Monday, July 11, 2011

Batteries and Things

After a week off, we did get a video out this week. It's a little hosed up. I have the new Final Cut Pro X editing software and it has a few problems I'm struggling with.

We had some news on the J1772 front. Clipper Creek actually made some sort of internal error and SHIPPED us one of their charge stations in exchange for money - about $3000. This is NOT a 70 amp charger as I said in the video, rather a 40 amp model CS-40

It uses the white Yazaki connector. It's pretty sturdy. And it's $3000. This will give us a baseline for J1772 charging.

For most of our viewers, somewhat more important is being able to fit your build so that you can charge using J1772 - and perhaps at a somewhat lower cost.

We have published the basics in the past with a couple of resistors, a diode, and a switch with a J1772 receptacle. David Kerzel is the President of the EV group down in Florida and a hobby machinist. He's done a lot of small engines and so forth. But today, he has ModularEVPower, a web site offering J1772 components for home builders and he has come quite a ways very recently.

We received one of his machined billet aluminum J1772 charge ports at $160. We can get these ports in a plastic model for $75 from China. But I am a sucker for machined aluminum pieces on our cars and this is very gorgeous.

Second, he has brought the resistors, diode, switch gig to a new level with what he calls an Active Vehicle Side Control Board at $30 each. This actually uses the proximity switch in the plug to activate the charging process. Once the switch is closed, it does the voltage divide down to 6v on the square wave to start the EVSE sending 240vac power. It ALSO closes a relay that has the common, the normally closed, and the normally open contacts all available for your use. You can use this to interlock a charger, light a light, or whatever. Very nifty and I think quite cool.

Finally, for $230 he has the guts of a the J1772 Electric Vehicle Service Equipment or EVSE - the EVSE Control Module.

This is what you connect to the plug end proximity pin (5) and pilot signal pin (4) to perform the J1772 interlock sequence. It features a switch allowing you to set the output current level which is encoded properly in the 12v 1 kHz square wave. It also has a 12v output to energize the contactors. You'll need a 12vdc power supply and contactors to switch the 240 vac phases, and you have, for less than $400, the equivalent to the Clipper Creek EVSE at $3,000. Of course you also need the plug and cable, which is about $175, and an enclosure. A GFI circuit breaker in the panel would be a good thing. But certainly for about $700 plus enclosure you can build your own. We use replica gas pumps. You can use a pumpkin if you like.

Duane Ball and Scott Smith engage us with a video report on the Porsche GTS (904) build they have been working on. This is the latest Chuck Beck replica of a 1964 Porsche race car that slightly over 100 were ever made. Special Editions has been putting them out and sold OUT of their first set of these. They have increased the price of this roller now to $63,000 to meet demand - that's with no engine. Duane has essentially completed his build to the point of first drive/EV grin and is working off some DC-DC converter issues and a problem with his tachometer sensor. Waiting for a VIN number in New York. But the car and build are gorgeous.

We also did a bit of battery testing this week. I have moved my test bench from the home garage to the new facility where I have LOTS of room and now quite a bit of electrical power. And so we did a bit of cell testing.

I obtained 16 of A123's 20 Ah M1 HD cells. This is a 20Ah pouch "prismatic" cell that is definitely a Lithium Iron Phosphate variant, which we like and use. The spec sheet has it at 3.65 and 2.0 v. We found it operates just as our Thunderskies with respect to voltage, charging and discharging although the climb and dropoff are a bit alarming at the end. That may just be our inexperience with such small 20Ah cells.

In any event, their claim is POWER and they do indeed deliver it. We were seeing sag voltages of less than HALF what we get from the Chinese variants for the same current load. They are very stiff. In American fashion, they don't quite put out 20Ah as advertised. We like the Chinese style claim where they call things 20AH and provide 21 or 22. Not on these.

Of course, A123 has been unbelievalbly snotty about the whole thing. We got these from OSN Power, a Chinese company ironically, They sample at $50 per cell, or $46 in quantity.

Two problems here. One is that they are heroically expensive at $2.25 per amp hour. Chinese cells are going to $1 Ah very quickly now. And secondly, they would have to be mounted in modules of some sort, which incurs some work, some expense, and some weight.

One of the advantages of these cells is a 127 wH per kg energy density, along with the power. Any weight from the module build would detract from that - and any expense would exacerbate teh first disadvantage.

If we can figure a super cheap super light way to do all that, we may pursue a build with these. But it would be to get titillating video. I don't see them as an alternative frankly. They are just too expensive and too complicated to use. They might get you a few more miles down the road, but they simply cannot compete with our lego block cells we use now. If you want to build a racecar, perhaps. But a viable vehicle? I don't personally think they warrant either the price or the complexity. They are incrementally better at some things with an exponentially higher cost to deploy. We have followed this company for years, and they remain a LOSER in battery development. The only company they've really sold on the cells are very small outfits and Fisker and they did Fisker by giving them the batteries essentially - a $30 milion investment in Fisker if they would use their cells.

John Plasma Boy Wayland and Rich Rudman have done something similar with Dow Kokam Cells and for the same reason - power output, once John's "Dream Came True" and they gave him the cells gratis. Kokam also designed a module for them which they had most of built at a custom plastics CNC plant (again gratis) with hardware done by Jim Husted. The result is about 10 lbs of module for 16 of these cells, pretty much negating the density improvement and is just hysterical. Forgive me, I just had to laugh. This is so overengineered and overbuilt to no known purpose that I can't deal with it. I SAY that overkill is ALWAYS appropriate. In this case, they have taken this concept to extremis and beyond. You HAVE to see this

If we have to do all that to use Kokam or A123 cells, it likely won't happen. But I have some ideas for something lighter and less costly ( cost is no object when using other people's money). I don't know when we'll get to them.


  1. I do not know why $3000 cost, where Nissan charging station under $2000 installed ...

    Sorry Jack, I bit rip off here...

    My 2c..

  2. Jack: in the video you mentioned trying out the A123 cylindrical cells and found them wanting. Can you go into a bit more detail? I can guess, but I'd like to know.

    Also did you check the capacity of the two year old cell? Again I can guess what you might have found but nothing beats data.

  3. Hi Jack,
    do you have a 10C discharge curve for the A123 pouch cell? Is there a dip in the voltage curve at the start of the discharge?

    And how about the tabs, are they plated copper and aluminum for the anode?

    I would love to see the voltage drop at 30C for these cells.

    No frills info packed session. GREAT.
    Keep them coming :)

  4., Jack paid for a 70A charger. Is the Leaf charger so big?
    What a beautiful battery case for such a small pack. A real "Look at me!" Jobbie.

    I would be using a lighter, solvent weldable plastic with 4mm correx separators between every 2nd cell to allow warming/cooling air flow. Easy, cheap and quick to fabricate.

    All pouch and cylindrical cells have lost favour with me on the price point. However, 123 pouches at 27% extra storage by volume and/or 10% less voltage sag at high C rates..... You pay for the pleasure.

  5. If you read the last photo's comment it mentions this box is being replaced with a half pound one. The copper is for > 2000 amps - but I wouldn't know if that is overkill or not.

  6. I was intrigued with the mention of the DIY charger in the latest video. Where can I find more information on it? All a Google search did was pull information on the bad boy charger or something similar.

  7. Randy , Here is the original thread from diyev forum:

    I built it and made some improvements and its been running in my car perfectly. Valery took over just recently and has really brought it on to be a great system:

    Also I have a couple videos on my website of development and testing

    Damien Maguire

  8. Jack, you have a talent for one-upping me. When I buy the 550amp kit, you come out a few days later and announce a 650 amp kit. I've been wrestling with installing a RF-50A, and was almost heartbroken when you had announced you had a 70A charging station. At least I still have the upper hand. =P Now about my unit. It's by an upstart company, they really don't have their act together. They seem to have fairly good hardware techs except when it comes to connectors. (I just had to RMA my front plate/cable because the connector came off the cable. =( ) Their software is a real quick and dirty hack. -- they don't really have any, they just crammed a little debian linux machine in there and only have a generic administrative console. The network setup is really bad, they expect you to set it up wireless networking using a laptop. I chose to remove the network interface and then hard-wire my network to the ethernet port. Problem is that it isn't set for auto-configuration that way. If the power goes down, I have to connect a machine to it, log into the admin console, and kick it over to the correct network. =\ I have numerous concerns about how it's circuit protection works. The circuit breakers are inaccessible unless you remove the cover, the LV circuits have no fuses at all, and there are no boiler-plate labels. My electrician had concerns about its grounding scheme when I had it installed. The company is slow to respond to communications, I typically have to harass them for 3-4 days straight before getting service. Yeah, this really is a nascent industry. I wish some of the big-boys would stop dicking around with puny-amp crap and get serious about 50A and 60A charging stations.

  9. Jack, you said serveral times in your latest videa, the charge voltage means nothing. But you also state that it is the result of the amps.

    But that is if you use a power source that delivers current. If you have a power source that delivers a voltage, like a solar power plant, it is very relevant.

    If the charging voltage is a result of the charging amps, the amps are the result of the voltage of the voltage source.

    Am I right? And can you there for translate the charging curve in a charging resistance curve?

  10. I''m not at all certain I understand the question. Power sources deliver power. We can measure this either with an ammeter or a voltmeter.

    In the case of a solar plant, with no load applied, we will get an open circuit voltage. If you connect this to series resistance, this voltage will drop and the current will increase as a function of the ESR of the source.

    In the case of a battery, it is a little bit different. The battery is also a source. And so if we connect one source (solar panel) to the other source (battery) we get a little battle going on between the two. Strongest source wins.

    Assuming the solar panel is the larger source, the apparent voltage or the voltage read, will be the result. As the sun wanes, this voltage will drop. On the other hand, as the battery energy increases, this voltage we read will increase. But it is simply NOT the voltage of the battery cell. The cell will never have a real open circuit voltage over 3.4v in the case of these LiFePo4 cells.

    To increase the amount of current going into the cell, we indeed have to increase the voltage of the source so there is a larger disparity between the source and the cell. But you won't read that source voltage because it is loaded - delivering current to the cell.

    So it's a bit of a special case and quite different from an inductive or a resistive load that we are all familiar with.

    So you are quite right, but I don't understand this "charging resistance curve".

    You will have a charge curve. And as the pack energy approaches maximum, it's ability to absorb power/current decreases and so it's apparent "voltage" or "resistance to charge" increases. This we can detect with a voltmeter, but do not confuse this voltage with the pack voltage - the open circuit voltage of the series of cells.

    This is why I describe this as a procedure - a recipe using a known "meter" that you can follow. The voltage is NOT only not sancrosanct, but it is the actual voltage of NOTHING. Neither the source, nor the cell.

    It is a CHARGE voltage generally termed.

    But if you hook up a meter and observe that, and put in ALL the current you want until you reach the recipe charge curve voltage, and then put in WHATEVER current/power is necessary to hold that voltage, until that current drops to 0.05C generally, (5 amps on a 100Ah cell), then you have reached a close approximation of a fully charged cell.

    THEN, if you let the cell rest for an hour, and the diffusion/surface charge is absorbed, you will read the open circuit voltage of the cell - usually just under 3.40v - 3.37 or 3.35. We undercharge a bit and usually see a 3.32 or 3.33v.

    Jack RIckard

  11. I like this iteration of your definition of Voltage. It was clear enough for a mech. engineer to understand. If only AC was that easy.

  12. To explain (for instance) a solar panel and a cell as two containers of water. Voltage is the fill height compared to each other and amps as the width of each container.

    AC is waves without an added dc component. ;-)


  13. Great show Jack! Love the battery testing. Any test in the future for Headways which are very easily procured.

    Thanks, Jim

  14. I love your show, and am starting a project car of my own. I have a question about solar cells, would they work to charge the batteries while my car sits in the parking lot at work? If they will, how much mileage could I add to my daily drive? I know you can't give absolute answers, but a nudge in the right direction would help.

  15. As soon as your car goes over a bump, the heavy glass panel breaks.

    Oh hum.

  16. Glenn:

    Welcome and we do sincerely hope you start a project soon.

    Solar Panels on your car roof. This is actually pretty easy to figure. If you work a full 8 hours and all of it is in direct sun, it should add about 80 feet to your range. Just as a rule of thumb, I would estimate 10 feet for each full hour of sunlight.

    Jack Rickard

  17. This comment has been removed by the author.

  18. I bought a poorly-converted '73 Beetle. Nicad flooded cells. I wouldn't make it up my hill. I just received 36 100ah Winston cells from China. The insert says to charge to 4.0v before ANY discharge. this seems to conflict with your bottom-balancing. What do I do?

    I'm getting strange looks from a local converter I'm considering hiring about no BMS. He insists miniBMS is the only way to go. I can' do it myself, 'cause I broke my shoulder between ordering and receiving the cells. Do I tell him forget it?

  19. You do not need to charge to 4.0 v ever.

    You do not need a BMS. It will NOT make your shoulder feel better.

    If the local converter will not do the work you want him to do, why would you hire him?

    And finally, if ANYONE insists there is only one way to go, you are talking to a moron. Why would you invest your time in such an activity.

    If you WANT a miniBMS, you should have one. If you do not want one, you should not have one.

    I show what we do and why we do that. It is YOUR VW and if you want to power it with lead, burn it to the ground with a BMS, hook all the cells in parallel and charge them to 80volts. Then I think that's just what you should do.

    At this point, we pretty much bottom balance everything. And we're pretty much giving up on the concept of advising people on electric vehicle builds.

    There now appear to be a sufficient number of "experts" in the field that our input is no longer needed. Indeed, they've reached such a number that you should be able to hire one full time for about $1.25 per hour. We're seeing guys at stoplights wearing signs "Will advise on EV's for food". Indeed one expert estimates that the number of EV experts now exceeds the number of operational electric cars by THREE to ONE.

    As I get about six e-mails weekly with advice from Dan Friedrickson, who's never built one and indeed appears to have never ridden in one, all at no charge, we're thinking of hiring them by the dozens and renting them out at a very modest markup. Kind of a clearinghouse for that sort of thing.

    Take MY advice and do whatever you like. But note that IF you park the cells in your garage, unopened in the box, and they sit there for another year, that when your shoulder heals and you decide you want to finish your car, they will all be fine. We predict you'll find them all at 3.30 volts.

    Yesterday, I got 98.5 AH out of a 2.5 year old 90Ah battery. These are GREAT cells.

  20. You've got a poorly converted VW that won't go up a hill. How could that be with so many experts who know "the only way to go".

    My friend, this little industry phenomenon is rapidly filling with charlatans and confidence men who know little or nothing about these cars, which is all the more remarkable in that they just aren't that complex. An EV is a pretty simple beast. As Brandon Hollinger puts it so aptly, "The stuff I took out was a LOT more complciated than the stuff I put in."

    Of course, if it WERE truly difficult and mystical, an expert could probably get $25 per hour from some hapless yuck for $1.25 worth of work.

    The guy who designed and built the miniBMS sold Dan Blocher's VoltBlocher circuits that actually ruined every battery they were ever installed on. He then took a survey on DIYelectric as to what people wanted in a BMS, and had these circuits available for delivery 28 DAYS LATER. HE is the expert YOUR expert is citing as "the only way to go." I spent an hour with him on the phone trying to explain how the basic charge process worked, he insisted he knew that and went on to describe a theory of it so bizarre I was in shock that anyone could have such detailed information on such total nonsense and I told him so. We've never spoken again.

    And there are HUNDREDS just like him on DIYelectricjunk and EndlessFear, and EVILDEAL.

    And so it goes. I'm discouraged by it, but shall not relent. We get in, we turn the key, and we drive away. We do it every day. And we're very happy with the outcomes. We share this information quite freely.

    But if you want me to advise you on what the rest of the world, including many many hundreds of experts, believe, I'm simply not up to the task. There is not enough of me, and I am not learned enough to do this. I have neither the education, the experience, nor the knowledge to take this on.

    Here's something I might suggest. What would happen if you waited until your shoulder fully healed, and then did the car yourself. Every morning, you take a DVM and measure the voltage of each and every cell. And every evening you did the same. And you charge it and discharge it and you watch those cells like a hawk. And you keep doing that until you simply don't care what hte voltage is because it really really doesn't matter anyway. And you are completely and thoroughly convinced of it.

    The other alternative of course is to Google it. And that seems to be the more popular one.

    Jack Rickard

  21. Wayne,

    The charger itself is a far more important issue than is BMS.
    A charger that charges to the correct voltage AND THEN SHUTS OFF is the way to keep your batteries healthy.
    Save the money on the BMS and spend it on a Brusa charger.



  22. Jack,

    Just watched your new show. Cool Show. Lot's of interesting stuff going on.

    I was thinking as I watched you do the fuel cap on the Cobra that maybe it would be better if you used the LED strip in a different way. It's a very minor point and your probably not going to care, but it would be better and more intuitive if you had the LED strip gently flash during charging. Flashing signifies activity. You could then have the LEDs stop flashing and stay on when the car was fully charged. When nothing is happening the lights would just be off. You could also do something with the timing of the flashes to signify the level of charge.

    ....My personality flaw is that I tend to over complicate things. I appreciate that you have already put it in and it works fine the way it is.

  23. Nickkkk.

    Actually, that's the magic of doing your own. You can go overboard to whatever degree you like, and it can be very cool. I think it will be nice to just see the EVSE come on by lighting the ring. So that's what we did.

    I do have some other LED plans for the car though.

    Jack Rickard

  24. Jack, your explanation of charging as a procedure really helped me understand what goes on. Thanks for your show, and for your kind advice.

    Mark, sounded good until I saw the price. Yeow! But thanks.

  25. Those proposed black Sino Poly cells, around 2/3 the size and Volume. I really do need them already.
    Jacks turning the DC-Cobra into a LED Sled.
    Last weekend. Went to the local Nissan garage to see if I can get a Leaf for the wife.

    I thought it was a good part-ex deal.

  26. Wayne,

    Get an Elcon once you know the AH rating of your pack and voltage of your pack. Get at least a 3000 or maybe even a 4000 model for faster charging. Mine works great.

    Pete :)

  27. EVCCON looks like it is going to be a great EVent!

    Can I camp in my trailer at the airport? It's self contained, I wouldn't need a restroom or hookups or anything. The trailer would do double duty hauling a lithium electric car.

    Will we need to be an SCCA or ECEDRA member to race? Will racing cost extra?


  28. Jack,

    I have, what I think would be, an interesting test for you. Perhaps you won’t agree, and that’s ok, but I just wanted to put this out there. Since you are doing the buddy pack for the Cobra, it brings up an interesting design question. You indicated at least one of the reasons for doing “buddy” cells of 90Ah instead of 180Ah is to help reduce cell variance as it relates to Ah.

    You alluded to this in your video. The 90Ah cells dropped 13.7% at 500amps compared to the 180Ah cells dropping 20% at 1000amps.

    How does the power output from a 40s1p (180Ah cells) pack compare to a 40s2p (90Ah cells) pack? Are the equivalent cell resistances the same for a 180 and a 90 Ah cell? If they are then you’ve essentially cut your cell resistance losses in half, which would mean you have less voltage sag and more power available to move the vehicle. If they are not the same, how do they compare? What about a 40s3p, 4p? Is there an advantage to parallel strings? Obviously this add considerable expense in battery interconnection costs, and if you are putting cell level instrumentation it would add up quickly, but given that these cells do not seem to require cell level instrumentation, could we gain some power by running parallel strings like you are putting into the Cobra?

    I’m sure you can rig up a test in your battery lab to find out some real data. Thanks for your consideration.


  29. You can camp in your trailer at the airport. I see no problem there. There IS an RV park with hookups in Cape not terribly far away. But i'm sure we can find a place for you to park at the airport. We will have a number of charging stations there, though not really an unlimited number.

    I don't think you have to be a member of anythiing to race. I don't think we've decided yet on any charges for race participation, but if we did it would be something nominal. I think $25 is traditional, but we haven't really decided on any of that. It's sort of a function of SCCA and ECEDRA more than anything.

    We have pretty much decided to have an EVTV OPEN as one of the races - full bodied 4 wheel cars with Lithiums and electric drive. BUt I'm pretty much decided on putting up a $2500 purse for that one race.

    We're also getting a lot of local interest, so we'll do some radio and so forth that week and I think we'll see six or eight thousand people from town out to watch.

    All the races are Friday evening.


  30. Palmer:

    There is no end to all the tests we could do, which is ok since we keep doing them. What you propose takes several thousands of dollars worth of additional constant current load. I actually have another 300 amp device in the stack, just haven't learned to use it yet. No knobs. I have to put in a 0-10v analog voltage to control it. I just need to rig up a pot but I don't have the proper mating plug so I'll have to jury rig something up.

    That would give me roughly 900 amps max full bore. As you can see, we can't really do a severe test on a 180Ah cell even then. And one of my 300A Acme's kind of comes and goes.

    I guess I was surprised at the apparent less sag on the Thundersky's myself so we'll be doing something. Perhaps the ESR is cummulative in strings. That would sort of make sense. So some side by side tests are in order.

    All complicated by new cells coming available from Sinopoly. So I don't know what we'll be testing next.

    I know I got 98.5 AH out of a 2.5 year old 90Ah cell. I LOVE these Chinese batteries. I think the OEM's are simply missing a trick overlookign these guys.

    Jack Rickard

  31. sky thinking...what about using the Soliton1 as a load. I'm not sure what you would do with the power, but you indicated that it does a pretty good job of controlling and documenting amperage. If you rigged up a current meter that put out a signal to the soliton that replicates the tachometer input you could let the Soliton1 "idle" and control current. For example a sensor that read 500amp and put out a signal that the Soliton1 thought was 500 rpm. 1000amps and 1000rpm. With this setup you would just change the idle RPM setting for the Soliton and it would control the current. Again, I don't know where you'd put this power, but it might be a way of testing and getting some crude numbers without lots of expensive equipment to get better accuracy.

  32. Palmer:

    I don't have ANY problem measuring the current, but it has to GO somewhere. Into some sort of load and it is much more instructive to do so into a CONSTANT CURRENT load. So the Soliton offers us nothing in this situation.

    We basically burn this power off as heat. I have a 240v inverter, but it has a capacity of about 40 amperes. To burn off 1000 amps, we have to have a car on the road, a car on the dynomometer, or several constant current loads in parallel. These devices cost about $5 per amp. If I can get all of them to run, I can do about 900 amps right now max in theory. The highest levels we've done is a little over 500 amps.

    Yes, I coudl hook up a piece of galvanized wire and stick it in a bucket of water. Perhaps blow up a battery, but it would at least destroy it. And while I'm not averse to this, to learn what again?

    If we demonstrate a certain sag at a certain load level, you can pretty much extrapolate that foreward at no expense at all. The cell with LESS sag at 400amps isn't going to reverse and sag MORE at 800 amps or 1600 amps vis a vis the other cell. So I'm a little lost as to what you seek, and unless it leads to something I too am curious about, I don't quite see destroying cells to gain YouTube viewership or something.

    Jack Rickard

  33. Regarding the "charge to 4 volts" phenomenon, I'm fairly convinced this was a misinterpretation of a poor translation from Chinese to English, and completely unnecessary. Originally it was supposed to be 4.25V, then 4.2V, now 4.0V, and was "required" to "set" the chemistry of the cell. Of course they already fully charged and discharged the cell at the factory at least once to see if it worked and to measure the actual capacity and resistance. So since it's already been fully charged at the factory and recommendations for longer life are to avoid fully charging the cell on a regular basis there would seem to be no logical reason to charge the cell to 4.0V that I can see.

    On an unrelated note, Jack mentioned killing Chennic DC/DC converters and I mentioned that mine was still working fine, possibly because I used a battery as well. Of course I spoke too soon. After sitting on a shelf for a few months out of the car it no longer functioned when I hooked it up again. I took it apart to see what I could see, which was a block of black epoxy potting which I have no interest in chipping away at. The dream of a cheap DC/DC seems to be dead for now.

  34. Sorry for the confusion. I do not intend for you to destroy cells. I misunderstood what you said you were lacking and was just dreaming up a crazy way of controlling the current for your testing.

    I'll try to avoid taking this site on wild goose chases in the future. Sorry.

  35. lol@blue sky thinking.
    I'm still waiting for an enterprising 'H' bridge controller manuf'rer to come out with a charger based through the regen circuitry.

    Why pay twice?