If you’re not living on the edge, you’re taking up too much room. Or so I’ve been told.
This week’s show is almost entirely devoted to watching paint dry. But it’s fast dry paint. We attempted to charge a CA40FI series cell at a 3C 120 ampere rate.
The Chinese manufacturers of the Thundersky, Sky Energy – now China Aviation Lithium Battery Company, have always listed 3C as the maximum charge rate. In the past two years, they have largely not touted that very loudly and in fact it appears to have disappeared from what little they now publish by way of spec sheet. We really don’t have a proper spec sheet from the manufacturer on these cells.They now include a little booklet that appears to have rounded up every myth known to batterydom into one single grossly inaccurate document for your perusal. Please ignore ALL of it.
We have previously not had either a power supply large enough or a cell small enough to accomplish a 3C charge rate, much less charge a car at 3C. I had done some earlier charging at 1C and higher and had a sense of it not being terribly different actually. But we had just never been in a position to test 3C charging.
The Society of Automotive Engineers, largely in response to Tesla’s Supercharger unveiling, hastily released their J1772-2012-10 standard this past month. That rather makes real the concept of a standardized FAST charge ability for the first time.
The Japanese Tokyo Electric Power Company (TEPCO) had in fact installed a number of fast chargers in Tokyo terming their “standard” ChaDEmo. A play on words meaning have a cup of tea and alluding to the fact that you could charge your car in the time it took to take a cup of tea – a common stop in Japan akin to our swing through the gas station for a big gulp soda. But to get a copy of the actual standard you had to join their club at $150,000 and so we never did quite get to read the actual document. I find it curiously annoying that a body such as SAE purports to publish a “standard” actually charges for a copy of it. It’s simply inappropriate. It should seek as wide a circulation as possible if it is held out to be a standard. That they also employ some sort of DIgital Rights Management and FORCE you to actually install the enforcement software on YOUR computer to download it is simply evil and despicable. But the $66 is a little better than ChaDEmo anyway.
A few ChaDEmo charge stations HAVE been deployed in the U.S. One in Oregon, I think one in California, and a couple in Chicago. Nissan had heavily told the story of the devices at ALL Nissan dealers across the country, which would have by itself established a fast charge network, but like all things Nissan, this appears too to be a bit overstated. We’ve seen NONE of them. Indeed, they had announced availability of a fast charge station device at a very modest $9,900. Again, unobtainium nonsense. Carlos Ghosn really needs more of a plaid sports coat look. He dresses to well for his business practices which do border on the used car lot end of the autospace.
[jwplayer file=”news102612-iPhone.m4v” hd.file=”news102612-1280.mov” image=”http://media3.ev-tv.me/news102612.jpg” streamer=”rtmp://s3einxnpkaij93.cloudfront.net/cfx/st/” provider=”rtmp” width=”950″ height=”584″ html5_file=”http://media3.ev-tv.me/news102612-1280.mov”]
If it hasn’t become apparent to some of our newer viewers, I suffer fools and bullshit artists poorly.
As I think all the hype on public charging infrastructure, certainly Level II AC charging, is poorly directed, why the diatribe on FAST charging standards?
There is little advantage to public charging at 240vac 80 amps. You can do that at home. And largely do. And it is a rare instance in a car with 80 mile range to have any NEED for public charging. The only real scenario is when you are traveling non-local distances – such as between cities. And then it is too little too late. An eight hour stop to charge after a one and a half hour drive just doesn’t make any sense at all.
Fast charging, on the other hand, given the touted 20 minute scenario, is somewhat enabling. It would allow you to drive 80 miles and charge in 20 minutes. In reality, it takes 20 minutes to gas your car – not the 5 minutes universally quoted. I’ve NEVER been in and out of a gas station in
five minutes. But 15-20 is normal. If this was extended to 20-30, no problem.
Unfortunately, what SAE adopted is a 200 amp standard. On higher voltage OEM cars, this might indeed do 80% charge in 20 minutes. But it really doesn’t for our cars, which tend to be 200v or less. Our most popular cell size is of course the CA180FI. And at 200 amps, it would take most of an hour to recharge.
But even that would be enabling. It would turn a cross country trek from a total endurance test to mildly annoying. We are not GOING to have a national network of fast charging stations without a fast charging standard. And so while I abhor the 200 amp limit (how hard IS 300 amps), and likewise share Elon Musk’s opinion of the ridiculous charge connector – just an engineering and design mess – I’m pleased to have a standard at all.
But it brings up the question as to what effect fast charging has on lithium battery cells in general, and of course specifically the cells we use in our converted vehicles. From my biased perspective, the China Aviation Lithim Battery Company CA series cells are the only ones that matter.
They are THAT much better than the THundersky, Winston, Sinopoly cells that no amount of discounting by those guys could realistically move me to go to the trouble of a conversion with those earlier chemistry cells.
And so I’ve kind of myopically drilled down to WHAT specifically would be the behavior of the new CA series cells in a fast charging situation.
The key is temperature. So we invested about $600 in a Raytek MI series temperature sensor. This is a very accurate and very quick infrared temperature sensor. We would love to know the temperature of the anode inside the cell, but we can’t get it without compromising the cell. So we settle for taking the temperature of the anode terminal bolt head.
This stainless bolt conducts electricity or course, but also conducts heat rather well. The actual charge current goes from the copper terminal lug of our charge cable to the copper flat of the anode terminal. The bolt just holds it in place. It may pickup a little heat from this interface under current, but not much and it really shouldn’t be conducting any current at all if our interface is clean. It will however conduct heat from the anode pretty well and so we think it represents anode temperature pretty fair. As it radiates somewhat better than the anode foils, it will be smidge lower than the true anode temperature. But the relationship should be pretty linear.
Why the anode? In charging a cell, this is where the heat shows up. As lithium ions diffuse into the graphite anode, there is a heat gain. And so you will observe higher temperatures at the anode in charging than on the cathode.
Of course, the question of cell damage is really a question of heat. While I am somewhat blase about the ambient heat of Phoenix and its possible effect on the cell, internally generated heat is pretty much an indication of wear and tear on the cell. Ambient doesn’t HELP this situation, but it really isn’t entirely related. We’re interested in the effects of high current charging on the cell and that means the cell internals really.
We have done some casual temperature measurements of cells while charging and discharging before. Usually by taping a probe to the case or shooting it with the DeWalt infrared. But we really weren’t focused on it much. I had kind of a sense of WHEN the temperature rise occurred during the charge cycle, but had never really tested it in any methodical fashion.
So we had questions on two points:
1. What level or percentage of charge could be accomplished with a 3C charge? This goes to a very tricky question. In order to charge fast, you have to have maximum current. We can do that up to our Constant Voltage switch point. After that, we go into Constant Voltage phase and as the energy level in the cell rises, the current MUST diminish to maintain that voltage. That inherently reduces the charge rate and so extends the charge time. Fast charge is about FAST so there is no point sitting there charging at lower rates. If you charge at max current up to the CV voltage, what percentage charge do you accomplish. The common myth is 80%. A very believable number since in normal charging that is about where you are when switching to CV.
2. Does fast charging damage the cell? This would best be answered by doing cycle tests for 100 cycles or more to 100% discharge while charging at 3C. Not convenient. So we look for two things – temperature on the anode while/after fast charging, and an immediate retest of capacity after fast charging. Any serious damage should show up there with one or the other indication.
This doesn’t mean that without temperature or capacity decrease there is NO damage to the cell. But none that needs be immediately addressed and likely, if it doesn’t show up at ALL there, it is unlikely that it is damage that needs to be addressed – probably at all.
Prior to the test, I would have believed any result except the one we obtained.
In regards to the first point, we APPEARED to have charged the cell to 97% capacity by charging at a steady 3C (120 amperes) to a voltage of 3.65v. This is quite surprising. In fact, I almost fell over. Comparing the amount put IN to the amount that comes back OUT on a subsequent capacity test, and repeating this whole thing 3 times this week, we are getting a very steady 39.685 amp hours in capacity OUT at 30 amps. But we put about 41 and change IN when charging. I think we are seeing normal charge efficiency losses here. And so it really is NOT 97% charged – probably more like 92% charged. That is STILL a very astounding result.
With regards to temperature, even though I predicted that the temperature gain would be greater toward the end of charge, I was absolutely astounded at what happened. FAST charge at 120 amps to that 94% charge level resulted in a temperature gain of 10F (5.5C) from an ambient of 73F to a finished temperature of 83C. No real delay here either. When we shut off the charge, the temperature immediately starts to cool.
But AFTER fast charge, and after cooling back down to 73F. we put the Powerlab8 on to charge an additional 2.5 Ah into the cell in the normal CC/CV fashion at 30 amperes. The temperature peaked at 3.65v and 30 amps at 126F. This represents a 53F gain. This is a FIVEFOLD greater gain than in the fast charge scenario. Slice that, dice that, and discount that to any level you want, it will leave a remainder that is still multiples of the fast charge temperature gain. And even though I predicted it, I did NOT see that coming. I would have guessed 50% gain in the first 90% charge and the other 50% in the last 10% and would have found that DRAMATIC. This was not dramatic, it takes us to an entirely OTHER conclusion.
We are overcharging our cells. Obviously we are at 3.65. But I suspect at lower voltages as well. The last portion of the charge curve is obviously damaging. The further away you are from it, the better. And it couldn’t possibly be overstated given this evidence. Remember we were NOT fast charging at all during the Powerlab8 portion of the charge. In fact we were at 1/4th the current seeing 5x temperature gain.
The apparent voltage while fast charging IS signficantly higher. And so while we would slow charge to 3.50 we can probably safely fast charge to 3.55 x Number of cells and will wind up with about an 83% SOC at that level. That would be our real world FAST charge scenario for these cells. At 200 amps on a 180C, we would probably be back at 3.50 v per cell.
That greatly simplifies fast charging. You simply charge to your normal CC/CV voltage, but then cut it off once it is reached with no tapering charge. And I would at this point rest easy that I could do as much of that kind of fast charging as desired, with no thought to battery cycle life. At this point, I have NO (read zero or below) evidence or even indication of any damaging effect of fast charging on these cells.
That said, I am increasingly finding sufficient difference in behaviour in these CA series cells to admonish you that if you run Thundersky or even the SE series from CALB that these results may not apply to your car. Can we duplicate this test for those cells? Frankly, I’ve just lost interest. CALB continues to make the SE cells in a few odd sizes, but their availability has already dwindled. The Sinopoly and Winston cells and High Power and Headway remain available, but frankly we just don’t think they matter anymore. Until they counter the CA series, they are kind of yesterday’s news.
And THAT said, note that SAE Level II DC charging is going to limit most of you to 2C on 100Ah cells and barely 1C on the larger and more common 180Ah cells as well. So I don’t think that at the charging rates AVAILABLE that any of those cells will really have any problem either.
Bottom line, as I’ve always said, the cells can do fast charge now. The problem is getting the power. We’ll see if the adoption of the SAE fast charging standard leads to actual deployment of fast chargers.
Meanwhile WE intend to pursue it. There is not a week goes by that we don’t hear from someone wanting to charge from their own home solar charged battery bank. I think the PulsaR can do that as well as provide us a path to SAE fast charging. Ryan will have to work out this goofy PLC data protocol, but assures me that it really isn’t much more than an annoyance and he’s mostly annoyed at the $150 fee for the specification itself.
As am I.
The other notable element in this episode is a continuation of the sea change shift in the fortunes of the conversion crowd. By upgrading the nature of the builds, at a time when the automaker and battery company fortunes are in shambles, it has become ok to be a conversion shop. Steve Burns of AMP Electric Vehicles stopped by to note that they were really just like us in many ways and Fisker has announced they are showing up at SEMA. Audi has cancelled production plans for their E8 E-Tron with the usual greasy lies about always intending it to be an internal prototype. Nissan is heavily discounting the Leaf to comical levels and indeed rumors about of a DOWNSCALE LEAF if you can believe it for 2013. GM is going the other direction and doing a Cadillac ELR version of the Volt. This was my advice from the beginning two years ago. You can hide the cost of the drivetrain more easily in a car that is already higher in price. And you HAVE to appeal to well heeled early adopters – not the economy shopper. The case for an economic reason to drive electric CANNOT BE MADE until economies of scale come into play and they are NOT going to come into play by building a billion dollar plant in Tennessee to make 150,000 economy cars no one is going to buy. To break a chicken and egg standoff, you have to tap on the egg GENTLY or talk to the chicken PERSUASIVELY. Bludgeoning the two leads
to a messy dinner, but no more chickens or eggs.
I truly believe the number of people willing to pay a significant premium for the advantages of an electric car will grow steadily toward a tipping point. But anyone who wants to be a player has to offer a very desirable car to those who can afford it. In a season where “trickle down” is a dirty word, it is only dirty to those who truly do NOT understand how economies work – or disruptive technologies for that matter. But I see a sea change in the wind. I think we’ll be getting better components here shortly, and the DIY “hobbyist” conversion guys are going to eventually be recognized as the most persuasive and influential key to the puzzle. We are already seeing companies that recoiled in horror at being besmirched with the same brush come around to the reality that they are really just like us, but bigger. And facing the same problems. And the “we only sell to OEM” suppliers have learned a very valuable lesson in marketing. You sell to the guy willing to buy – not to some surreal imagined “target market.”
A 74 year old guy wandered into our shop and wanted to buy a Soliton 1, an 11 inch motor, and enough batteries to light Cleveland, all to run a Chevy S10 pickup truck. I told him I would be happy to sell them to him, but he would most likely hurt himself with such a grossly overpowered beast. He hauled out a 30 inch framed photograph showing him WINNING one of the largest drag races in the deep south with the same truck using an ICE engine. “Thought I might try to do it again in electric” he noted.
Noted. Here, I’ll help you load em up. Valuable lesson repeated. Remind me to practice my “sit down and shut up” moves more often.
As to the CA cells. We are about sold out. Pomona is sold out. I talked to CALB about selling their cells last May and wasn’t certain we could meet their required quantities to be a dealer. We’ve rather become their largest outlet. But there will actually be a bit of a shortage until mid-November. Which is just a couple of weeks at this point. WE are putting in a larger order this time in the hopes of meeting our viewers requirements timely. But the more we examine these strange new cells, the better we like them.
The final thing we learned in this test is just how very capable this Revolectrix PowerLab 8 Battery Workstation is. It has a LOT more utilty when coupled with their really very good software. You can use it to charge at up to 40 amperes. You can not only use it to discharge cells, but can discharge using constant current/constant voltage just as you do when charging. You can use it to perform complete cycles, charging and then discharging or discharging followed by charging. And we actually used it to monitor and graph our progress when using a DIFFERENT power supply to fast charge the cells. It makes beautiful graphs, but better you can export all the graph data for voltage, current, time, etc. to a tab delimited file for later imort into Excel or whatever. This thing is cool.
89 thoughts on “CHARGE FAST – DIE YOUNG – LEAVE A GREAT LOOKING CAR”
Bravo! No one should miss the significance of what you have demonstrated here (or the procedural elegance with which you demonstrated it). If later work confirms what you showed us, 20 minutes to 90% + charge with no capacity loss is a game changer. With the right infrastructure Petrol (Gas) stations could install fast charge stations which would make inter-city travel viable: and no one would gripe at a huge markup for the convenience on the three occasions a year that they make long trips.
The oil companies and the OEMs may yet kick EVs into touch but if they do so it will be politics and entrenched interests not genuine feasibility that is the problem.
Obviously I agree John. With the establishment of a standard form, even an ugly one with J1772 Revision B, and the ability of the batteries themselves to perform this act, we are a bit of hardware away from being able to do this.
I have been saying for some time that our range is more than adequate. But to enable full freedom of use, we need to be able to refuel it somewhat on par with gasoline cars.
But what this means is the car isnt’ the problem. It is again the infrastructure. To fuel a car with a 35 kw pack in 20 minutes does indeed require 100 kw. As few cars are 500v, I think we need yet another current level – something on the order of 300-400 amps.
If you have a 180Ah pack, a 20 minute charge means over 500 amps. And that isn’t happening even with Rev B. Until that point, we cannot refuel in the 20 minute time frame.
But here I think we demonstrate that the batteries, and thus the car, isn’t the problem. If we had fueling stations of sufficient power, the car will do fine.
DC charging brings up another point entirely. I should have probably mentioned this in the video and the blog. Why would we need a charger in the car at all? We basically need a data communications device and instrumentation to talk to the EVSE and direct the charge voltage and current. This is currently a couple of thousand dollar device in each car – the charger – largely because it has to have power electronics in it. If we move the power electronics to the charge station, we are left with very low cost data communications electronics in the vehicle, about a 20 lb weight loss in many cases, an order of magnitude less cost, and it is simpler.
A home EVSE could be made to have the same features as a fast charge station.
We are likely going to do this project, using a PulsaR and batttery bank to form a charging station, and a small PLC communications device that also measures current and voltage to communicate with it. In this way, the car won’t have a charger per se.
But the bottom line is that the only real drawback to an electric car is intercity travel. With a network of fast charge stations, this kind of goes away.
I could very easily see five or seven years out a network of such stations. With battery improvements in that time frame, assuming rate of progress of the past few years, 140-150 mile range becomes very doable.
Here in the U.S. we have 43,000 miles of Interstate Highway and we have already calculated that 1150 stations would place one every 50 miles on the ENTIRE SYSTEM. At $250,000 per station, we would be talking about $287.5 million dollars total. They have already spent $115 million on very dubious work at establishing a totally useless AC Level I and II network. In the scheme of things, this is not very much money to enable a vehicle class that addresses a $365 BILLION per year gasoline habit here.
So all becomes possible IF the batteries will do it. And my crude testing this week would seem to imply that they will. That moves the problem out of the car and onto the ground – a much easier place to develop things.
Iv’e been thinking about this infrastructure issue for some time and your figure of $250,000 per station doesn’t need to be all at once.
Because there are relatively few electric cars travelling long distances, a DC station capable of charging at 3C would only need a relatively small number of batteries. These batteries would be charged at an AC managable current so the batteries can then, in turn, deliver the DC fast charge. As the number of cars using the station increases, you simply add more batteries. This spreads the cost over time.
If the government built it, it will cost $250,000 if it does 4 amps at 12volts.
But what you say is quite true. I think a large cost will be siting. And getting electricity to it at all. But I’m basically seeing a small solar roof, a fairly large thing of batteries, and a couple or three charging ports.
This is one of the best shows ever, really interesting with great explanations and results.
On the subject of speeding up dc charging by using higher voltage. It’s a rather Rube Goldberg solution but you could use the old series/parallel dpdt throw knife switch to double the dc voltage while charging if you had identical half packs.
Jack, you know your “blow me” turbo fans for cooling the warp motors. Have you thought of using an intercooler designed to work with torbochargers to cool the air that’s being forced into the internal combustion engine to cool the air being blown into the DC motor(s)?
I am not Jack, however I do believe I can speak on this matter. It’s hard to cool air that is already ambient to below ambient by merely running it through a intercooler. Gas motors spin the turbo’s with spent exhaust gas in a conventional I.C.E. setup.
All the best,
Ahhh….and the penny drops. Thanks Aaron.
Pads, That would cost energy, reliability and complexity over a relatively few degrees.
I agree, I did’t quite think that one through. Well, at least I’m thinking. Sort of.
Jack — A great show. Lots of important information about batteries and fast charging. John Hardy said it all — Bravo!
A couple of observations. For fast charging at home, unless we find an alternative way to deliver 40kW to our pack in 20 minutes, the energy is not likely going to come to our houses from the local utility company. Most houses built in the last half of the 20th century have 200+ amp service at 240 volts. Solar/battery banks are a way get there, but the current solar marketplace is not demanding this right now. Just try to run a Solar City solar system during a power outage, even with the required utility disconnect.
Second as for the SAE specifications, one should not necessarily want to know how laws, sausages and SAE specs are made.
We get a lot of interest from those wanting to charge with solar from home.
At home, slow charging is pretty easy to deal with. After your third day with an electric car, you either quit worrying about charge times or else fall to the ground from sleep deprivation. If you can sleep, you can charge.
That said, an at home Fast charging station would be fairly trivial. A bank of retired batteries. Remember, it can be charging on a rickety old 1500 watt charger as it is charging 24×7. You dip into the mother bank for 20 minutes as necessary.
But I agree, I don’t really see the point. FAST charging is all about going intercity. I don’t NEED to do that. But it would be NICE to do that and it eliminates one of the psychological hurdles to electric vehicle ownership.
Jack, I am currently putting 2.2KW solar panels on my roof and with the old Thunderskys I took out of my car when I replaced them with the Sinopolys, and an 8KW inverter I have, should be able to charge my car daily. I have 3* 2Kw chargers, two of which are in the car.
Will let you know when its all going.
I love these kind of experiments. Makes you wonder though why CALB isn’t doing them and posting their findings. You think they would want to.
A bridge too far. By way of example, Ted Han AKA Ted Mosby is the head of marketing for CALB. He has never forgiven me for laughing at his very brilliant marketing idea for
the U.S. He wanted me to help with the slogan CALB – Cost A Less Bill.
The documentation with the CA cells is a little booklet that actually serves as a roundup for every single battery myth I’ve ever heard. It is comical.
And they do not actually HAVE a spec sheet of any kind for the CA series cells. They list the weight and dimensions – that’s it. No spec sheet available AT ALL.
That said, we’re kind of off their radar screen at this point. I would say they are very disappointed in EV sales. And their dreams were of course to land a U.S. OEM. I think
it has dawned on them that that is not in the cards. They HAVE landed an OEM car in China. But the talk now is kind of pie in the sky about grid storage and solar and
wind – large sales to big storage facilities. None known. But that’s where their head is at. There have been conversations implying I should just handle the EV space in the U.S.
so they can move on to these bigger and better things. I have noted that that’s all good, but not to count their chickens before they hatch on the utility/grid market.
I actually do have a document I wish I could share. Actually I DID share a couple of graphs from it that I wasn’t supposed to last June. That’s what started all the testing. They made
some preposterous claims for the cells without much in specifics beyond “70% more power” and “improved cold weather performance” that was barely in Chinglish. That’s what set me
off on the testing of both power and cold discharge.
Bottom line, their understanding of the American market is kind of deranged and their marketing is CALB – COST A LESS BILL. They wanted at one time for me to build a car and drive it around the country to OEMS and show it to them.
Naive doesn’t begin it.
But I think they are making an excellent cell. And we’re selling a lot of them at this point. In the world of U.S. automakers and battery makers and so forth, I think I’m sitting on a secret for all intents and purposes. It’s a very
How did they do it? There are veiled allusions to nanopowder sized things. I think they may have taken a page from A123 and given it a large format prismatic twist. But they just
don’t talk about such things.
Thanks for the great heads up on cell temperatures. Been waiting for this a looong time!
Tesla cells come with PTC’s, BMS man’frs are stuck on charging voltages… Temperature is the real deal that effects life, capacity, final voltage, charge uptake, power delivery and what-not.
All the people who wanted that tip top charging voltage instead of more capacity will always be heading for a pillow to cry into.
On the brighter side. Calb’s CA cells have done well in reducing heat sensitivity.
I do wonder how well a pack of many bottom-balanced cells in series – ie, an actual vehicle battery pack – will behave when charged at 3C? Note that this is a serious question; not an implied suggestion that anything untoward will happen.
It is an interesting question. We do not charge cells in series strings to that high a voltage for example. I would cut off at 3.5v as always and that is probably a little over 80% SOC for example.
How’s the CHARGER end of all this going?
We do not currently have a build going with the new CA cells. We are going to do a “mother bank” of assorted and varied cells we have laying around. It will probably do double duty as a power source for a dynommeter and fast charge power source.
As soon as we get a build done with CA series cells, we’ll test this further of course. But I don’t see any difficulties or even particularly interesting questions beyond what happens when you mix an odd lot of various cell sizes in the mother bank.
The other issue of course is actually CHARGING at 3C. Dubious. That would be 540 amps on a 180Ah pack. J1772 DC Level II is only 200 amps.
Peter McWade suggest there may be more to the temperature diffeetial than SOC. His theory is the higher current levels make diffusion/penetration EASIER for the cell than low current levels. Would that not be the crap if it
turned out to be true.
We may do some comparatives at 50% SOC at 3C and 0.5C to determine if there is an innate temperature difference in the two charge rates. THAT would be frightening.
I guess this is related to a question I had about your heating and fast charge experiment. Have you considered trying a fast charge to a higher target voltage, say 3.8 volts? I see the voltage was rising rapidly, but you where so close to full charge that you may make it all the way on a fast charge. I wonder if you would see the same kind of heat rise right at the end like you did finishing at low current.
Have you measured the terminal temperature for some time after the charge was complete to see how much longer it rose and how long it took to start back down? That hints at how effective heat conduction is inside the cell.
Jeffrey – FWIW in my testing I charged an 8-cell (Headway) string at 2C with and without CV phase. With no (or little) CV the cell voltages tracked each other pretty closely. If you open http://tovey-books.co.uk/attachments/File/Test_1_Summary_Data.zip and look at the Excel file for cycle 200 you will see a typical trace. Interpolating from the cycle 100 trace (in the same zip file) I was (Ball park) leaving about 10% of cell capacity on the table dispensing with the CV phase. This is absolutely in line with Jack’s findings with a single cell.
Very interesting episode. I’m the type that takes a little bit of knowledge (that I don’t fully understand) and proceeds to wield it dangerously. So given your anecdotal “night club” analogy and your test results, is it possible that the fastest and safest way to charge, might be a sort of constant temperature method (ie push current until you begin to see a slight rise in temperature)? Assuming an unlimited supply of power, how much amperage could you push into one of these cells safely? 4C? 6C? 10C? And could you push more in early and taper with the level of charge? Thanks for humoring my curiosity.
The rise in temperature to get the last bit of energy into the cell brings up the question: Is the wall outlet cost/watt of charging a cell for the first 90-95% of the charge different then the cost/watt of the last 5-10%? The temperature difference between the two test charges during the show would certainly indicate a difference in efficiency. Hooking up a JLD404 to the input of the Powerlab would show a difference in power source ah at different parts of the charge curve.
Excellent show. Fast paint drying is even better than slow paint drying. I was wondering if you planned to try any higher rates? Theoretically it looks like you could try at 5C if you wanted, with the equipment you have. That would d be a good way, and entertaining in my eyes, to see if it’s actually easier on the cells to charge at a higher rate than a lower rate and it could really show what these cells are capable of.
It would certainly turn the question of fast charging on it’s head if it showed a smaller thermal gain at even higher rates as opposed to lower rates.
Loved the show this week….!!!
It is great to see some real data vs internet BS. I kind of suspected that the prediction of damage due to fast charging were over blown simply because if it were true regenerative braking would more or less have killed the battery packs on all of the cars that can do it by now…
This does open up the possibility of long distance travel without having to drag a gas/diesel generator around with you.
I really cant wait to get my car finished. The PLC in the car will give me enough instrumentation to produce what I hope will be some very useful real world data. One of my reasons for building the cars was to be a rolling laboratory…..
The SAE spec of 500V at 200A is 100kw of available input power. If you can trick the EVSE properly you can buck the voltage down to your 38 cell level (139V) and amplify the current. Probably something over 500 amps (at 100% efficiency this would be up to 719 amps) which is not quite 3C but close enough.
My first thought was what about 4C or even 5C charge rates but the answer is not something that matters if all you have is 100kw to work with or even worse when limited to just 200 amps.
Because all the temperature rise is in the last 3 % of the energy you put in I am thinking that most of the 5% losses of the 95% charge efficiency happens in that last little bit and it probably doesn’t matter much at all what the charge rate is before you get to the CV part of the charge. I’ll put this on my list of stuff to look at once I get the car more complete (unless you do it first).
Have you got any 40AH Calb CA cells in stock you want to sell? I don’t see them on the store. I could use 4 to make a 12v buffer battery and one for testing. (I’ve got a 5V 400 amp power supply so I could test one of these at charge rates up to 10C just for grins.)
Well I guess my question from my previous post, of the differential between CC/CV has been answered. The actual reason I asked that, was because I had already suspected that a constant current fast charge would bring the cell very close to full. Many years ago before endless sphere and diy, I like Jack, bought some developers pack from A123 and did some bench testing. Since being an electronic tech and having some 20 of my earlier years in the automotive field, not to mention also an amateur radio operator, and therefor have always been interested in battery technology.
Since playing around with those A123 cells and wanting to know more about them, I discovered endless, dyi and of course EVTV. Reading stuff in endless and diy, I could not understand why they all insisted that a bms was a must on lipo4, my tests on the A123 indicated that that was not needed, and only a back up monitoring for equipment failure was necessary. Since I discovered by using a Cellpro4, which was the top charger from Revo at that time, all my A123’s were very close in capacity, it never occurred to me that I would have to bottom balance first, before using my 30A power supply to charge them to 3.6v per cell, but never the less while monitoring with the Cellpro, found that all individual cells pretty much came up at the same time, later on I decided that for safety sake I would charge them to 3.5 volt or so, to give me a little safety margin.
You can imagine my surprise when I discovered EVTV and started watching, “well I be I be damned, there is someone who at least testing these cells, and telling it like it is”, and I’ve been a fan ever since.
A biggish question:…..
I think we now have data which suggests that
1. under-charging is good
2. much of the cell temperature rise seems to be in the CV phase
3. wild cell voltage variances occur in the CV phase
…so should we just forget this silly CC/CV stuff and just charge to a voltage and switch off? It makes the charger a lot simpler.
I was wondering when this would come up. Basically yes. Then we could charge to the book voltage and just DROP the CV phase entirely.
It would bring us very nearly to the same place and cut the charge time at ANY charge rate.
FULL CURRENT CHARGING
Jack & John,
Actually I already made this observation (to drop the CV phase) in a comment to your previous post. http://blog.evtv.me/2012/10/1821/#comment-7554 😉
What I’ve been wondering is what would be the simplest way to charge from battery bank at the garage to a car. If I had a bank with a higher voltage and amp hour count than my car pack could I just connect it and let the electrons flow into the smaller pack in the car and what rate would this happen? As fast as the bank car discharge?
Correction to last sentence: As fast as the bank can discharge?
Quite right Jarkko – not sure why I missed it. A case of thinking alike rather than plagiarism
Quite right on the first bit that is…
On the second part about charging from a battery bank you would still need a way to limit current I guess unless you wanted to experiment with 50 C charging
Indeed more like thinking alike. Too much current from the bank could be the problem. You’d need a pretty big and reliable contactor to handle it as well. Perhaps if you used inferior batteries that wouldn’t be capable of too high currents?
This lead to another idea. If the charging from a bank to a smaller pack idea works in the first place, you could use a similar scenario to test even faster charging. My blue CALB SE40AHA cells will put out 400 amps. Connect two of them in series and connect these two in parallel to a third one and it should theoretically charge at 10C, right?
How about make it easy. Connect the mother bank to the input terminals of the Pulsar. Connect the output terminals to the car. Set up the configuration for the voltage and the current of the car charge. Let the PWM control thte voltage and the current.
Hey Jack, about your comment here about using the pulsar…. I think I remember Jay Whitacre stated in his Lithium battery lecture that charge rates could be raised 40 or 50 % by pulsing the current to the battery. Is this part of the idea in using the Pulsar and have you tested this theory?
He didn’t really say charge rates could be raised that way. But he said it might be useful in reducing heat and stress due to diffusion delays. Try this, fully charge your cell. Then let it rest a day and try to do it again. It will take another couple of Ah with no apparent ill effects. So was it full before or was it not?
The ions rearrange themselves in the matrix. Pulsed charging might have the same effect at the right frequency and power level and no one has any idea what might be optimum for that.
Surprised nobody caught this. I made a huge booboo on the test. It doesnt’ really have anything to do with the fast charge results. But that 73 to 126 F temperature gain with the last 2.5Ah and at 30 amps – likely bogus.
If you notice, I’m charging with steel spring clips onto the steel bolt head that I have the temperature gun aimed at. Steel is not a very good conductor.
I’ll have to repeat the fast charge to get to the same place, and then do the final charge with the clips on the cable bolts away from the cell to get a true temperature gain on the last 2.5 Ah at 30 amps.
Now the next question, or as Jack says “there is more”…..
A while back when I tried to figure out what battery pack I wanted to put in my little vehicle, I was adding diffferent pack sizes. Originaly I wanted to go 48 cells, so figuring 3.5vpc that came out to 168v, soon as I saw that figure I said “168 volts thats interesting”, since 120v mains voltage rectified comes close to 168v dc. Now mind you someone in diy claimed he had tried that, and if my memory is correct, claimed he used a 50 amp breaker and charged his pack directly from it. I dont recall how large his pack was but he claimed it charged it close enough so he could use it as a fast charge to make another trip, but not quite full. I also dont recall what amps it charged at.
So the question is, has anyone here tried this ??? Or want to try it.
Roy, that 168v is open circuit voltage. It will drop when a load is applied. Also you will not get a good DC voltage off the bridge. It will look like little bunny hops on a scope. You have to add capacitors to level out the bunny hops.
This type of charging would probably simply not work well for many Other reasons….
Actually, it has been done and works pretty well. It’ has come to be called a “bad boy charger”. If you rectify your AC, even 240vac, you can apply it to the batteries and they will indeed charge. There is no particularly known effect on cells due to the “bunny hops” and some theorize it may actually be advantageous – due to the inherent diffusion delays. That theory is based on lower frequencies yet, like 3 Hz or something.
In any event, it has certainly been done in a pinch and it works. Not terribly safe. Not terribly elegant. But yes, a circuit breaker and a bridge rectifier kind of make a charger in its simplest form.
This retest would make a good live streamed event. I would pay to log in and watch the testing live as I am sure most of the other viewers would.
Wait, 500V @ 200A = 100kW. PulsaR can handle up to 150kW… I presume lower voltage means more available current to dump into the pack. On a 36 cell 180AH pack @ 128V (as would be used on a Curtis AC setup) that’s over 780A. That’s a potential of up to 4.3C charging. Only the higher voltage DC motor setups would suffer a 2C or lower charge rate, and that’s assuming the car uses same 180AH cells. A 60 cell pack of 100AH cells would still be in 4C charge range.
You are quite right. The PulsaR is in a buck configuration and can convert higher voltages and lower currents into lower voltages and higher currents. Obviously there is some efficiency loss but yes, we could get most of the 100 kW using PulsaR.
Speaking of dropping CV phase, something I noticed about the 3C test… The end of charge voltage curve seemed to begin much later in the charge than in lower current charging. Makes me wonder if the amount of charge you can add to the pack in CC phase is related to the amount of current being applied… and thus the reason Jack was able to achieve 97% SoC @ 3.65V. What then is the state of charge at 3.65V during a 0.8C charge?
This is great, I’ve been doing this anyway through having to just cut the charger at a set point, I was going to work out a way to do constant voltage now i won’t bother, I know you only sell the CA series but why not test some Sinopoly cells, you could even stock them.
I only sell the CA cells for a reason. I guess we could sell whatever we liked. Sinopoly is now an inferior technology from yesterday marketed by a known loser company. We won’t carry them. And we won’t test them. Moving on…
I’ve always wondered whether holding a higher CV over too long a time would unbalance a bottom balanced pack. However CALB state no cell should be held at over 3.4V for any extended length of time.
No idea if it was true but one pro-BMS propagandist proclaimed (to me) he left a Thundersky at 3.65V with a low output power supply over a weekend and it blew up like a balloon. Boiling a big cell with miliiwatts to me seems far fetched. Maybe it was done one bubble at a time?
You’ve been on here quite some time to have this surprising question. The charger procedure is a PROCEDURE. Leaving out the part about CUTTING OFF THE CHARGE AT 0.05C is like baking a cake and leaving out the part about the FLOUR. Or turning on the oven.
Holding this artificial voltage requires energy which is absorbed by the cell. You are continuing then to CHARGE the cell past its boundaries or limits. That is OVERCHARGING. Overcharging is a no no.
The 3.65 is, and never was, a REAL cell voltage any more than 4.2, 4.0, or 3.8v ever were. These are instructions. A procedure you can follow using values you can measure, to achieve a particular SOC. It is not the open circuit voltage of the cell AT that SOC. We are NOT able to measure that while at the same time adding energy to the cell. Ergo the procedure.
AND THE WINNER IS: 1. Charge at constant current to 3.65v. 2. Hold 3.65v constant voltage. 3. When current level decreases to 0.05C, terminate all charging.
If you do 1 and 2, and forget the 3 part, you WILL blow up a cell and at some point could in fact cause a fire. The RATE at which you do this, and the time it takes, are of course related. At 10 amps soon. At 100 ma, later. But the outcome is the same.
Jack, breathe slowly..
I’ve always stated there is no need to tip top charge any cell. My last post on last weeks blog is confirmation…. and I did make clear Calb’s 3.4V rule in the previous post!!
Been drawing up the CC route to a set pack Voltage for a while but a full suite of family problems have practically stopped me dead in my tracks to the point I’m forgetting where everything is drawn up to.
“At 10 amps soon. At 100 ma, later.”. One bubble at a time 🙂
Apologies if I wasn’t clear enough and made you re-state the obvious. Consider it this way:-
Read a Chinese website which manufactures LiFePo4 pouch cells a while back they said these cells eventually “self balance” over time because of the CV charging in the upper knee. Sounds daft unless we consider charging efficiency drops as after the knee, due to wasted energy heating the cell. This is where I considered a bottom imbalance could occur.
Conversely on the low end.
Some people. (One you know). Has stated he discharged A123 cylindrical cells, the voltage came back up then slowly died off to nothing and could never be brought back. Happenstance upon cooling after internal damage?
Andy, if I’m understanding you correctly, they will indeed self-balance at the top. They always will in as much as the perceived imbalance is only and entirely detected through measuring the voltage before and after the diffusion delay, as it’s been coined. You can measure a few thousandths of a volt difference at the top of a balanced pack while charging. Give it a few hours with minimal or no influence and they will rest at a very consistent voltage. Sounds like this audience is better informed about these cells than the spec sheet author.
My post was not a tutorial on how to charge directly with line voltage, I’m fully aware of what a signal looks like after rectification. The only reason I stated 168v was because I’ve build so many power supplies, that voltage looked so familiar. The voltage would vary depending on the incomming line voltage, since it can vary from 110 to 125 volts depending on the area. I can asure you that the voltage drop would be my last concern considering the amperage available.
It was just an idea in theory of how if one would select the right pack size, it could be charged directly. The amperage would depend on how much higher the voltage would be versus the pack size.
I did not mean to insult you. You are correct in that voltage will determine the current to the pack. I would simply do not feel it was safe to build a system like that without some current limiting device in the mix. Dealing with the inrush current might be an issue as well.
I have charged batteries with a bridge/Cap and variable tap transformer. I usually just turn the volts down to about the pack level and apply power and raise the voltage to get the desired current. Of course, this was with lead acid batteries. As the pack charges the current will naturally drop off a little.
As for the bunny hops I guess I over thought that a little. I built a lot of unregulated power supplied for stepper and servo motor applications. In the control industries I work in Bunny hops = bad result. I guess I tend to think they are always bad. After reading jack’s comments I guess it would kind of be a 60 cycle pulse charger….
I would still prefer a PWM current regulate charger, but I guess it would indeed work…
Changing the subject slightly, a couple of weeks back Allan Høj commented on a Norwegian “study” (http://blog.evtv.me/2012/10/evccon-2012/#comment-7306). This purported to show that EVs had a greater environmental impact than ICE cars. There is a very interesting expose here: http://llewblog.squarespace.com/electric-cars/2012/10/11/the-truth-will-out.html. To pick two random examples, the report assumed a 1000kg motor in the EV. It was also written by a group with strong links to Statoil. Lies and propaganda happen, but I am very embarrassed that the BBC publicised it. As an Englishman it pains me to say it, but it is hard to avoid the conclusion that in publishing it the BBC were either corrupt or incompetent.
BBC also carries Top Gear, which routinely ridicules all things EV. While some say it is meant for entertainment only, so slander is okay, most people in the 20-40 bracket get their news from John Stewart or Stephen Colbert while the more sophisticated among us get their news from Leno or Letterman.
The power lab 8 is a great tool for EV owners, if you cut the boards off and lengthen the wires you can discharge up to 9 cells at a time, it’s also pretty good for fast charging Helicopter batteries.
I would be very hesitant to discharge 9 cells at a time, beyond about 2 amps per cell, using those little wires. On the other hand, we have ordered the little cables to include in our kit because you CAN monitor nine cells for various purposes.
Just to avoid spreading incorrect info, the power lab 8 will charge/discharge 8 cells (hence the name) the power lab 6 will do everything the 8 will do, but for up to 6 cells. They both do 40A charge/discharge etc. The PL8 does have a higher discharge if using the internal resistor instead of regen discharge though and is the only difference for someone processing one cell at a time. It’s also a bit cheaper, you may consider carrying that in the store as well Jack.
I’m using a power lab 6 right now to discharge a 6S3P section of a larger A123 pack at 38A. The sensing wires are for monitoring/balancing only and do not carry more than 1A of charge/discharge current and only if you have balancing turned on, which is pretty useless at the cell sizes we are dealing with.
Jack can I also suggest that you create a preset that you either load onto chargers that you sell or make available for people to download. As you know all the existing ones are for RC car/plane size cells/packs and it can be tricky for a new person to get the settings right.
Well to avoid spreading incorrect info, I think you said it backwards. You can do up to 40 amps USING regen into a battery – NOT with the internal load which is limited to about 10 amps.
The rest of your message is quite on point. I have at least one viewr that has burnt up a 6P and we haven’t tested it, ergo the silence on it. We’ll carry the 8P.
But as soon as I can figure out how to save presets to a file, which it does do, you are quite correct. I need to do some basic presets for the 60, 100, and CA180FI. Different capacities actually cause
different options to pop up on this software which is excellent, but does require a learning curve.
We can do that.
Sorry Jack, that is how I intended to say it. 40A charge/discharge (using regen) and 8A for the PL6 using the internal resistor, or 10A for the PL8.
The PL8 is definatially good, I wish I had one of those right now as I discharge groups of 6 cells measuring the capacity looking for “bad” cells. I have some A123 3P groups that will only hold 32Ah instead of the ~56Ah I should get from a 3P group…. CALB CA here I come!!
And for the record the bad cells are in groups that I wasn’t monitoring with the Cell Log 8, however if I go with CALB’s I won’t feel the need for any monitoring, A123’s are just to unstable/unpredictable.
To save a preset to the computer simply select the preset you want to save as if you were going to run it, then “File” – “Save Preset xx to File”, and to load one just, “File” – “Open File to Preset xx” I believe you can save/restore all the presets in one batch which would make it easier for preparing a PL8 for sale.
If you are indeed able to save the presets to charge and discharge cycle of a common CALB 100AH and 180AH single cell to a preset file, you would make this easier for us!
All the best,
We will make it so.
If you are indeed able to save the presets to charge and discharge cycle of a common CALB 100AH and 180AH single cell to a file, you would once again make this easier for us!
All the best,
One of the nice features about using the sensing wires is for bottom balancing. In monitoring mode it gives you an voltage reading on up to eight cells to 3 decimal places. This makes syncing the cells a lot more convenient.
The article is well written.
It’s just like when the damage is done, it is all too late. Although if the false words would be withdrawn. it doesn’t matter, The vast majority of people have already got it stuck in the brain for good.
The article is well written.
It’s just like when the damage is done, it is all too late. Although if the false words would be withdrawn. it doesn’t matter, The vast majority of people have already got it stuck in the brain for good.
Given this new info, could the PulsaR in future be used to convert 200Amp higher voltage from public fast chargers to 300Amp lower voltage for charging? This would allow reducing charging times up to 33%!
It would be possible to charge a 180Ah string within 36 minutes. Does the PulsaR have the physical components to do this?
I wanted to let you know that at my last count, there are 158 CHAdeMO stations in the United States (source: PlugShare.com).
Also, the dealer discounts that you’ve seen for the Nissan Leaf were part of their annual “Summer Sales Event” and were comparable across their entire product line. A lot of dealers also calculated in the tax credits in those discounts, so that enhanced the perceived discount on the Nissan Leaf vs. other vehicles.
In clusters and not well spaced for optimum use. None close to me. Bummer.
Being a CALB dealer could turn out to be a very good thing indeed, if the latest silicon-graphene anode turns out to offer long cycle life. A poster here
shows that Calbattery is connected to CALB.
@ Warren – interesting. Silicon or tin anodes with that kind of gain in specific energy must be coming soon from somewhere. It will be interesting to see how the specific power, self discharge, temperature tolerance etc works out. Another learning curve!
I saw a web site the other day advertising production of a new battery chemistry. What caught my eye was the CTO was Jay Whitacre. They said they were starting production in 2013. Aquion’s Aqueous Hybrid Ion (AHI) http://www.aquionenergy.com/technology/
There are over 1100 Nissan dealers in the United States. Your plugshare shows very few of the Chademo stations are at Nissans. A grouping in the north east and another group in Dallas/FortWorth. Very few of them are Nissan dealers. The Nissan CHademo in every dealer thing apparently didn’t happen.
I’m not sure the point of the referende as to the Summer Sales event. Nissan is generally leasing their cars for $69 per month???
Oddly enough I am finding difficulty locating the plastic flooring style plastic you are using for spacers at the bottom of battery boxes. Possible to stock in the online store?
All the best,
Search “composite decking” at your major hardware retailer.
No magic to it. It is just plastic porch decking. The particular kind we use is made by Weyerhauser and it is called ColorFast Decking. Lowes carries it.
But TREX is probably the largest manufacturer of plastic decking. It is relatively cheap and we use it to space things. It quiets and insulates. It is virtually indestructible. So we use it where you might a piece of wood.
Many thanks, got it!
All the best,
Not sure how the new CA series batteys respond to initial charging but…. I recieved my Powerlab8 and went into the accurate charging profile and set charging to 3.5v @ 20A and the 100AH cell got to 122F anode temp and I could hear faint audible gasing in the battery so I stopped. Any ideas?
All the best,
PowerLab 8 Preset file for a 30amp charge 30amp discharge. It is set for CC only for both charge and discharge. I have been using it for charging and discharging one of my A123 cells. At 30 amps the little cables do get mighty hot. It may be better to do this with a separate CC power source. If you do be sure your battery is fully charged and is a deep cycle setup. I am using 4 100 AH High Power cells and during my last charge cycle it went below 10 volts. You need to be sure your pack is fully charged. One little Golf cart battery may not be enough. I’d take about 4 and put them in parallel and then use that for your charge/discharge pack. I have the file set to shut off if the source voltage reaches 10 volts so with my cells that is 2.5 volts each. Pretty much depleted. I now must recharge the battery . I did use it for some time in the Bug too and have only charged it once when I first put it together.
You must set your settings for battery power deep cycle.
My testing within the safe range does 18.5ah into the cells but if I truly use them in a safe range I’d consider these pouch cells truly 15ah cells. If used that way these suckers are pretty nice. However I do really like the long flat curve both charge and discharge of the CA cells and with the C rate that can be pulled when needed these are just perfect for street and I’d bet for short trips down the track.
So here is the link for the PowerLab 8 preset. Give it a try and let me know how you like it. email@example.com
Thank you for everyones help. I have seen the errors of my ways. It was the “High capacity” preset, I think number 3 that I was using. I see that had a charge voltage of 4.2v. I hope I did’nt damage to much. A cycle test is in order now. I should have been playing with a SLA battery!
All the best,
@Aaron: I don’t know the Powerlab 8 so this might be a silly question: you didn’t by any chance pick a lead-acid profile that attempted to trickle charge rather than terminate at 5 amps or whatever?
You should be working with an A123 profile, you can restore or copy any of the original presets from the library section and paste them into one of the first 25 presets in the menu. Modify the preset in the first 25 preset section.
You probably want to change the descriptive text in the preset so that when you are running the Powerlab the preset number you are actually running is the one you intend to run. You can modify and update the presets but another preset number could be selected to be run on the Powerlab so check that on the Powerlabs LCD display.
I think that cell is OK. I removed 98.55AH from a 3.5v charge to 2.5v. I will let it rest and try charging to 3.5v and then 3.65v to verify.
All the best,
Inspiration on how an electric motor should BE torque wise,
1000NM per ton
2,000 miles on one tank
70 mph at 1000 rpm
“its better to have a big engine that barely works then an efficient small engine that works so hard”
all the best,
Got my AC50 three phase motor running on the bench with the 38 CALB grey cells today. One of the lads (my son’s contemporaries) was a bit rough with the accelerator and came fairly close to knocking over the bench to which the motor was attached. On reflection it might not have been a brilliant idea: if it had gone to full power due to a wiring fault it would probably have gone through the garage wall.
It was useful though – it sure helps understand what all that mess of wires is for
Imagine the sight – an AC50 running down the street followed by 38 grey cells.
In other news GWL Power has been doing some destructive testing on LiFePO4 cells. Unfortunately they don’t for specify which cell they tested. Quite bizarre. http://gwl-power.tumblr.com/post/34903913779/lifepo4-cell-test-data-abuse-and-extreme
That’s very interesting Jarkko: thank you for pointing it out. I wonder if they were old stock of Thunderskys (I think they were said to have Yttrium in them)
hey Jack and friends,
I been watching for couple years now but never commented, I was wondering how can you protect your lowest capacity cell from topping out while charging at 3c??
if you get such a high temperature rise getting the last few aH in the cell could you see a time where the lowest cell getting over charged?
A side note: I got a few of your A123 cells (thank you) and read that the edge of the cell has exposed aluminum and can short out your cell if allowed to touch the battery box, so I checked the voltage along side of the cell and get .5 volt from the anode and cathode,
I don’t understand where you could get the same voltage difference from both sides, and could see a potential for a problem if all these sides could get in touch with a metal of the battery box.
I would like to see a test to see if this would be a big deal or not.
but for everybody using A123 cell I would suggest putting strip of duck tape folded over the side of each cell were the exposed aluminum heat spreader of the cell is.
A great show this week I enjoyed it very much.
More hands on and voice over in the background about the topics while showing work being done 😉