Posts
Well, we have posted our first show of 2011 and it is pretty good, if I mention it myself. An hour and 48 minutes.
We had recently posted a couple of photos of a horrific electric vehicle fire that occurred on an Oslo Ferry, burning up the car, the car next to it, and extensively damaging the ship they were on.
It was an almost suspiciously serendipitous occurrence then that a couple on an around the world tour in an electric car, just happened to be passing through Cape Girardeau Missouri the very next week, and in an identical car from the same company. As such, we got to both examine the car in detail, and put our shop at risk by allowing them to charge there overnight and unattended. Living dangerously.
Predictably enough, the car features a REAP Systems BMS controlling two Brusa NLG chargers. I found it odd that with sufficient chargers to be running at 20 amps, the car only charged at 9 amps. The laptop Hjalte hooked up to the BMS actually had a MARVELOUS and very informative display - if it were not for the part about burning down cars, buildings and battery packs, I think I would like this REAP Systems thing.
I found it EXTREMELY interesting that they noted that two of the cars had burned up and that the company, who could program their car remotely, had set it to this 9 amp value because of the fear of the fires. Second clue, they also noted that the company was very unhappy with this BMS and would be changing to a new system from Lithium Balance.
So it appears Soren Ecklund's public pronouncement of an adapter cord being the cause of the fire on the Oslo Ferry was a bare faced lie to cover his embarrasment, and he KNEW FULL WELL the real cause of the fire was the REAP Systems BMS. He already has plans to replace them, and to TRY to preserve life and limb on this world tour vehicle, he's remotely programmed it to cut the charge current in half - unnecessarily DOUBLING the charge times for this poor couple's journey.
The irony of course is that the Brusas, well able to charge this system at over 20 amperes, would pose no fire hazard at all if they could just put them in automatic mode, which Herr Ecklund could easily do by satellite if he had a clue.
Splendidly poor behaviour by a conversion shop on all fronts and I might as well publicly decry it as a sterling example of cowardice in the face of very nearly nothing.
But the visit with this lovely couple was very uplifting and we had a great time. They joined us at the house Sunday evening for dinner of Beef Stew which my wife actually does extremely well.
We also received 50 of the CALB LiFePo4 180AH cells this week and 30 of the 100AH cells from their new CALIB facility in Pomona California. Notably, we ordered those on December 23, received them on January 6, didn't have to do squat about shipping/import duties although of course we ultimately paid for it, and paid on American Express. Keegan Han has been a delight to work with and we can recommend this battery source quite enthusiastically. You can pay easily with recoverable funds in the event of mishap/nondelivery, we got them in two weeks, and they are direct from manufacturer.
It also provided us an opportunity to compare a brand new cell fresh from China to our existing stock of 180AH cells which we received in October 2009 - some 15 months ago. Voltage of the new cells: 3.300. ALL of them are at exactly 3.300. Voltage on the cells that have been stored for 15 months: 3.298. ALL were 3.298 or 3.297. The myth of self discharge and all you need to do to care for these cells in storage: BUSTED. If you see someone online going on about charging and equalizing and so forth to preserve these cells: note as moron and disregard all future postings.
Brain and Matt have been struggling to make my dreams of Porsche Speedster Belly/Battery/Box come true. They won. It projects a scant 1.75 inches below the existing frame of the car, is quite secure, holds 11 cells, and is essentially invisible. I had plans for the cables to and from: unnecessary. Our wiring tunnel bisects the box - an obvious fact I had mislaid mentally of course.
So we are at 171 volts and 51 cells on the Speedster rebuild. I haven't mentioned it to Matt and Brain yet, but I have an idea about mounting the Soliton1 on top of the motor, freeing the rear shelf for another six cells, which would bring us to 192volts and a pack size of 34,560 wH. We would undoubtedly have a shot at a 150 mile drive on a single charge.
AND about 175kw of power to the motor, including sag, for 234 HP to the wheels. This is OVER THREE TIMES the power we recorded on the dynomometer in this car previously. It provided exhilarating acceleration and top speed then, so I'm not sure what the point is. Perhaps will take on Tesla at Monte Carlo in April????
We also mounted a J1772 connector in the car, our two rear battery boxes holding 11 cells each, and are just generally making famous progress on this rebuild. It will be grossly overpowered and a bit overweight to my tastes, but with a very different level of range for bragging rights.
Really more of a can we instead of a should we, I do fear.
We've also started a new Texaco Fire Chief charging station for the 601 Morgan Oak facility, with J1772-2010 plug in this case. We're going to have a couple of Kilovac contactors for the phases, a 12v power supply to control them, a NEMA 14-50 in addition, a power meter, etc. It should be good for giggles - a better thought when drinking heavily of course.
But that's what we like to do - get liquored up, play with high voltage, and then drive our cars. Not PC, but whatever.
Jack Rickard
Subscribe to:
Post Comments (Atom)
Hi Jack,
ReplyDeleteIt would be interesting to see how much charge the old 180Ah cells have compared to the new 180Ah cells. You have stated that the voltage says almost nothing about the SOC/SOH of the cells. I have two TS-90AHA cells which I can use for long term testing. A way to really see if there is a self discharge is to discharge the cells downto the lower knee point and let them sit there, since the voltage dependency is more visible there than on the 50% SOC point.
I have just had a failed charger which I will have to replace before I can use my car again. The charger has been unstable for a long time, but have finally died of old age (it is from 1996).
I have ordered two 2 kW chargers (TCCH-H130-18). I ordered two because of redundancy and the possibility to use only one of them or both depending on the grid connecting available.
Martin.
Martin:
ReplyDeleteI don't recall saying that voltage will tell you nothing about SOC. Actually it's a great measure of it. Unfortunately, it is not very useful during the part of the discharge curve we are interested in, because it is very flat. You have very small changes of voltage to indicate relatvely large changes in SOC.
Then too, it is difficult to really achieve voltage measurement accuracy in a device you would install in a car. It is much more accurate and telling to count AH, however imperfectly.
But voltage DOES indicate SOC very well. In this case, we are measuring with a good meter, but understand that even a Fluke does not HAVE an accuracy to within thousands of a volt. What it does have is RESOLUTION to within thousands of a volt. And that is indeed signficant with SOC measurements - QUITE indicative.
So let's throw out accuracy. Because in this case we don't care. We only care about resolution. I am COMPARING the SOC and the voltage of two known cells. One is at precisely 50% SOC and is brand new. I can take a meter of sufficient resolution, and measure it with that meter. I don't care about accuracy. It could read 3.300 volts or perhaps be a little off and read 12.300 volts. Doesn't matter.
If I go to the second cell WITH THE SAME METER, as long as the meter is CONSISTENT in its measurements, if the voltages are the same, the SOC is the same. In this case 12.298 instead of 12.300 or just as well 3.298 instead of 3.300. We don't care about the meter at all, our CAL standard is THE NEW CELL. And we are comparing it to a 15 month old cell of the same manufacturer and size to detect SHELF LIFE.
So there is no need to go to the curve. In the first place, that isn't the question we are addressing. We already know the two cells will have two different capacities, though my sense is less so each time - the CHineses are getting better on the consistency question. We don't care about capacity. In this experiment, our ONLY question is about shelf life and self discharge.
And the result is VERY clear. There is not sufficient self discharge to bother trying to measure, much less fix. And yes, voltage in this case is an EXCELLENT indicator..
Jack RIckard
We've been discussing the self discharge topic at DIY recently so this video is very timely. I too thought it would be useful to discharge the cells and see what their actual SOC was as a verification of what the voltage was showing. I agree that though the curve is fairly flat it is not truly flat so voltage to 3 decimal places should have meaning.
ReplyDeleteJRP3
The REAP system, the name makes it sound like some sort of battery "harvester" ...last stop before that big battery box in the sky!
ReplyDeletegreat show!
JP:
ReplyDeleteIn NimHd and NiCAd batteries, as well as Pb cells, there is actually a chemical reaction termed a "shuttle reaction" that causes what is termed "self-discharge." There is no such reaction in Lithium chemistries. It's not that it is reduced, it does not exist.
This is terribly confused by the Chinese insistence to LIST a self discharge rate. They are trying to tout it as a feature in that it is a much higher value on other cell chemistries. But technically, it doesn't exist at all and is in fact chemically impossible.
There IS a diminishment in capacity in the first two weeks of life, which is why they "cure" LiFePo4 cells at the factory. It has nothing to do with self-discharge per se. It is more an adsorbtion of active LiPF6 and cathode/anode materials until a stable balance is achieved.
This is what has all the lead heads wrapped around an axle over cell "drift". Pb cells do indeed drift, as do NimHd cells - directly as a result of self discharge chemistry.
For these reasons, it was ALWAYS very questionable that we have any evidence of self discharge, and any evidence of cell drift. It's not really part of our cell's chemistry at all.
Ironically, you have people ACTIVELY designing BMS systems that have no clue about any of this. They don't know what causes self-discharge or cell drift in the first place, so they ill understand why we don't have it. And lacking that understanding, they have an entire mythology based on the fact that they DO have it, which they never did.
It makes my head spin. I hope you haven't been reading Dmitri and David Andrea's endless posts again on Endless Fear and DIYElectricjunk.
It's very difficult to prove a negative, and VERY hard to measure something that doesn't exist. Ergo, the opposite MUST be true and you really NEED a BMS to protect you from the things that you don't have anyway, which you can't prove you don't have so you must have them?????
In any event, I thought the voltage demonstration sufficiently pointed and dramatic to make the point without an organic chemistry degree, which I don't have either. Here's one we KNOW to be 15 months old, and here's one received yesterday. Here is their voltage. End of self discharge.
There IS no self discharge. There IS no cell drift. If you see someone referring to such, note their moronness, and ignore any further posts from that source. Period. They simply do not know what they are talking about, and how MUCH they talk about it won't change that.
Jack Rickard
JMS:
ReplyDeleteYes, more like the GRIM REAPER than REAP Systems.
Hey, guys. Nobody designs BMS to burn down cars. And this guy didn't design a car to burn up a ship. But engineering isn't about wishful thinking. You design, you build, you test, you rebuild, you retest, you rebuild, you retest, you rebuild, and sometimes you have to abort that design and start over.
That's long before you are selling stuff to an unknowledgeable public with the claim that it is providing SAFETY for their system. But when a failure does occur, they should step up to the plate and admit it.
By the way, I today learned the official Police Investigation indicates the fire did NOT start with a cord, but appears to have started in the batteries.
If you overcharge your cells, that is where it starts.
BRUSA chargers do not inherently overcharge cells. You have to program them to do this. Or you have to control them by CAN bus to do this. In this case, they were controlled by the BMS via CANBUS and I have seen this connection and setup with my own eyes.
There is NO question it was the BMS that failed, the cells WERE overcharged, and the fire started there. And the only adult in the room at the time was the BMS. It was solely in charge of terminating the charge process. It failed to do so.
I'm also gaining some sympathy for the position of some of these companies such as Elcon and DeltaQ who simply won't let you reprogram their chargers. If you hose it up, it is still THEIR charger that burned the car to the ground. I get it. It's terribly inconvenient, but I get it.
Jack Rickard
Another cracking good film Jack. I bet the Dutch couple will also hold you in great regard. I envy their adventure.
ReplyDeleteIts a wondrous world when batteries cannot be charged from a temperature of +40C. It's not so much the temperature but the mindset to say "no" rather than consider some way of solving the problem of warm batteries. Read on..
I like the kokam battery but as it is, it seems delicate and naked. It's really only one component towards an encased battery system. One could easily weld up a structural plastic casing. Interleaved between the batteries should be plates one can pass warm or cool air between as and when required. Something thin, simple, light and cheap like 2mm or 4mm Coroplast should work fine.
Tesla's new car is a fine example of the advantages of designing an electric car from the ground up. I think it's not looking so much Aston Martin but possibly the finest car ever to be born on US shores.
Hi Jack.
ReplyDeleteAny chance of messing around with fast charging from a battery bank one of these days? It could make an interesting topic for a show and you seem to have a lot of spare batteries lying around to play with.
Nick F :-)
Jack,
ReplyDeleteThe fill bottle on the speedster should be mounted at the highest point, it looks like the controller will be fitted in a higher position compared to the fill bottle, the water will drain from the controller and seek the lowest point (the fill bottle)
Hi Jack,
ReplyDeleteIf you want to ensure there is no water locks, or air in the system on speedster uno, the fill tank has to be positioned at the highest point, it looks like the controller is going to be mounted higher, if so, the water will drain from the controller back into the fill bottle.
Jack,
ReplyDeleteThe Toyota Prius invertor Water pump P/N I supply you is the new and improved model from the recalled version, the old P/N was G9020-47030 and the new P/N is G9020-47031.
I strongly suspect this is the reason why the pump is so cheap.
Simon Wilkinson
Your New Zealand Correspondent
Hi Jack!
ReplyDeleteNice show again! The Uno looks like its going to be a real killer.
Now I have been thinking about the SOC of the cells that you are talking about.
I had a look at the tests you did in the April 16 show last year. It shows that the terminal voltage goes below 3,15V at 50% discharge. And at 50% charge(from fully discharged) the voltage is over 3.45V. How long does it take for these voltages to get back to the 3,300 that you have recorded on your cells now. And in addition to that: What happens if you discharge 5-10ah from one of those cells and the let it sit for that amount of time again to let it stabilize. what voltage would you have then, Would it be detectable if you added or removed 10ah at or around 50% SOC by measuring voltage after letting the cell rest ?
I plan to test this with Thundersky cells.
But if you have any info on that already it would be good to know. Otherwise I will just let you know as soon as I have conducted my test.
Best Regards
/Per
FILL BOTTLE. Well, we'll see. Yes, I think the controller WILL drain into the fill bottle. Now with it drained, the fill bottle will be filled to a certain level (like the level we fill it with.
ReplyDeleteIf we turn on the pump, it should take it from the fill bottle, put it through the heat exchanger, and into the controller. It then returns to the fill bottle. As long as the pump runs, there will be fluid constantly moving through the controller. If the pump stops, it will drain back into the fill bottle.
But the fill bottle has much more capacity than the controller. The difference is unlikely to be a half inch. So how is this a problem?
Now I have air in the controller each time the pump starts, but it is a single lane system. As soon as the pump takes fluid from the bottle and starts pumping, the lines fill up - all the way back to the bottle.
We'll actually put the pump on a little cheap thermostatic switch I've found that we'll attach to the controller - 45C it will connect the pump ground.
What am I missing here?
Jack
FAST CHARGING. There's every chance we will mess around with this at some point. I'm just not sure what the point will be.
ReplyDeleteRight now, I have in my head a charging device that is broken into two pieces. The buck/boost circuit would be in the car. The rectifier would be on the wall or a Texaco Fire Chief. Kind of a two piece charger.
The rectifier would make 330vdc. The buck boost would work from any DC source - theoretically 12vdc to 600vdc. A battery bank could not only be connected alternately to the 330vdc rectified, but in theory, I could even charge the mother bank with the 330vdc rectified whenever it wasn't hooked up to the car.
On hooking to the car, fast charge current would come BOTH from the rectifier and the AGM/Pb battery bank.
So I'm picturing an onboard charger that would work from any DC source. In this way, I could also use one car to "jumpstart" another, even if they have different pack voltages.
It would also allow me to then migrate that to a ChaDEmo or other fast charge public system.
So that's the thought on fast charging. It is easy to do a little of it. But I'm still letting this concept gestate a little.
In the interim, it is just unnecessary. At home, you don't really need it. I just have no need for fast charge.
Where it would come into play is intercity driving. And none is available as yet. So it's kind of a pre-pregnant idea. I'm interested. Just not feeling much of a time pressure on it.
Jack Rickard
Per: I have two TS-90AHA which I think I will charge to a 100% and discharge 4.5Ah (0.3C in 10 minutes) each day and measure the OCV right before the discharge. Should we coordinate our effort ?
ReplyDeleteMartin.
ooerr! I thought a 1c full charge or discharge takes one hour.
ReplyDelete4.5ah/90ah = 0.05c? Or invert it for a 20 hour discharge.
How many years do you expect this experiment to carry on for?
I have absolutely no idea why people want to play with maxing their batteries away or measuring the nth degree voltages on part charge.
If Jack is spot on about battery charge as received. And one battery 'might' be weaker than all the others. One only need to monitor that battery for top and especially bottom charge.
Personally, it would do my head in.
Hey Jack, you might be doing this already but for your "BMS" Mil spec connector on the belly pan cells, I would add a small 1K resistor on each of the wires, right at the cell's poles. You can then safely short the wires at the connector point (or anywhere along the route) without fear of melting something.
ReplyDeleteJR
Andyj: You don't get it. This is not about charging cells to the nth degree, but trying to get solid data on OCV (Open Cell Voltage) versus SOC (State Of Charge). And yes, it will take about 20+ days to do, but it will be an automated setup. But in order to get a known SOC to begin with the cells will be discharged a bit (10Ah or so) and charged with 0.3C in order to have the cell at the defined 100% mark. A defined mark could also be done at the bottom at the charge curve (2.8V in this case) but then a defined charge would have to be applied afterwards. This first approach is more time efficient.
ReplyDeleteI think I will discharge one of them to 95% and let it rest in order to track the OCV as time goes by.
Martin.
JR: A 1k resistor is not quite enough. A setup like this will require more like a 47k resistor in order to be safe. Assuming the worst case the voltage is about 163V (51 cells at 3.2V each). If you put 163V over a 1K resistor you will get quite a lot of heat in a small space (27W). A value of 47k would only burn up 0.6W and even that can pose a problem in some confined spaces.
ReplyDeleteDepending on the measurement equipment and setup, this series resistor might ruin the measurement. I will be using fused connection in my car if I get the time to add the measurement equipment... ;-)
Martin.
Hey Martin, Jack wants to measure the voltage on each cell so typically 2 wires are run from each cell to a multi-point connector where they can be individually measured. With this configuration and a 1K resistor, max current is ~3mA. I wouldn't doubt he already does this.
ReplyDeleteJR
JR: If I attach anything more or less directly to a cell I will make d... sure that whatever it touch either by purpose or by accident (failure) it will not burn down my car. I don't know where you get the 3mA from, but it is not from a worst case scenario. Normally you will have to assume that any wire can more or less short to any other wire in your car. So look at the surroundings and try to imagine where a wire can touch. A car is a rather harsh environment and insulation on a wire is not very thick... If you think a 3mA is the worst case scenario, you'll be in for a surprise some day.
ReplyDeleteJust my $.02
Martin.
Hey Jack
ReplyDeletePretty brave putting a battery with bare terminals into a bare metal tray!!!
You would have had an interesting introduction to welding aluminium (why do you guys pronounce it ALUMINUM - shouldn't other materails be called TITANUM, PLUTONUM or LITHUM :> etc. in that case...). I hope you would have posted the video if anything actually had short circuited!
Jack - thank you for the piece on self-discharge. That looks like a critical result because it would be the only mechanism for drifting out of balance in a Pb-ish sort of way.
ReplyDeleteOh and yes I was mortified to hear an Englishman at Tesla talking about "Aloominum" rather than "Al-you-miny-yum". Can't have everything I suppose
The 47k resistor is actually a very good idea. It won't effect the measurement at all. I'm not charging/discharging through this connector, just measuring voltages. They would limit current quite nicely. Yes, it will throw the measurement off, but with a 10Meg input impedance on even a cheap meter, you won't be able to detect it.
ReplyDeleteAs to shorting out the cells on the alyouminium, not really a problem. I used some 1/16 inch high molecular weight polyethylene to line the boxes. The boxes are kind of narrow, and this stuff is stiff and tough. So I just layed it across the boxes, and started setting batteries on top. They didn't even cause it to sag. This was going to be tough.
So I took a MAAP torch to it. Amazingly this stuff nearly enough plastiformed to the box. The cells drew it right down into position. I folded the excess over the top, and heated that too. Then put more cells on top of the HMWP until it cooled. I've got virtually shrink wrapped batteries. This stuff has great dialectric strength, good thermal properties, and doesn't weigh much.
I used two gland nuts in one end to run a 1/0 cable between the two boxes and made the cell connections there.
We're fashioning little copper straps to interconnect the lengthwise cells in the smaller box now.
I think it's all going to work out.
Jack
One thing that adding wiring to cells can create is a ground fault. Also having the cells under the car with a less than perfectly sealed enclosure will also provide a potential for a ground fault. Higher voltage packs make it even easier than lower voltage packs. Some chargers measure for ground faults, some motor controllers measure for ground faults. If they find one either operation stops. Even having exposure to morning due can create ground fault paths that you cannot even see.
ReplyDeleteIsn't their a manufacturers requirement to capture these large format cells in compression so they don't expand?
You probably know this already but it's a safer idea to make sure the most positive pin in your connector is as far away as possible from the least positive pin. Even though connectors are often rated for voltages higher than this application, they will find a path between high potential pins.
Glad to hear you liked the Reep Systems BMS gui, Jack. There are other brands out there with nice front ends that might not have the burning car stigma.
Sounds like you might be softening your position on battery monitor use Jack. I shudder to think....
"Martin Zacho said...
ReplyDeleteAndyj: You don't get it. This is not about charging cells to the nth degree, but trying to get solid data on OCV (Open Cell Voltage) versus SOC (State Of Charge). And yes, it will take about 20+ days to do,.... "
Too right I don't get it! lol.
Too many variables for a bit of data that is useless in real world use. You cannot simply stop your car at the lights and state you have x% charge left! You'll have to include temperature and pressure, differing charge/discharge rates and frequencies, the age of the cells and so forth. It will need to stand for a good while too!
The data is not solid data. Google it ;)
mikep_95133
ReplyDeleteThe BMS was remotely programmed to reduce charge demand to the point where only one of the two very expensive chargers were required. All due to BMS issues. This is why no real world manufacturer has or will ever use them.
What a waste! Money, complexity, volume and dead weight on a round the world tour. :(
Jack was being Jack, a great guy with a lovely couple. He did say on here, "REAPer". >:-)
There IS a requirement to clamp the cells if you use an unregulated charger:
http://gwl-power.tumblr.com/
A bit like needing to sit in a tank while at war.
AndyJ:
ReplyDeleteActually pack voltage is a very good indication of SOC and it stabilizes quite quickly when the car comes to a stop. Here's why we don't rely on it solely. It is terribly nonlinear.
During the first 80% of your pack capacity, it very slowly but stably declines. And the voltmeters we use aren't that accurate to give you a real "fuel gage" effect. During the last 20% of discharge, they are really quite useful. But if you're maximizing pack life, you really don't normally want to be there at all.
Used in CONJUNCTION with an AH meter, the two are wonderfully interactive and informative. You'll quickly get a sense of where you are in the pack. It's not nearly as mysterious in practice as it sounds discussing it.
Jack Rickard
Martin:
ReplyDeleteVoltage vs SOC. We've published dozens of Xcel graphs of cell discharge curves. Here's why it's a losing proposition for me:
1. We all use different cell brands.
2. Those using the same brand use different sizes.
3. We all have different pack sizes.
The result is whatever we develop in regards to cell voltage vs. SOC is nonsense for everybody else. It's a very good indicator, though very nonlinear. You wind up with a very slow decrease for 90% of the energy in your pack, and a very fast decrease for the remaining 10%.
But it is entirely worthwhile to do this yourself on YOUR cells and YOUR pack and indeed, I assume that you almost can't avoid doing it. If you have an AH counter of some sort, and an voltmeter, you'll naturally develop a feel for it over time. But sure, you could graph it.
The problem is with a different brand of cell, or a different chemistry of the same brand, or even a different size of the same brand, you get a different curve.
But the curve is very informative. It's just of very limited application to the car you're doing it for.
Jack Rickard
Andyj: You still don't get it. The test is not intended for use in the daily use of the car, but just to verify if the OCV gives any clue about the SOC of STORED cells.
ReplyDeleteMartin.
Nope. I don't get it Martin. :)
ReplyDeleteJacks right. I wouldn't test one cell then multiply that figure. Use the whole pack.
Once your pack is outdoors on your famous frosty mornings, its all out of the window.
Jack says you can use of it. I agree and hope you succeed but it will never be commercial quality.
@Jack and Andyj:
ReplyDeleteIt was not a test to attemp to have a voltage graf to use for driving. But more a way to actually be able to verify that Jack is right when he says that there is no self discharge on those cells, if the OCV cannot tell you if your at 40% SOC or 60% SOC, how can you tell if there has been self discharge or not by measuring OCV in the first place. The only way to actually measure if there was any self discharge during these 15 months is to do a discharge to empty and measure the Ah that was left in the cell.
That was why I proposed the bleed a known amount of Ah and let the cell settle and then measure OCV to see if it changes, or if it is more or less the same between 30 and 80% SOC.
The reason why cells are delivered from factory at 50% SOC is because they store best at that SOC. It gives the best Shelf life.
But I suspect that the LiFePo4 could also handle a higher SOC for storage since they are more stable than LiMn or LiCo.
Regards
/Per
Per.
ReplyDeleteActually you've got a couple of things mixed together here and not precisely on the ball.
First, voltage does not have a spread as to SOC. It is quite exact. The reason we don't use voltage to determine SOC in a car is that our ability to measure it fails in the face of the scale of it. IT is NOT true that there is no relationship between voltage and SOC. There's quite a definitive relationship.
Determining SOC by voltage can be problematical unless you have a very accurate voltmeter and know what the voltages are across the SOC curve.
The reason the cells are shipped at identical SOC is not so much a function of "that's a good SOC to ship." That's true. But the result is that they have a fairly automated test and conditioning cycle for the cells that simply winds up at the same place on each cell - because it's automated. The purpose of this is actually to sort and ship cells by a selection process because their manufacturing process is not very consistent. But the result is the cells all come at almost exactly 50% SOC. That's a byproduct more than a mission.
My demonstration is not dependent really on determining SOC, and there is no purpose at all in then measuring AH out of the cells really. I already know the answer. You can choose to accept that or not.
ReplyDeleteI found it indicative and noteworthy, that in a pack of 60 cells, undisturbed and unused for 15 months, they are ALL within one onethousandth of a volt, meaning very specifically all at the same SOC. And further noteworthy that they were all within 2-3 one thousands of a volt of a shipment of identical cells received less than 24 hours previously.
To my way of thinking, that is the END of both cell drift and cell self-discharge. You all can (and will I am frustrated to admit) debate this ad nauseum quite without any further input from me. But you are counting angels on the head of a pin. It doesn't matter about the meter, the temperature, or the phase of the moon. As a fortunate matter of circumstance we happened to have 60 leftover cells from a shipment 15 months ago, and 50 brand new delivered ones, side by side. We took the measurement, and the results were truly ridiculous.
No drift. No self discharge. Nada. Knowing as I do that after the SEI layer is set, there is no electrochemical possibility of such and never was, I thought it a good time and manner to illustrate this in terms that anyone could view and understand.
That there will be endless debate on Endless Fear and DIYjunk is inevitable. I'm sure I was standing in the wrong position, had too much natural gas heat going, and not enough light to satisfy the demonstration. It was probably all done with mirrors anyway, I am sure you will hear.
What I cannot comprehend is the desperation to cling to voodoo mythology and embarassingly amateurish explanations by David Andrea, Dmitri, et al. Posing as professionals and even "experts" they continue their witchcraft and lunar worship like a bunch of pagan wickens in the woods.
We don't actually have any vested interest in this beyond the following:
IF you succeed in converting your car to electric drive and operate it successfully in public, there will be more of this. That is a good thing for us.
If you fail, and hobbyist converted electric cars come to be viewed as dangerous experiments posing a public safety issue, we lose.
Disinterested but imminently knowledgeable parties such as Jay Whitacre have publicly acknowledged my positions on this. The ONLY place you will hear an opposing view is from people selling BMS systems and the odd person online posing as a a knowledgeable source/expert by typing themselves smart and repeating the BMS designer mantra.
So I'm a little frustrated at all the confusion. My motivators are:
1. Avoid needless and stupid regulation limiting activities in electric vehicle conversion.
2. Further where possible the abilities of individuals to successfully make those conversions.
Those two things met, we will thrive at EVTV.
The opposing camp also has a motivation - a prepaid order, usually by check, in non-recoverable funds. No warranty express or implied.
As to credibility, why would we put it in a video at all, and risk credibility, if we didn't already have pretty good assurance of the answer? I can put anything in the video I want - including cartoons starring "Carbon Footprint."
Jack Rickard
AndyJ: No real world manufacturer uses a BMS? You have not done your homework. Are you talking hybrids or EV's? Both are very different applications of the same battery technology. Hybrids don't use a BMS due to super shallow charge/discharge cycles. The hybrids do eventually fall out of equalization. It just takes 100,000 miles or more. Ev's have very deep charge/discharge cycles. Totally different animal. ACP still uses a BMS after 17 years. A really nice one at that.
ReplyDeleteJack: You don't have any end of life pack experience. That's why you don't know the obvious. All cells diverge in capacity and impedance over time. It's your money. I think you are spending and video taping yourself smart. Accusing those of what you are guilty of yourself. Classic. You'll cave in to at least a monitoring system someday. You've already started down that path by adding wires to your battery tray. If you really trusted your beliefs that lithium does not need to watched over, you would not be adding those wires. We will welcome you to the darkside, someday....
Mike:
ReplyDeleteACP has been in business in electric vehicles for 17 years, and has basically left a tiny smear on the road - they hardly exist. Their "BMS" caused endless problems with the MiniE and a lot of their controllers had to be replaced. Citing them as an example is a very poor choice.
You are hopelessly confused. There is nothing wrong with monitoring cells as long as you don't destroy them in the process. That doesn't quite comprise a BMS as I understand them. But yes, in order to declare victory, I expect the BMS crowd to gradually retreat into "monitoring". The shunt balancers are already more or less disappearing and the charger supervisors will be next to go.
Your divergence theory is scientifically implausible. Where do you get such information? No one has more end of pack experience than I do. I ended one set of GBS cells in three days. You have jumped in here and blindly started typing in every direction without watching any of the videos or reading any of the blogs apparently.
I've asked you several times to return to the forumsewer from whence you came. What's the problem pard? If I'm so errant, why do you keep hanging around in a forum so clearly out of your intellectual league?
In any event, for your edification to the point where you can absorb it, there is no divergence in capacity nor in impedence. They do not self discharge. They do not drift. All of that is the direct result of your misguided attempts to MONKEY WITH THEM LIKE AN ORANGUTAN. If you will remove the BMS you are using to measure all of that, it all goes away, because the cells never did any of that. You're moving the cells around with your shunt balancers. There is no mechanism in these cells for this drift at all. They are NOT Pb or Nimmy cells. They never were. Your insistence that they act like the cells you are familiar with in the past has lead you to read messages from God in cloud formations.
Jack Rickard
Jack.
ReplyDeleteWhat are your thoughts on smart controllers and smart chargers made from the same manufacturer with data acquisition? I apologise for the vagueness of the question.
Padraic
jack forget the small wires in the flat battery tray. put holes near the terminals and put in rubber plugs to make them watertight.
ReplyDeletedhiller
My thoughts on smart controllers and smart chargers made from the same manufacturer with data acquisition?
ReplyDeleteI have no such thoughts. I don't even know what you're talking about.
Did I cover this?
The heroic flaw in most of the BMS systems with regards to charge termination, is that the charger is set to RELY on the BMS for an active message to CEASE.
This is heroically stupid. But it is very easy.
The proper way to do this, at its most basic level, is that the BMS should command the charger to CONTINUE. Lacking such a command, the charger should instantly SHUT OFF.
This is because software has "bugs" and often they are catastropic in that they cause the microprocessor or microcontroller to "crash". There is no real debris, the microcontroller or microprocessor exits the programmed command stream and is at a little loss for the next instruction. On your PC, it feels like it is locked up.
Now before all you software geniuses jump in to explain how to avoid such bugs, I would point out that my first NONMILITARY software adventure was on the original Sinclair ZX80 kit, using a Zilog Z80 microprocessor in assembly language. I've written a lot of software since, none of it without little "aberrations" from time to time.
Today, I use a program called Final Cut Pro to edit these videos. It's been out over a dozen years. It is on version 7.02. I can crash it at will so thoroughly it exits the program entirely. This is VERY mature software used by MILLIONS of people.
Software is, well soft.
The charger manufacturers do pretty well using a couple of techniques. First, the software is very limited. Chargers don't do very much, and they don't have many features. Limiting the scope of your program is an excellent way to ensure reliability.
As an adjunct to that, it becomes easy to test. And so they test them rather thoroughly before release of the product, largely because of the huge liability in selling a charger that can destroy valuable batteries and houses.
There isn't really much to pick from. We are badly in need of a good charger solution for these cells. We're working with a company now that assured me they were "shipping" last August. it was a bit of a lie fortunately, because we immediately found a dozen problems with it, and they are indeed working on them as we speak and I expect a new unit this week hopefully.
ReplyDeletePutting a BMS in charge of a charger is lunacy. In this case, the perfectly good Brusa was put under control of the REAP systems via CAN BUS.
IF the REAP crashes, it provides NO further signals to the Brusa. The Brusa sits there a charging and a a waiting, waiting and a charging, waiting for that command to shut down. It never comes. But all the fires of hell soon do - right about the end of the normal charge period.
An improved way to do this is to have the Brusa in a constant state of shut down almost from the beginning. Each second the REAP would send a keep alive to prevent termination. If it crashes and never sends it, the BRusa just stupidly shuts down entirely. Your car doesn't get charged, and you need to notice that in the morning before your 100 mile commute, but nothing burns up.
In real hardened systems, you actually have two entirely separate microcontrollers, with entirely separate inputs. They would monitor each other, and the charger. If any controller notes a failure of the other controller or the charger, the charger shuts down. The charger also monitors the two controllers, and if either fails to respond, the charger shuts down. This is the three-man vote. Quite common in military systems requiring a high degree of reliability.
The simple minded BMS systems I see offered, are just hysterically amateurish - particularly given the horrific consequences of failure demonstratted by two Qashqais's, the LincVolt, and a half a dozen others - all of whom any moron can trace directly to overcharging as a result of BMS failure. The Manzanita's and Brusa's involved in these two cases I can personally attest are quite reliable in their ability to terminate a charge sequence by themselves. While the Manzanita has some quirks regarding voltage measurement, the disparity in not such as to be catastrophic to the point of fire.
So I gather you are talking about an advanced charger, with more data acquisition abilities. Not a bad idea. But remember that one of the reliability factors is to keep the original software scope very limited.
Jack
Jack
ReplyDeleteThank you. You have answered my quesition. I know the question was poorly worded.
I am not an expert, and I hope I have never presented myself as such. I do hope I am not wasting your time because I will now reveal my ignorance by asking a stupid question. But my original question may make more sense when I ask this question. Would it be worthwhile (or possible, or needless?) for the charger to measure the current used to charge the pack, and subsequently the controller to measure the current taken from the pack? To measure voltage and current either side of the pack. If anything, one can gain an understanding of the efficiency of the particular pack, using particular cells in a particular car.
Feel free to call me ignorant, but to deny my ignorance would only serve to preserve it.
Padraic
Jack: I've always been a proponent of monitoring and managing. No confusion here. I think you confuse BMS with BMS reliability issues. I still think you are creeping over to the dark side. We'll have some wine for your arrival.
ReplyDeleteDivergence is what people who have been around EV's for a while observe, pack after pack, car after car, owner after owner. It's as old as EV's themselves. I didn't invent divergence, not by a mile. But I watched and learned from it.
Your constant jump to conclusions is the main tell of your level of experience to me Jack. It's easy to tell what you know from what you don't say as well as what you do say. You can't spend your way smart with EV's either.
Mike - Spending yourself smart is the ONLY WAY to get smart! It's called R&D in my profession.
ReplyDeleteAnyone who does not require adult supervision is clever enough to assimilate and analyze Jack's data and experiences without bias. If you so vehemently disagree with Jack's point of view then why are you hanging around his web site like a bad smell. Your post are not going to change Jack's opinion and frankly it won't change mine either.
Per: I think we are on the same page here :) Could you email me so we could try to coordinate our effort ?
ReplyDeleteMartin.
>The proper way to do this, at its most basic
ReplyDelete>level, is that the BMS should command the
>charger to CONTINUE. Lacking such a command, the
>charger should instantly SHUT OFF.
Elcon (now TC Charger) ahve their chargers setup just like this.
short the enable pin to ground, and the charger just does the cc-cv curve.
open it and it shuts down.
that way, if a BMS locks up or otherwise fails, its no worse than if there was no BMS at all (ie no fire because the charger just followed the normal cc-cv).
I hope that Lithium Balance has improved their system since July 2009 when it did not stop the charge of a battery pack in a converted Jaguar XF Thursday the 9. of July in 2009, and the car burned to the ground.
ReplyDeleteIt was the first time that the Car should charge up fully, the fire broke out at night at the end of the charge, the Police have concluded that the fire started in the batteries, and it was because the charge was not stopped by Lithium Balance BMS system.
The car was converted by http://cleantechmotors.com/ and it was supposed to transport VIPs for the climate summit COP15 in Copenhagen in December 2009, and afterwards Cleantech Motors wanted to build and sell these converted Jaguars. Unfortunately the project is stopped completely because of the fire.
Lithium Balance will not acknowledge that there system failed, and that is was their fault, but they might have learned something from the fire anyway and have improved the BMS system. I hope that they will not cause any more fires, these fires are not helping.
/Carsten
Mike I do not understand why you keep making these undocumented claims. You are just writing words on a web page, you can write it over and over again, that does not make it true, it does not make it right.
ReplyDeletePlease stop the undocumented claims, there are two ways as I see it, bring some documentation to back up your writings, or stop writing undocumented beliefs as the truth.
Believing something and writing theories about why it is true, does not make it true. You have a lot of theories, my suggestion to you is that you try to describe the experiments that you will or would do to either confirm your beliefs, or discard them as being wrong, depending on the result.
/Carsten
ok, every body. Jack has proved with the horses mouth himself. The man who knows everything about batteries. He says BMS are NOT required on LiFePo4's. If a world leading professor on the subject says so? I believe him.
ReplyDeleteThe issue of batteries losing sync with others. JACK! HEADS UP!! IT REALLY DOES HAPPEN!!!
To everybody who has a BMS strapped across their batteries. hahahahahaha! Imagine my shock!
Jack knows this full well. He had a little instrument across some, not all of the batteries for a few months. Then found those cells drifted down.
Earlier, talking about "no BMS" cars; I meant REAL electric car manfr's. Toyota's Rav4 EV, GM's EV1, Tesla cars. None burned down and none have a BMS.
Andyj:
ReplyDeleteYou are precisely correct. Divergence is VERY real, if you are monkeying around with your cells. That's the only cause I've found. Yes, we left two seemingly innocent Cell Log 8s plugged into the Mini Cooper, to measure cell drift if you can picture the irony. The cells didn't drift enough in two months to make even mildly interesting data. But instead of plugging them in and unplugging them each day, I just left them connected for ANOTHER two months. Sure enough, they bled those cells down over time.
SO those 16 cells DID diverge, because I had a parasitic load connected to them. No matter how miniscule the current drain, if it is 24x7 it adds up over time.
Jack RIckard
Carsten:
ReplyDeleteFascinating. Mostly fascinating because Nina Rasmussen told me that because of the TWO fires with the REAP Systems, Ecklund was planning on replacing the REAP BMS with a Lithium Balance BMS system.
I've heard of out of the frying pan into the fire, but how do you jump out of the EV fire into the EV fire. And what sense does THAT make?
Jack
Mikep:
ReplyDeleteYou can't spend your way smart with EV's - what, by trying different equipment and measuring results? But you CAN by endlessly typing nonsense into a computer screen in the hopes that by repitition you can alter reality?
And YOU, el brilliant one, can tell all by what I DON'T say. I think I'm getting a clearer picture of the problem here. Are you on medication? Or more to the point, are you TAKING your medication...
Jack Rickard
Ironically LITHIUM CARBONATE is used in the treatment of mental illness. Maybe Mike could open up a cell and lick the anode?
ReplyDeleteRegarding the Lithium Balance and Jaguar. As far as I know the BMS was _not_ connected due to a human error. I do have it from the people involved with the post-mortem. The BMS did order the charger to shutdown and had for several hours, but a signal cable was left unattached....
ReplyDeleteThis just put an emphasis on having redundant and independent systems when dealing with batteries. The charger should not be able to do any harm on a nominal system (CC/CV with a timed shutdown is a good choice). Then attach a simple voltage and temperature supervision in order to shut down the charger if the batteries does gets to hot or rises above a certain level. All three systems should be totally independent and have a failsafe de-energized state (e.i. if a fuse blows to the voltage supervisons, then the charger should be disabled).
Martin.
Jack,
ReplyDeleteIf you were to run cables for your Celllog to get a snapshot of your cells in the Belly Pan, What would you use to fuse the line?
AussieJoe
Completed and installed the belly pan yesterday. Very successful. Decided NOT to wire for monitoring. Monitor for what? Nothing is going to change.
ReplyDeleteJack Rickard
That's good to hear. You have eliminated, what, about 2 dozen points of failure from UNDER THE CAR?
ReplyDeleteI'm relieved, and I just have to point out you guys are really maturing as car builders. Any hot-rodder can see the quality going into your Speedster. It used to be expensive parts hacked together. Now Brain, Matt and you are really putting things together much more artfully.
The belly pack is a great example. One box and it doesn't hang too far down. No more monitoring spaghetti a la the Mini. You now have the confidence to think through and reject throwing failure points into a spot because you can't easily get there to monitor. The unit looks good so far. Melting the plastic around the batteries is a great idea. The only thing I would have added is a pair of vertical longitudinal aluminum skid skegs, hanging 3/4" or so below the pan on either side, to protect the cells when the inevitable high-center bump happens. And well-considered pan drainage.
Two things that would improve your systems routing within the car: Nutserts and Adel clamps. The nutserts are great where you want to screw something to your fiberglass or sheet aluminum panels, and the Adel clamps are the best for restraining tubing and wiring runs without chafing them. As the power goes up on this hot rod, the physical demands on all these things are going to increase, too. Consider that regardless the powertrain, you are now building a track-day car, not a daily driver, and its a heavy one, too. Are your brakes up to the task?
Nice work at any rate, men. It really shows.
TomA
Jack wrote:
ReplyDelete"The reason the cells are shipped at identical SOC is not so much a function of "that's a good SOC to ship." That's true. But the result is that they have a fairly automated test and conditioning cycle for the cells that simply winds up at the same place on each cell - because it's automated. The purpose of this is actually to sort and ship cells by a selection process because their manufacturing process is not very consistent. But the result is the cells all come at almost exactly 50% SOC. That's a byproduct more than a mission. "
I do understand the selection process. And it is really good that it is automated because it gives us a good start when assembling the car.
But I can be pretty much sure that IF the best way to store these cells where at 15% SOC they would all come at that SOC.
Regards
/Per
Martin I have also talked to the owner of the Car, it is some time ago and I did not know what I do now, therefore I did not know what specific questions to ask. But after the conversation it was my clear understanding that the BMS system was the cause of the fire.
ReplyDeleteAs a side note, if a faulty wire could cause a fire, burning a car to the ground, that is a faulty design in my opinion, even if it is not installed. The wire or the connection in either end of the wire could fail at some point, and then your car would burn down the next time you charge it, what is that about???
/Carsten
Jack:
ReplyDeleteYes it is a little strange, but I think that you have explained it your self. These Lithium cells are very different from the “older” chemistries like lead Acid and NIMH, and EV builders just do not know.
It is not a secret that I have probably all of my knowledge about batteries and the way they work from your, the videos and this blog. If I did not have this knowledge, all I would see is that almost everybody that speaks up in the EV world, says that you need a BMS because of … Why would I not trust that? The truth is that I would, the reason I do not, is because I now have some degree of knowledge about Lithium chemistries.
As I understand it batteries and battery chemistries are very different from anything else in an EV. If you understand the mechanics and electronics, there is no reason why you should know much about battery chemistries, then you need to trust that others do. If you make the reasoning that there would not be “professional” companies that make BMS systems if they were not needed, then you trust that they know what they are talking about.
My two cents
/Carsten
Jack:
ReplyDeleteYes it is a little strange, but I think that you have explained it your self. These Lithium cells are very different from the “older” chemistries like lead Acid and NIMH, and EV builders just do not know.
It is not a secret that I have probably all of my knowledge about batteries and the way they work from your, the videos and this blog. If I did not have this knowledge, all I would see is that almost everybody that speaks up in the EV world, says that you need a BMS because of … Why would I not trust that? The truth is that I would, the reason I do not, is because I now have some degree of knowledge about Lithium chemistries.
As I understand it batteries and battery chemistries are very different from anything else in an EV. If you understand the mechanics and electronics, there is no reason why you should know much about battery chemistries, then you need to trust that others do. If you make the reasoning that there would not be “professional” companies that make BMS systems if they were not needed, then you trust that they know what they are talking about.
My two cents
/Carsten
"But I can be pretty much sure that IF the best way to store these cells where at 15% SOC they would all come at that SOC."
ReplyDeleteI do agree here. There is a huge penalty in term of money and time in charging (or discharging) these cells any more than absolutely needed.
I they can cut production time and/or cost by 5 - 10% by simply shipping the cells with less charge - they would do so.
Martin.
Tom:
ReplyDeleteThanks. I think there has been a progression and rather a change of mission along the way more than anything.
Our first car three years ago was kind of an experiment to see if these cells actually made an electric car viable, where it had not been before. The cells became available to mere mortals, and we started playing with them on a couple of GEMS, comparing them to AGM, flooded lead, and even capacitors.
The original speedster was to put all that in a car and see if it would WORK. We were stunned and delighted with the results, and wanted to share it with the world. But the message wasn't that we were ace custom car builders. Indeed, the message was that a 54 year old guy working half days falling down drunk and wearing yellow shoes could cobble together an electric car with 100 mile range and 100 mph speed in his own garage. Can electric cars be built? That's the end of THAT question.
And our point all along is that ANYONE can do this. Yes, it takes time. And yes, it takes some resource. But I didn't change my own oil when this started. We are NOT expert mechanics here. I don't actually know now what a Nutsert or Adel clamp IS.
But yes, along the way, I guess we've gone beyond the experimental "can it move" stage and we want to make better cars. And so we are having to acquire more skills. To address the chafing/movement question, we've gone to what I call gland nuts but I think are more properly "cord grips". They are a bit pricey, but really isolate the cables and allow you to pass through thin walls and things without chafing and much better than the grommets we were using.
Once I learn what a Nutsert is, I'm sure it will get better yet. Because the mission has rather moved in that direction. And you and other viewer input is largely what has moved us there.
Jack Rickard
Jack
ReplyDeleteWill we see an almost complete MK II in the next show? Or am I way off? Or should I just wait and see?
Personally I can’t wait.
Plus. Has Matt any plans for a build? What car would he consider converting?
Any news on Brain's Smart? That is an interesting build in my mind because it's such a small car. A good demonstration of what a few men in a shed can do with the technology available. Also, I can't see it been done with Pb.
I apologies for all the questions.
Padraic McDonnell
Jack:
ReplyDeleteThis is such a recurring theme it can't be said often enough- don't build a crappy car.
Every EV converter I've seen who has provided some post-build feedback has said: "I should have started with a better glider."
Every car hobbyist I know, starting with me, has said about a completed car: "I should have started with a car in better condition."
Your experience is quite similar, but coming from the opposite direction. You've always had new, nice components, and you started with a $22,000 glider. Your take on 'don't build a crappy car' is more like "we can do much better on the third, or even the fifth one," which is obviously true, but you've also changed in the way you experience the cars. You just won't do things the old way any more.
The bottom line is truer than ever, though: DON'T BUILD A CRAPPY CAR. If you start with junk, you'll wish you hadn't; if your build quality is low, you'll want to redo it.
And Jack, you fellas are going now going down a special developmental road with your little red rocket- I call it the "What's Next" game of serious hot rodding:
Tripling the output of the motor is awesome, (it would have cost you TEN TIMES more for similar gains with an ICE,) but unless you dial back both the rate and limit of power delivery in the controller, you're going to play What's Next, i.e. what part is the next to fail, every time you stomp on the accelerator.
My money is on the clutch, then the coupler or the transmission, but ultimately- both. Nothing on that car is designed for the torque you can make. I've already hinted the brakes won't be up to the combination of the extra weight and higher speeds. Ditto the dampers, and probably the springs, though they are somewhat adjustable. You'll need better tires, too.
This isn't a bad thing, just what happens when you massively increase the power. Its the hot rodder's critical path. I know you never wanted to join NEDRA, but you're probably going to wind up asking those guys about driveline parts. Maybe the time to check with them is right now, before you break whatever is next.
In these respects a hobby EV is, at long last, no different than any other hobby modified car- you need to cope with higher performance levels- and that's refreshingly cool. Big change from your Pinto days, yes?
Now go grenade a spider gear...
TomA
Tom:
ReplyDeleteI'm afraid you're right. T his was George Hamstra's discussion of going to the 1/2 inch terminals. The terminals were blowing off in some high end regimes. He can cure that by going to bigger terminals, but it tends to just "move the problem" to some other component.
In this case, it has been more of an experiment to see how far to take it. The Soliton1 was simply unavailable during previous builds. Matt was talking about how needlessly LARGE it was today. I was explaining to him thermal issues are the REAL issues with electric cars all across the board, and that I liked the dual air/liquid approach they had taken and why that simply leads to bulk in the device itself. But if you are going to claim 340kw, that's pretty much where you wind up - mostly heat sink.
The amazing part of this "rebuild" which started out pretty modestly, is the batteries. I thought we had both versions of Speedster pretty much packed full before. We just worked out mostly another six cells in back that will actually fit pretty neatly by putting the Soliton1 on top of the motor. We're up to 57 of the 180AH cells for 191 volts and a pack size of 34,380. Total cell weight is up to 712 lbs.- replacing 540 for an additional 172 lbs. It's probably better distributed than it was, and the springs and brakes should be fine as they were a little over the top before. Our 100% maxed out range is 137 but even our 80% range is up around 110 miles. This probably makes a drive to St. Louis quite doable.
The tranny is kind of a beefed up version of the VW transaxle now. I kept the original clutch. It was already some sort of racing clutch. But yes, I intend to use it as a fuse.
Mounting the Soliton1 on top complicates the motor removal slightly. But I can live with it. Recognizing the potential in the Soliton claims, we want to be able to replace the clutch and/or motor without undue heartache.
One thing we learned by acquiring the Spyder, you can build a car TOO well. Apparently this Scott Smith, the fabricator on that project, is quite good. Every wire is EXACTLY the right length. And the layout LOOKS great. Until you try to change something. To remove any part, requires the removal of six other parts, which of course each require the removal of six parts. The result is kind of alocked up rubick's cube and you can't work on it at all.
Jack Rickard
Why are they shipped at a certain SOC?
ReplyDelete“There is no shuttle-based self-discharge reaction in the Lithium Ion cell like that found in the NiMH and NiCd. As the cell ages, the self-discharge eventually becomes zero. Initially the cell suffers from irreversible capacity loss. This is a reaction of the electrolyte with the the active components if the cell. It occurs more rapidly with increasing temperature and cell voltage. For this reason, cells should be stored at 24C or less and between 30-50% state of charge. The lower limit is chosen because they are often stored in packs witch circuitry that demands a small drain on the battery. When one considers the circuitry needed for li-Ion, it becomes the most important source of self-discharge.”
From:
http://www.gpbatteries.com/html/pdf/Li-ion_handbook.pdf
Food for Jack’s opinion.
Regards,
Jan
Jan:
ReplyDeleteI think it is again hysterical/historical thinking. But how this has evolved, on these specific prismatic cells from China, is that they have approached the problems of manufacturing consistency through a selection process. And instead of highly automating and regularizing the manufacturing process, they have automated the selection process. That simply leaves them at the same state of charge and 50% SOC makes sense. It has to be something. ANd top and bottom don't sound good.
My point was and is, the chemistry for what we traditionally think of as both self discharge and "balance drift" don't exist in these cells. And so the many expert geniuses attributing things to that and "having personally observed it hundreds of times" are just generally bullshit artists.
But I detect evidence of a learning curve among the Chinese manufacturers themselves.
Jack Rickard
“That simply leaves them at the same state of charge and 50% SOC makes sense. It has to be something. ANd top and bottom don't sound good.”
ReplyDeleteJack, I don’t understand your reaction. They simply state:
“You should store them with the lowest voltage possible. But we know, that you morons will store them with some stupid electronic circuit that will eventually kill them, and that’s why we ship them with at least 30% and max 50% SOC.”
Well, that’s what they would have written if you where the technical writer.
Doesn't sound hysteriocal to me.
In that handbook not a word that comes close to: “Though Should Use A BMS” It gives you some numbers that are normally used IF you want to use a BMS. And they write that IF you want a BMS, you can do it at cell level OR for cells in series. They are pretty clear about it. No Cell Level BMS Is Required.
The part of stay away from the Top AND the Bottom is not clear to me. That handbook I hyperlinked says the cells age faster the higher the voltage. Nothing about low voltage problems. The Liionvideoprofessor said that these cells will be damaged by reversal. But that’s a long way from a low voltage. A stored cell can never go into reversal.
Could you be more specific about this issue? It’s an important property we should know, considering top or bottom balancing. Well, I know you’re a supporter of bottom balancing, but it gives it an extra strong argument.
And it is important to know how to store a EV for a longer time. Is 50% SOC really the best storage voltage? Maybe not.
Regards,
Jan
LiFePO4 battery builders consistently show increased cycle life going from 100%, 90%, 80%, and 70% DOD, which is probably why they wouldn't store a cell much below 50% DOD. If you have no load on them you'd probably be fine storing them at 40% DOD but I wouldn't go any lower.
ReplyDeleteJRP3
That should have been 40% SOC. Oops.
ReplyDeleteJRP3
Jack
ReplyDeleteThe nutsert that Tom A mentions is also called a rivet nut. I have also heard them called jacknuts, and blind nuts.
The Adel clamp is also called Vinyl-Coated Loop Clamps, or Rubber-Cushioned Loop Clamps.
Both of which are available through McMaster-Carr.
Barry
Something to consider before mounting the Soliton1 over the motor is the amount of waste heat the two devices will produce in normal operation: 0.5-2kW for the Soliton1 but 15-35kW for the WarP-9 motor. One thing people tend to forget about heatsinks is that they are just as good at picking up heat as shedding it... putting the Soliton1 on top of an impromptu 35kW space heater might result in the early onset of thermal derating, even if liquid cooling is used (after all, the liquid cooling loop was designed to remove ~3kW of heat, not 10x that amount...).
ReplyDeleteOne correction, if you will... the CLAIMED rating of the Soliton1 is 300kW, not 340kW. 340V is the maximum allowed input voltage; 1000A is the maximum allowed continuous output current; you can't realistically have both at the same time, but if your pack DOESN'T sag down to 300V when pulling 1000A from it, current is otherwise limited to 900A at 310V or higher to ensure that the 300kW limit is respected. Losses in controllers go up with voltage, too - a fact that seems to not be well-known or appreciated.
Jeffrey:
ReplyDeleteGood points on both counts. I was frankly unaware of the 300kw limitation to 900 amps exceeding 310 volts.
The heat could indeed be an issue. I really do like the location of the Soliton1 over the motor for a number of reasons. First, it actually provides for circulating air UNDER the Soliton1 where on the shelf it was rather closed in.
It looks a lot better on top of the motor, more central to the drive train and kind of "on display".
And finally, there is no artful way to store batteries over the motor. But we can add six to the shelf rather easily.
As to the heat given off by the motor, it is signficant. There is about 1 1/2 inches between the motor and the aluminium plate the Soliton1 is mounted on. I'm hoping this is enough circulating air to negate the effect. But I could be wrong.
We will of course have it liquid cooled albeit interlocked to a 45C thermostat so it does not pump in the dead of winter. But if the Soliton plate goes over 45C, the pump will come on.
If it proves that the motor heat does effect the Soliton1, we have some fairly easy get well. We are currently provisioning a smallish heat sink. This is sort of an experiment in what we can get away with. We're using it with another heat sink on the shelf that is more like a foot square with a powerful fan on it originally procured for this purpose.
So the heat is sort of an exercise in what will be required. I would love to liquid cool the motor as well but no possible design has suggested itself.
If it would generate sufficient heat, we'd actually use it in the passenger compartment. Speedster Part Duh, much lower power, simply does not. In winter, we can barely coax it up to 35C.
Jack Rickard
I have a similar situation with my AC31, the Curtis is right over the motor. I'm considering a sheet of plastic in the space between them as a barrier, maybe foil faced, and a blower to exhaust the motor heat out of the vehicle.
ReplyDeleteJRP3
Even just an aluminum sheet between the motor and the controller with an air gap will cut down nearly half the radiated heat reaching the controller. It sounds like the mounts would easily accommodate something like this quite easily.
ReplyDeleteGizmo: yes, a radiant barrier would be helpful here, especially if it also blocks the exhaust air from the motor's internal cooling fan.
ReplyDeleteJack: most people end up mounting the Soliton1 on top of the motor (or very near it) without incident, I just wanted to point out the need to be mindful of all the sources of waste heat in the engine compartment so you can ensure that, e.g., the exhaust air from the motor doesn't blow straight onto the controller's heat-exchanger or heatsink.
As for the need to derate at higher voltages, this is true - or should be, anyway - for all controllers. For example, the highly-regarded Zilla Z1K is limited to 885A max at 300V (veryu similar to our derating to 900A at 310V). This is an artifact of needing to stay with the "Safe Operating Area" of the 600V IGBTs used in both controllers (though they are very different IGBTs to be sure).
Controllers that don't derate allowed current input voltage goes up are either leaving a lot of extra power on the table at low voltages or destined for a short, unhappy life (like Francis Macomber) at high voltages.
At any rate, I admit there is but a vague mention of this on p4 of the manual and no real mention of it on the website, but modifying the latter has been an incredible thorn in our side since it is all Flash based and none of us have a clue about web design, etc., otherwise the content would be a lot more actively managed... Anyway, time to go back to making the donuts...
The Soliton is mounted on an aluminum sheet - that's what holds it in place and it should defer heat to some degree.
ReplyDeleteDerating is central to longevity and durability. We do a bit of it ourselves. We'll be at 191v at this point, and shouldn't strain it unduly. I doubt we'll ever approach 1000 amps except on the dyno. With the natural sag on these cells, I think we're looking at 170kw top end on a 300kw device.
That's more than a little headroom. In fourth gear, at about 160 mph, we MIGHT get there.
Jack RIckard
Jeffrey:
ReplyDeleteDoes the ethernet put out a serial stream of data including AH? We may want to do up some software to display this data. The documentation is a little spartan as to what can be ported out this way....
Jack Rickard
Yes, all sorts of information is continuously streamed to the ethernet port in UDP format. You can get a rough estimate of Ah consumed by doing your own integration of motor amps x duty cycle over time. However, I wouldn't use this for anything but a very rough estimate of Ah as the Hall effect current sensor inside the controller is not particularly accurate at low currents (specifically, it tends to read too high). I calibrate for accuracy at the high end of the current range because that is what is important to the semiconductors (and for delivering on our promise of 1kA).
ReplyDeleteThe reason there is little documentation on the logging function is because it is constantly getting tweaked. Indeed, we provide example code for a basic logging application, along with a description of the data format, in a zip file, but it has been rendered totally obsolete in the last couple of weeks. These changes will be in the next revision of the controller code (1.3) which, with any luck, will be done today.
Your operating voltage and mounting arrangement sound perfectly fine. Seb has much more experience with the actual implementation and deployment of the Soliton1 in the real world so if you want to learn from his mistakes... ;)
Jan, there was an error in your quote from the pdf you linked to. The sentence starting with, "For this reason..." should read, "For this reason, cells should not be stored fully charged at temperatures approaching 60°C. Optimally they should be stored at 25°C or less and between 30-50% state of charge."
ReplyDeleteDavid
Hi Gizmo, Yes, you're right. I copied it by typing. But the reason why I find it interesting are the next sentences.
ReplyDelete"The lower limit is chosen because they are often stored in packs witch circuitry that demands a small drain on the battery. When one considers the circuitry needed for li-Ion, it becomes the most important source of self-discharge.”
What would be the advised lower limit, if they are stored without circuitry?
Regards, Jan
Jan,
ReplyDelete"The lower limit is chosen because they are often stored in packs witch circuitry that demands a small drain on the battery. When one considers the circuitry needed for li-Ion, it becomes the most important source of self-discharge.”
That text answers itself. If you don't have circuitry across the battery terminals; then you will not have a problem.
Some not so small points Jeffery
ReplyDeleteNo way can a warP 9 ever constitute a 35Kwh heater in reasonable use. Its an electric motor. Not a heat engine! Its heat output at any given speed/load (eff%/100)*Input(Kwh).
Another error. "heatsinks is that they are just as good at picking up heat as shedding it."
Not so. Unless its a flat plate in a vacuum. Fins, shade and air passing through -- different ball game.
Solitons' heatsink faces upwards and I'm sure Jacks 350wh/mile consumption at 60% eff. (140watts heat) will be manageable.
I do worry about the speedsters Soliton mounting plate if made with aluminininunmm. If its in any way bouncy it will crack and break.
Don't worry about the mounting plate. It's 0.80 aluminum but the edges are turned down which makes it quite stiff. No bouncy. On reflection, we should have turned them UP as it makes reaching the mounting nuts underneath a little bit of a task.
ReplyDeleteThe Soliton is surprisingly heat sensitive. Just read through the manual and they start current limiting at 55C. We'll see what we wind up with but might need to go to a bigger heat sink/fan arrangement if it poses a problem.
We'll be pumping quite a bit more than the 4 lpm they call for - more like 20. Unfortunately, I'm hoping we're not over their 5 psi limit for the system.
Jack
Jeffrey:
ReplyDeleteThe user manual really doesn't document the logging function at all that I could find. Is there a document that describes this?
jack
This appeared in the latest EVA newsletter:
ReplyDeleteI wanted to comment on your article in the latest newsletter regarding an EV burning up from lack of a BMS. I had a fire that was caused by the BMS!
The fire in my LiFePO4 pack, was fortunately restricted to the BMS circuitry/wiring itself...mostly. In my case though, the fire was caused by the BMS. I will not name the manufacturer, and in fact believe the problem is inherent to any similar BMS regardless of manufacturer. I had been warned by Jack Rickard of EVTV.com that he had heard of at least 8 car fires started by cell BMS circuit boards and the asssociated spaghetti mess of 22ga wiring needed to connect the low voltage, overvoltage, and power circuits to 50 (or so) Li Ion batteries.
Shortly before leaving for China in October (to look for more Li Ion batteries), I took my newly completed E-sports car for a test drive with my mechanic/assistant builder in the passenger seat. Less than two miles from home, while accelerating to around 45 mph, he smelled something burning, and soon the cockpit was filling with choking white smoke. Since the weather strip for the side windows had not yet been replaced and we were concerned about scratching the glass on the old, hardened window seals, he partly opened his door as I searched for a driveway to pull off into. I pulled into a church parking lot and he jumped out asking if I had a fire extinguisher. I said, "yes, at home", so he started searching for some loose earth to through on what was now the foot high flames in the trunk battery box. By now it had burned through the polycarbonate lid/trunk bottom. He could find no soil, and poured a handful of (fortunately) wet mulch and sand he had gathered from under a tree onto the fire. The flame went out and only a small trickle of smoke continued. Since we were now only about a mile from home, I decided to try to drive home where we had tools and fire extinguishing equipment.
By the time we pulled up to the garage, the flames had reignited, and after finding our fire extinguisher had been discharged and was not working, dumped a box of baking soda on the flames which put the fire out out for good. After cleaning out the mulch and baking soda and examining the damage, I was relieved to see that, amazingly, though a couple of cells had blackened/scorched tops, the fire had been restricted to the BMS circuit boards and small gauge wiring associated with them. The batteries appear still functional, and the only thing damaged other than the BMS circuitry was the polycarbonate battery box lid, which doubles as the trunk bottom.
Whether a BMS board became detached from a cell top (possible, since they are held there by adhesive backed clips, which don't like the cell vent "hump" on the top of Thundersky cells) and shorted against a 4/0 power cable, or whether the short was internal to a BMS board is hard to say. What I can say is that having hundreds of yards of small gauge spaghetti wiring attached to dozens (50 in my case) of batteries that can put out hundreds of amps (or 1200 amps in my case) is a recipe for fire.
When I get back home, I intend to remove the BMS and all the spaghetti mess, bottom balance the cells, buy a programmable charger to keep track of the charge and a good battery monitor to keep track of the kilowatt hours discharged during driving, and have a much simpler and safer EV.
....and just in case, carry a newly charged fire extinguisher or box of baking soda in the car. Once a fire is going emergency disconnects don't do much it seems.
Sincerely,
Creighton Freeman
Columbus, OH
AndyJ: The efficiency of a WarP-9 motor plummets as current exceeds 1000A. We've measured 65% efficiency at 1000A on our dyno. John Wayland has reported, IIRC, 55% efficiency at 2000A. This shouldn't be surprising given that the 1 hour (S2-60) rating of the motor is around 225A. Finally, I don't mind a good old-fashioned argument - unless it's over perpetual motion - but I do resent being lectured to just like anyone else, especially when I am trying to be helpful to a customer outside of the "proper channels", so please keep the snark to a minimum or I'll simply put you on ignore.
ReplyDeleteJack: Yes, we begin derating at a heatsink temp of 55C. Please note that other honest controller manufacturers like Cafe Electric do the same thing. This is because an experience power electronics engineer knows that if the heatsink is at 55C then the semiconductor junctions that are generating the heat will likely be at 80C, and that is when derating typically begins for them.
The documentation for logger is usually available from our website in the downloads section, but as it is in the process of being revamped you might want to shoot me an e-mail for the latest zip file (just waiting on our programmer, Martin, to finish writing the release notes).
I found that there was very little improvement in heat transfer above 1 gpm of flow, but you can certainly exceed it if you like. Likewise, we test the cooling loop for leaks by pressurizing it to 100 psi while the enclosure is submerged under water, so 5 psi is more than a tad conservative - the reason for such a low pressure recommendation is mainly to steer people towards the right kind of pump (ie - centrifugal, rather than some form of positive displacement pump like, e.g., a Shur-Flo).
Jack! This guy by your own incredible standards has to be a moron! He put out a fire and then continued to drive home, only to have it re-ignite again?? Are you kidding?? The BMS board 'could' have shorted to a 4/0 cable or 'could' have come loose??? Seriously? Are we not talking workmanship standards here??? A BMS is only as good as the nut installing it! I can't wait to see picks of this guy's pack. It has to be a work of art!! Please, the same idiot that can't assemble the hardware, also diagnoses that the BMS was to blame!??? Jack, this guy has to be one of your kindred spirits!! Are you sure he's not a cousin?? He sounds just like you! Or maybe he's taking lessons from you!
ReplyDeleteMikep: Actually he sounds like he was taking lessons from you -- remember you'll ruin your pack in short order without a BMS!
ReplyDeleteJeffrey:
ReplyDeleteCurtis is pretty experienced. We've actually experimented with the trip point on their current limit, which is not quite as smooth as advertised, but is right at 80C on the heat sink.
Rinehardt is also at about 85C claimed, though I haven't actually tripped it like the Curtis. We're running 55C INTO the Rinehart successfully - kind of combining our electric water heat system for the car with the COOLING system for the Rinehart and at least in winter, it is working remarkably well. We get some heat from the Rinehart, and heat the water to about 65C in cold weather. Coming out of the heat exchanger in the car, it is more like 55C into the controller.
The reason this is of interest is that we plan a similar system on the Escalade, only larger. It will feature TWO Solitons and TWO Warp 11's. It's also going to feature a rather massive water heating system on the order of 15kw. But it's also going to be routed through a 50x50 inch battery box in the back of the truck to warm the batteries. This should be a pretty interesting heat sink by itself before routing the water to the Solitons. We're going to line the entire bottom of the box with PEX tubing much as you would warm a house floor. An aluminum false bottom will then lie on top of that.
The 1000 amp scenario is hopefully mercifully brief on both Speedster and Escalade. The Speedster will be at 191 volts nominal. If we sag to 170, that's 170 kw at 2400 lbs. I don't know about NEDRA, but that's staring to calculate out at felony ticket range around here in six or seven seconds. We'll see.
Similarly in the Escalade, it's going to be slightly lower voltage 170 or so, but potentially 2000 amps with two Soliton's and two 11 inch motors. We're planning on a 400AH pack using two strings of 200AH CALB cells, though those cells aren't precisely AVAILABLE yet. But I can't picture more than six seconds of such activity.
As to pumps, we're using the pump from the Prius Inverter. It's really quite quiet and will do 20 lpm in a free run - much less under pressure obviously. But I don't know what the pressure is. If you're testing to 100 psi, I find 5 psi a little beyond "conservative." It would have helped to know "tested to 100 psi" if that's what it is.
Jack RIckard
Jeffrey:
ReplyDeleteI can't shoot you an e-mail because I don't have an e-mail. Your "blogger" id is closed.
Jack
Curtis is experienced all right... they've been making the same dumb-as-a-box-of-rocks controller for the last, what?, 30 years or so? I, personally, don't think it is acceptable for a heatsink to be at 80C before derating kicks in just for burn-hazard reasons, much less the unavoidable penalty to semiconductor (and capacitor) life courtesy of Mr. Arrhenius.
ReplyDeleteThe Rinehart Motion product looks nice and their cooling recommendations seem eminently reasonable. After all, 4 lpm of water will only rise 5C while absorbing 1350W.
Your plans for the Escalade sound distinctly unconventional but I don't see any big red flags. Once again, when it comes to the actual implementation of the Soliton1, Seb has lots more practical experience than me - I'm just the power electronics engineer here.
You can reach either of us via info@_remove_evnetics.com which circulates to everyone pertinent at Evnetics. I'm certainly not going to put my direct e-mail address out here in public... I get enough dumb physics-defying questions and male enhancement spam via e-mail as it is now.
To the armchair engineer AndyJ:
ReplyDelete"Its an electric motor. Not a heat engine!"
Any device aimed at converting energy from one form to another produces heat. The heat is caused by something called losses, or inefficiency, due to imperfections in the process. Worst case in the efficiency department is the ordinary incandescent light bulb which is actually nothing but a radiator that also happens to produce light as a byproduct.
An electrical motor is a gadget that converts electrical energy to mechanical energy at a peak efficiency of roughly 90%. In an EV this is seldom achieved because a flaw in the construction that's usually referred to as the driver or, in more rare occations, a traffic jam. When the motor runs outside of its optimal torque/rpm range the efficiency plummets and can even reach as low as 0% if the motor is stalled. That usually happens when the driver can also be referred to as a moron. There have been several scientific studies of how common the average moron is in traffic, but my suspicion is that those studies are too conservative and that the incidence of morons is far more common than what's currently believed by scientists.
"Unless its a flat plate in a vacuum. Fins, shade and air passing through -- different ball game."
If we, for a second, ignore the fact that no air can pass through vacuum, it has been known for several centuries that hot air passing over a cooler surface will heat the surface quite like how cold air passing over a hotter surface will cool it down.
If you don't believe me, kindly visit your favourite hair dresser and ask politely to be exposed to a device called a hair dryer. For added pedagogical effect, please moisten the area of your body that will be exposed to said device before the device in question is applied.
"I'm sure Jacks 350wh/mile consumption at 60% eff. (140watts heat) will be manageable."
If Jack will be content with driving at the not quite thrilling speed of mere 1 mph your calculation is accurate, however I doubt that that's the case. The original post indicates that Jack's wish for speed exceeds this moderate, although admittedly quite safe, sub-stroll speed and thus the heat radiation will exceed your expected numbers.
Anonymous:
ReplyDeleteA bit snidely worded, but yes. I guess my greatest surprise in doing electric cars is that motion is easy, batteries aren't as hard as they are made to be, but thermal issues are INCESSANT. And they are in all directions.
The driver is a necessary part of it, no matter how "moronic" I might be. My concern with range decreases in direct proportion to any deviation from 74F. Either hotter or colder. And truly hot (95F+) or truly cold (25F and lower) and I forget ALL ABOUT range. I want to talk about heat, or air conditioning.
Our motors do put out heat. They get rid of it "pretty well" too. But after our chill plate experiment with the Curtis 1238, which contrary to Jeffrey's competitive pronouncement, is quite different and quite advanced from the usual Curtis 1231, I have become a big fan of liquid cooling. Water has a heat index about 3600 times greater than air.
You can't destroy heat. But you can move it. And you can spread it out. It shows up as a problem in concentration. Glycol/water is just an excellent medium to carry it and both aluminum and copper are really very good as well.
George Hamstra and I have had several discussions as to how to get a Netgain into a liquid cooling scenario, and I just don't have any good answers, and he himself is seriously seeking them.
Actually, there are no real limitations on power handling in electric motors - they can convert infinite amounts of electrical energy to infinite amounts of radial torque - up to their heat limitations. They are simply limited in how quickly they can rid themselves of heat and that IS the limit on what you can put into them. In practice, their "efficiency" is all over the place millisecond to millisecond. We can calculate average efficiency over specific operational scenarios. But it is ALL about heat dissipation.
So the thermals for the driver, the occupants, (my wife isn't concerned about range at 25F either), the semiconductors, and most especially the motors, essentially trump all other considerations - now that we have sufficient battery energy density to have such conversations.
I would predict the future of electric cars will revolve around materials science, both in the body/frame and batteries, and thermal management. No breakthroughs in motors or control electronics are likely or even particularly helpful. It will all be about weight and heat.
I was interested to learn that polycarbonate has finally been approved in automobiles even for windshields.
Jack Rickard
Jeffrey:
ReplyDeleteMy e-mail is rather publicly posted. Of course, I'm not a big time controller designer such as yourself and not nearly as important so I don't have the same issues I suppose..... It must be a terrible imposition.. I can just imagine....
Jack Rickard
Jeffrey.
ReplyDeleteTell me, did I say using 1000amps at 65% efficiency? No, I stated an average real life consumption rate, (350wh/mile); the favourite format. Which is also more useful working out your heat dissipation rate.
If Jacks car uses 5Kwh@60mph and he has a good rev fit he can be on the 90% efficiency area. That's 10% heat on the motor.
Throw a hissy fit if you wish Jeffrey. You are talking of a series wound motor throwing out 100% power as heat while returning an output power of 45%. That's 145% from 100% -- Bought any JP Morgan Silver yet?
I'm fully aware series wound motors are most efficient from static and return less efficiency at speed. (Inverse to the cooling fan). They are NOT a good bet for racing. They produce most of the heat IN the rotor assembly. If you want a heatsink for a massive continuous current consumption? Good for you. Go AC with its external cooling jacket.
Armchair engineer? This may be not be my ball game but I am a fully accredited aerospace engineer with 35 real years experience.
Errata:
ReplyDelete"Throw a hissy fit if you wish Jeffrey. You are talking of a series wound motor throwing out 100% power as heat while returning an output power of 65%. That's 165% from 100% -- Bought any JP Morgan Silver yet?
To the armchair engineer AndyJ:
ReplyDeleteNo matter how you turn it, "Anonymous" is not spelled "Jeffrey". Even if you believe that Jeffrey doesn't have the guts to take you to task with his real name attached to his post, I am quite sure that Jack can verify that our IP addresses don't match.
"They are NOT a good bet for racing."
Should I tell John Wayland that or do you want to do it yourself?
Sure I'm an armchair engineer when it comes to EV's. Not interested in spending myself wise.
ReplyDeleteThis as my classroom. There are some cheeky kids disrupting lessons talking about perpetual motion machines who would rather attack the person than listen to the math. >:-)
Yes, John Waylands car is quite fast. Love its take-off. I wonder if he has noted those UQM motor graphs yet?
My very simple point is efficiency. Under load, batteries get hot, so lose voltage and capacity in creating heat. Then there is the efficiency differences on rpm/load between different types of AC and DC motors. Please note Netgains performance graphs. The rpm runs backwards.
John Wayland is not doing distance at speed.
Its horses for courses really. For MY real world use, Its looking like a Netgain through an automatic gearbox from a diesel car.
If I had the need for constant highway use, it would be a water cooled AC motor. They give higher power and greater efficiency at speed.
Don't believe me? Go see for yourself.
p.s. Jacks car should be ok for airflow cooling in the real world. Providing the airflow under the car will not work against the motors internal fan.
Andyj:
ReplyDeleteThere is a slight efficiency gain in induction or permanent magnet brushless motors in general. It is not sufficient to really make a difference you would notice in a car.
For racing, or pure power, there is no physical way to get more torque out of iron motors than a series DC motor. Whatever power you put in shows up in both fields to the same degree. I cannot think how you could improve on that for torque. It just cannot be otherwise.
Currently, AC systems start at about three times the price of DC series motors. There is one exception, and that is the HPEVS we used with the Curtis 1238. It is a marvelous system, but can barely reach 50kw as it stands. For small cars, it is very economical.
THe main advantages of polyphase brushless designs can pretty much be summed up in three areas.
1. Higher voltage. Without a commutator and brushes, the motor can be powered with higher voltages quite easily, and indeed there are systems operating as high as 900v in cars. For a given power output, this means correspondingly less current. The heat is caused entirely by the current part of the equation. And for this reason, you can use much smaller cables and conductors.
2. Lower maintenance. Without commutator or brushes, you're pretty much down to bearings as failure components or simply burning the insulation off the windings.
3. Regenerative braking. Very easy with these motors, essentially a loser with series DC. We did some testing with regenerative braking and were very disappointed in the results. The 10=15% efficiency gains are just pure myth. More like 3 to 5%. And for some driving styles, negative numbers.
The most attractive feature of AC motors can hardly be termed an advantage actually. THey are much easier to liquid cool and I like liquid cooling. But as a design goal in and of itself, it makes no sense.
For these reasons, we continue to be interested in polyphase brushless designs, but frankly they don't pass the value test in an electric car. The simpler construction of the motor should kind of offset the slightly more complicated costs of the control electronics. And I would find a 120% pricing of DC series systems about appropriate. 3X and 5X and 15X is what is out there generally. So aside from HPEVS, we don't see much useful out there.
Our adventures with TIMS600 MES-DEA actually point up the difficulties in long supply chains and tiny customer base logistics. We're going to try to avoid those mistakes in the future.
Jack RIckard
No arguments there jack :) Have you compared consumption at speeds between Uno and deux?
ReplyDeleteOn complexity and outlay. My real world preference is DC. They say it takes 6hp to do 60mph but that barely shows on the efficiency graphs. I'm a natural born free wheeler anyway. Power should be mainly for hills and acceleration. Do you agree?
I love the idea of using juice only when and where its needed. Something your ordinary car does not do.
Off subject:
One small thing. With your new gearbox on redux and no clutch. The revs between the lower gears is greater, what will the wind down time to get into the next gear be like? You can visualise this by looking at the ratios against rpm on paper.
No need to look. I still have a clutch. But that gets into the areas of ADVANTAGE of the series DC. I can shift without using the clutch, and I can freeroll. Actually this is a delight to me and I prefer to drive this way. Nothing like cresting a hill and ACCELERATING down a long mile long hill using nothing, hitting the light just right, and rolling halfway up the next hill with nothing on but the radio.
ReplyDeleteYou don't GET to feel this with an ICE car or with regen. It just rolls and rolls and rolls. This is probably why I get negative numbers on regen.
It is a delightful feeling of freedom and silence.
Jack Rickard
I'm sorry, thought you didn't have a clutch.
ReplyDeleteFreeroll/freewheeling.
Point taken well and truly. The same message is mirrored by so many people, even with fuel cars.
I would like regen for one thing only here in the overcrowded UK with too many radar/laser traps and steep, winding roads.. Not allowing acceleration once the throttle is released. Switchable, of course! The uk actually has an old law banning free rolling but its almost impossible to implement. Unless some sharp witted policeman takes note of a series production car with no regen. The driving licence testers instruct us all we'll lose control of the car if its out of gear. Which is a nonsense in all cases this day and age.
Point of interest:
On noise levels I find some vehicles can be more appreciated in free roll more than others. My Nissan, no. It's all (cheap) tyre noise enclosed in a roof. My BMW motorbike, yes. Choosing the right vehicle seems very important.
Hope I've given you some food for thought. Waiting in anticipation for your next video. In mind to make a contribution one of these days...
Here in Ireland we have summer tires, and winter tires. I have never been able to lay my hands all weather tires.
ReplyDeleteI experimented with a petrol car. I drove very conservatively without dropping the clutch, and very conservatively while coasting where I could. There was a slight improvement on consumption.
I was always thought that coasting was dangerous because it increased the chances of a wheel, or wheels locking. I think this is only a worry in adverse road conditions. During the snow here in Ireland I needed a little brake to slow the car when going downhill (steep ones) even when in a low gear. I daren’t coast down a steep hill in snow and use the brake alone to ensure I do not pick up excessive speed; it is more difficult to brake without braking excessively and perhaps skidding.
Is this a concern for those of you who have no choice but to brake while coasting in wet weather etc? I know the correct tire choice would make a difference.
Or is this simply not of concern?
I’m very interested to know, and by the way I’m for coasting. I’m worried though does the braking system for an ICE need reengineering (for lack of a better word) for DC run EVs?
Padraic McDonnell
Patrick:
ReplyDeleteYou know, it doesn't come up much. We're just not real enthusiastic about driving a Porsche convertible in the snow.
I can tell you that the regen on the accelerator on the Mini I thought would pose a danger on ice or snow as only the front wheels have regen of course and I thought it would cause a real problem. So I have a cutoff switch for the regen.
In practice, the regen on the accelerator seems to be quite an advantage on light snow.
Jack Rickard
There is so much mythology around driving on water, snow and ice that its laughable.
ReplyDeleteThe trick to maintaining control is moderating inputs of all kinds- throttle, brakes and steering- so as not to break the tire contact patches away from the wet or icy surface. ABS does that with fine, automated control. As a driver, you want that, too, but it doesn't really matter whether it comes from regen resistance or braking force. If you have regen working on the rear wheels, it could be ideal, because now you can balance the braking energy away from the steering energy, and presumably have a better range of control over the four wheels in extremely marginal traction conditions.
Of course, all of this is theoretical crapola. It comes down to how well you can modulate the controls and the keep the contact patches adherent, which is so highly variable depending on conditions, tires, chassis and driving skill that no generalization like "regen yes" is of much value.
The bottom line for a road EV is you want to be able to turn regen on and off, and when its on, you probably want some pretty fine control over it- like a pot would provide. Mostly, though, you want to play with it, because you won't really be able to predict how it behaves.
We were all actually instructed to do this many years ago by wise upper midwestern parents: "The next time it snows a few inches, you guys go take your cars down to the mall after it closes and before they plow the lot, and learn how to drive in it. Keep it slow and be careful, but have fun."
Almost 35 years later, I still do a version of that exercise with every car I acquire. Its still fun, too...
YMMV,
TomA
Jack.
ReplyDeleteI can’t imagine you in your convertible driving around in the snow. That image made me laugh so much.
I used the snow as an example, but I’m also curious not so adverse conditions such as a wet road. Have I cause for concern where I do not have the option of regenerative braking? This question is for anyone.
I agree that the ability adjust regen is important.
Tom. The crabolgy surrounding snow driving is laughable.
My advice. Put winter tires on your car even if its four well drive.
Low gear going downhill.
Imagine that there is a baby’s head between your foot and the pedals. i.e. gently does it.
If you loose traction going uphill, decelerate, do not accelerate. Etc, etc, etc, I could go on, but there’s no point
Padraic McDonnell
Getting a bit off subject but is salient for battery positioning. I would be wary racing Redux.
ReplyDeleteApplying an engine for braking or acceleration on the bends DESTABILISES a well balanced car.
I had a car few people outside the UK have seen. It was a very small hatchback called a Rover Metro Gti. This little car had tyres like a road roller, no weight and over 100bhp. It was a car thieves best pick for a fun ride.
My favourite trick was in the wet on a tight corner. Go in fast, let off the throttle so the car ran wide. Cry out to the passenger(s) "Oh no!". They braced themselves for impact them I applied some juice eased the steering slightly. It gripped again and drove round. "haha. Got ya!". The car was useless in the snow. Yet the dry grip went beyond its handling qualities to the point where you can actually feel body flex.
ABS and traction control work on limiting the wheel spin relative to the road speed to maximise grip. This business about free rolling being dangerous is as daft this day and age as requiring a man to walk in front of you waving a red flag.
Andyj - I don't know how I can make this any clearer or plainer: I used a Soliton1 to deliver 100kW of electrical input power to a WarP-9 and my hydraulic dyno reported 65kW of mechanical output power. That's 35kW of loss, or 65% efficiency. Yes you can turn a WarP-9 into a 35kW space heater if you run it at 1000A, and yes there are bad consequences if you do that for too long (like more than 15 seconds, I'd say). This is, after all, a 4-5x overload of the 1 hour current rating so it definitely falls into the "motor abuse" category.
ReplyDeleteMy points earlier were simply that if you use external blowers to force air through the motor you can run it at high enough power levels for long enough to heat up the entire engine compartment, and if the Soliton1 is mounted on top of the motor, such that it gets blasted by the hot exhaust from the motor, it might go into thermal derating prematurely, despite being liquid cooled. It's not a matter of conjecture or theory, it's something I've personally witnessed on a Volvo V70 crammed to the gills with some 38kWh of LFP cells that weighed in around 4700# and still went up and down the local interstate highways here at what Jack likes to colorfully refer to as "felony ticket speeds". Including the 4% grade of the Sunshine Skyway bridge, btw.
That car melted three WarP-9 motors and could heat every surface in the engine compartment (as reported by an IR thermometer) to 75C/165F after a 20 minute run. Probably a larger motor should have been used, or even two of them, but the automatic transmission needed to see at least 6000 rpm to shift and there wasn't any room for two motors, even if they were smaller in diameter (e.g. 8").
Such are the vagaries of the real world.
And I am not anonymous, btw. I don't have a "blog id" or whatever, but I am most assuredly Jeffrey Jenkins from Evnetics.
Jeffrey:
ReplyDeleteAs I said, I can't make heat go away, only move it around. We are looking at a scoop design to help the Warp9 along with air when in motion. But yes, if it heats up the entire engine compartment, we have a problem Houston.
On the other hand, all may be right with the world. With the Soliton1 over the motor, if it gets too hot, the Soliton1 goes into current limit, I kind of have a closed system here that regulates itself. Not entirely a bad thing.
We have our DC-DC convertter installed, a live pack, and so last night I got out the horrid Windoze laptop to tackle the Soliton1 software procedure.
I've been so beat up with Rinehart Systems and worse, Current Ways, really ugly bad software that I dread these particular sessions. Couldn't we just bang on aluminum with a hammer for awhile guys while I hold it in my teeth?
To my surprise, and as the exception proving the rule, the Soliton software was EXTREMELY well designed and well implemented. Whoever Q is, he's done well.
The system uses an IP connection scheme that you all may not be familiar with, but I am. It's normally reserved for mesh networks and I had about six Linksys routers running some hacked firmware a guy did to make mesh networks. It used the same 169.254.xxx.xxx IP scheme.
The firmware update comes COMBINED with it's own "loader". In other words, you don't have to load a program, find the latest file, load it, and then "bootload" it. You just run the loader.exe. It launches, finds the IP number, connects to the controller, updates it, and then advises you it has completed. That's it. You do nothing but launch the program.
Next update: you just get a new loader.exe file with a different title. ALL firmware updates should be done this way, worldwide. And anyone doing it differently is publicly proclaiming themselves a moron hereinafter.
We then went to the web interface. Calibrating the accelerator pedal is the central issue in controllers. This process required two people and took about 12 seconds. Click on min throttle. Press pedal to floor, click on max throttle.
Ubelievable.
There area actually a lot of input options on the three spare inputs and a lot of output options on the three spare outputs. The documentation is a little vague on what is expected on the inputs and what is provided on the outputs. This needs some work. It looks very good. You get a little pick list on each of these. I just don't know precisely what to do with it.
And then there is logger.exe. After you have it running, you just launch logger.exe. It prints out a single line with current, rpm and temperature. Again, that's all you do.
We were not prepared for all this. We weren't up on jack stands. We weren't ready to test it. I assumed I would be wrestling with it for an hour or so.
We put it in neutral and played with the pedal. A little touchy. Changed the pedal profiling to 40 which indicates 40% power at the 50% throttle mark. Set to a very tame 300 amps per second slew rate. Still a little more responsive than we are accustomed to, but let's try it.
Brian rolled off the lift and drove away into the night. We've never seen him since. Some say he's headed for Mexico with the car....
Jack Rickard
Some say he is really the Stig....
ReplyDeleteSome say he was a masked man. I don't know, but I sure wanted to thank him.
ReplyDeleteHi ho Silver...... and away...
Jeffrey, I have every respect for the engineering of your product and come the day; yours is on my shopping list.
ReplyDeleteSorry I got steamed up before. The red glaze. I did not read the sender was "anonymous". He did not even consider doing the maths I set before him, nor make a sensible comparison. Every drag car is run beyond its design limits.
Your software guru seems to be well up to par. People have said the power delivery is nice and "oily". You are careful with IGBT life and the overall design, where most seem to lack. Qudos.
But surely, your note of 15-35KwH heat from the motor will not be met continuously. If Jacks motor rotor was so hot he would not dare simply stop the motor without keeping the motor revving under no load until it cooled sufficiently! At 1000A input its got to be better to fit a motor that runs efficiently at this power level.
Comparing Jacks car with say an electric drag car is taking dreamland a bit far. Some quick reasons why:
Battery type, they are boxed, alongside the motor, the overall weight, gearing, tyres. Makes for a car with road only pep and distance.
Has Brian come back yet? I bet he's been to the wineries without you my son :)
The Soliton fitted where it is will ensure "daughter mode" if it gets too warm down below. Think of the batteries!
Suggest? 6mm double foiled insulation for home DIY stuck onto hardboard and on inner sides of battery cases?
"Airtec Double Insulation 1.5m x 25m
25m. For refurbishment or new build. Suitable for dry lining and cavity walls. Meets current building regulations when installed in accordance with the manufacturers' instructions.
Thermal Resistance up to 1.455m²/kW".
Sounds like it belongs in your car anyway :)
Jack said, "so last night I got out the horrid Windoze laptop to tackle the Soliton1 software procedure."
ReplyDeleteFWIW, we now include Linux versions of the updater and logger programs (while Martin is an "enthusiastic" proponent of Linux, he hadn't actually made Linux versions available until v1.3 of the code). Porting to the Mac is on our collective to-do list along with a re-write of the long-neglected manual.
Oh, and thanks for the kind words. Seb was sure you would be pleasantly surprised at how uncomplicated it is to get a Soliton1 up and running (I confess I didn't think you'd care one way or other, so, I guess I was wrong on this count). The Soliton1 owes a lot of it's functional design to Seb's frustration with other controllers. We pretty much looked at what everyone else was/is doing and did the exact opposite... ;)
Andyj - none of my data comes from drag racing and at no point was I saying that Jack will have to drive the Speedster like a drag racer to cook his motor. However, it is definitely true that he will have much less chance of doing so than in the Escalade. Two motors and two controllers will produce a prodigious quantity of waste heat. A lot more attention must be paid to its removal and that was the forewarning I was trying to give here.
Yes, the WarP-9 can take a fantastic amount of abuse, but there is simply no contest between it and a Soliton1 - we'll beat that motor like a rented mule every time.
Jeffrey:
ReplyDeleteThe horrors visited upon me by simple configuration software SO badly done you would not believe and I find it infuriating every time I encounter it. True, you configure and move on and its not an everyday occurrence. But it is so needless. Configuration software doesn't do that much and shouldn't be so UNIVERSALLY BAD in any event.
We shot the show Friday, and it was a workout. Your controller is sufficiently simple that we were able to go into some detail, without it being a hopeless video mess. And so we took the opportunity to describe the motors and how controllers work conceptuallyu, and then went through each connection to the controller and most of the options on the software as well.
The controllers is working very well, We are only up to 600A per second slew rate and it is already very responsive. We're a little concerned with brush seating at the moment, not heat in this 35F ambient. But your points on heat and power with the Soliton on the Netgain are very well taken and received here. Of course, it's kind of a privilege to HAVE a controller I could blow up a Netgain with. And if we blow one, we'll get another and revise our approach.
The show has finished rendering and is uploading now, and should be available this afternoon. It's a bit of a problem in that its over 4GB and runs two and a half hours. We've gone beyond feature length film all the way into a miniseries format essentially. But I hear from a lot of people that don't seem to mind length oddly enough. And I think it's one of our better shows.
I would like to suggest a feature. We have a very nice liquid cooling system, with a very easy expansion to a larger heat exchanger if necessary. But we don't like to run it when the temperature is 20 degrees out and we're sitting at a stoplight anyway. So I have a little 45C snapswitch kind of JBWelded onto your Soliton case that switches ground to the pump.
Given that you have 12v PWM outputs available, and do report the temperature, it would seem to me you could improve on this situation rather easily. I don't know how much power you can put out for this, but a PWM to the pump to control temperature would be pretty damn fancy - basically turn the pump at whatever speed is necessary to cool the controller up to 95% or whatever of 12.8v.
Lacking that current ability on the output, a 12v relay signal would be less elegant, but probably every bit as effective. We just don't want to run the pump needlessly.
Yes, I'm aware of your trepidation through this project. I'm really not that hard to get along with, I just want UTTER PERFECTION in everything and ALL DETAILS MATTER. And after that I'm pretty easy really.
You win. The Soliton1 is the best EV controller in the world - period. The packaging, the hookup, the configuration, and the performance. You've got it all going on and I have NO problem acknowledging it - with the usual codicil of a never ending series of additional feature requests, which I and every one else will only provide when you're good.
It's a GREAT device. I love it. You need to prepare to make more of them. I think they'll be in short supply by midyear.
Do me a programmable charger - no BMS thanks. I'll write you a feature list. Use S1-S2 from the motor as your buck/boost coil and let me bypass the rectifier to provide high current DC directly to the buck/boost circuit, so the charger will work on ANY DC input up to 400 amps - as well as 120 and 240vac. You may call it the Oscillon1. Probably patentable. You're welcome.
It this way, in addition to the ability to quick charge, we could also use one electric car to "jump start" another - even if they are at different pack voltages.
It would also be VERY nice to have an inverter output so we could convert pack voltage into 60HZ AC true sine grid phased or backup power - all in a 4 inch by 4 inch package of course.
Jack RIckard
Jack, Are you thinking of getting a 400A panel installed at 601 Morgan Oak, or all ready have it? We're getting a panel installed where I work.
ReplyDeleteThe thought of a full wave rectifier on the mains feeding the right number of batteries has not escaped consideration because the money saved and volume reclaimed would be substantial.
ReplyDeleteA Soliton that handles battery recharging, its own cooling and mains output. Now that would place this device well above anything else on the EV market. Could some of this be done using existing major components within or a bolt on?
An inverter for *load balancing* can be profitable on one hand or great for keeping the Grandkids "ps3 quiet" in the back of a Escalade.
**http://en.wikipedia.org/wiki/Load_balancing_(electrical_power)
Better still, a free recharge point so you can take home all the free electric you want. w00t!