as possible - otherwise we will just confuse each other to an even higher level ;-)

    Martin.

    Martin.

    ReplyDelete
  108. The first law of scientific endeavor is:


    1) Do not deceive yourself.

    I hearby add the second LAw

    2) focaults' razor says: see above

    I have been an electrician for 25+ years and I always used WATTs not AMPs to determine what work is being done, then of course divide by Volts to find AMPs.

    Jack please explain why we should not use KWHRs.

    Regards, JMS

    ReplyDelete
  109. Theory: "Where does the energy go?". Heat. Either the brake pads and rotor (no regen), or the motor/controller/batteries(regen). Some of the energy gets into the batteries, but most of regen transfers to heat in the electric components, thereby reducing efficiency and causing more Ah to be consumed overall the more it is used. Since heat in the brake pads and rotor doesn't affect drivetrain efficiency, you need less amp hours overall. Regen causes all the drivetrain components to be less efficinet due to heat.

    ReplyDelete
  110. Let's be clear. The heat is not the cause of the inefficiency, it IS the inefficiency. And while I agree that not all of the regen energy makes it to the battery, SOME does. With regular brakes, NONE does. That's the point. Don't make it more complicated than that.

    ReplyDelete
  111. "Regen causes all the drivetrain components to be less efficinet due to heat."

    With a professional life as a R&D engineer at company doing lots of 4Q 3-phase inverters I tend to disagree with you on that statement. Our inverters are used for UPS's and not automotive, but that is, from a topological point of view, inconsequntial. The AC to DC effinciency is above 98% for a modern 4Q inverter. So the main losses are in the batteries and motor.

    Martin.

    ReplyDelete
  112. You guy's are missing something, and it is the same thing that has Jack and I scratching our heads. I may be way off in this thought experiment, but lets give it a go anyway.

    The car without regen took 78AH to go around the circuit. For conversation purposes say the car averaged 120v on the pack.

    Now you guys all assume that the car got a 15% gain in wH due to regen on the second run.

    The original run was 78*120=9360wH. If the second run had a 15% increase then you are saying that the car actually took only 7956wH. If this is true then the car would have to had an average pack voltage over the course of the trip of 102v (78AH*102=7956). There is just no way that was true. This is the crux of the problem.

    I do like the thought above regarding heat into the motor and controller. Regen is recovering 15% of the power, but maybe 7.5% is going into heating the motor and controller, and running a hot motor and controller gives 7.5% less efficiency so you end up with a zero difference.

    Is there a way to test for this? Or am I way off base.

    ReplyDelete
  113. PMD,

    I'm not sure where the 15% comes from and for my argument I don't claim any particular gain; just that there must be some gain with regen if you drive a route with significant braking. If you drive an identical route with an identical car you will use an identical amount of energy. The question is, when you get to the end of the route do you have more energy in your battery on the run with regen than the one with friction brakes. Jack proposes he has the same or less energy in the battery using regen, and I am saying I think he has more energy in the battery than he realizes. The reason he has more energy than he thinks is that when he is regening the voltage rises and pumps more energy into the battery than he accounts for since he is monitoring amps only. Amps is only half of the power calculation. If he could set up an experiment to measure Watt Hours instead of Amp Hours I think his results would be different. My argument is only that there should be SOME energy recovered during regen, not how much or if it is worth the money or the hassle or anything else. Most of all I argue that there is no way to get less efficiency than friction brakes.

    Respectfully,

    Dan

    ReplyDelete
  114. "Why do you completely ignore what I am saying is possibly the real culprit; namely AH vs WH? Do you have a technical argument that says WH is not relevant, or do you believe it is so trivial an argument that it isn't worth addressing directly? I gave you a cogent explanation of what might be happening with your AH measurement during regen and you ignore it completely. If it turns out that this is THE issue and you completely ignore it you will never find the truth. If its not the issue then give me an argument that proves it or at least provides some support for your assumption that it isn't important. I am genuinely interested!"

    Dan:
    I believe it is so trivial an argument that it is not worth addressing directly. I tried not to address JRP holding zero regen with his trigger finger and failed, but I do insist that I will not address this. It is entirely too trivial and leads to goofy discussions of things that cannot possibly matter to this topic.

    The results might indeed be different. On the order of milliamps.

    I've posted a continuation of all this on a new topic. This page scrolls forever. Please do view the latest video and I will be happy to resume the discussion on that page.

    Jack Rickard

    ReplyDelete
  115. Most of all I argue that there is no way to get less efficiency than friction brakes.

    And I've demonstrated that you can five times and over nearly 250 miles of urban driving.

    You wll NOT let go of the concept that the only drive that matters is the one with regen, despite the fact that you never did drive it. It was only theoretical.

    And if you compare it to ANOTHER drive that is NOT theoretical you will find that it nets out to near zero, and if NOT zero , on the wrong side of it.

    You have simplified this to "Does the act of braking using regen produce current." Of course it does. So does sunshine. But I can't use it in my car. The difference is I probably COULD use sunshine....

    Jack RIckard

    ReplyDelete
  116. This conversation reminds me of a famous description by Albert Einstein that I will paraphrase probably so badly that it won't work.

    "You see my dear, the telephone operates like a cat that stretches from New York, to Los Angeles. You pull the tail in New York, and the cat lets out a yowl in Los Angeles.

    Now wireless is exactly the same thing - but without the cat."

    ReplyDelete
  117. Found it.

    "You see, wire telegraph is a kind of a very, very long cat. You pull his tail in New York and his head is meowing in Los Angeles. Do you understand this? And radio operates exactly the same way: you send signals here, they receive them there. The only difference is that there is no cat."
    A. Einstein

    ReplyDelete
  118. There is certainly some losses through the motor and controller and indeed some heat. But it is like the wH, it would not account for any of this.

    To my way of thinking, there IS a cost to regen and that is mechanical strain on drive train components and the inadvisability of braking on two wheels vice 4. Well worth it for even 6-8% gain. We have no gain. That's what started all this. That's why the testing.

    The testing puts more definitive numbers to it. It no longer "seems" like we're not getting anything looking at a couple of drive logs and some calculations. We're now not getting it to a tenth of an amp hour.

    So you're point as to heat is technically correct, but I don't think signficant in magnitude. It doesn't explain this situation.

    Jack Rickard

    ReplyDelete
  119. Hello Jack and greetings from England. :)
    I've attempted to run through these posts looking for comments on charge/discharge charactoristics of Thundersky type batteries.

    I might be wrong but read somewhere LiFePo4's don't accept a charge immediately after discharge. It is why PML Flightlinks Mini project was fitted out with 4 Farad's of supercapacitors to soley take the regen charge. Big enough to accept 70-0mph against its mass.

    After all, batteries are a chemical process, not a capacitive one. Have you done any experiments on this?

    I'm definitely a backer for all wheel regen braking into a dummy load even if it only means reduced service/replacement costs of brake pads and discs!

    All the best Jack, (Brian& co.). Your show is a revelation to me.

    ReplyDelete
  120. I'm sure you read it, but no, there is absolutely no problem with THundersky's accepting charge before, during, or after charge. They accept regen quite readily and there is no need of supercapacitors. Supercapacitors can accept all the charge very quickly and without the losses of a chemical conversion, and they can also make that charge available for acceleration very quickly.

    Reduced brake plads? Common. This is a $5000 drive train and a $3000 transmission and we want to save wear and tear on a $18 pair of brake pads???

    Jack Rickard

    ReplyDelete