A pleasant week in Missouri and for EVTV. The temperature has dropped dramatically to low 80’s from well over a hundred degrees fahrenheit and we can work again in the shop.
I took a day to fly the King Air 200 with my old mentor Kenny Hoffman to pick up some friends undergoing medical treatment there. This puts a point on how every deeply I resent the total nonsense and mess that our government has made of our commercial air carriers and airports. They are essentially unusable for anyone but particularly for the elderly or the sick. And they accomplish nothing. Nothing is even marginally safer. They can’t make you safe. They can only make you take off your shoes. This is our government in action.
The level of critical thinking in our government, and perhaps in our country, has reached what I term the MORONIC PLAGUE level. It’s actually pandemic, incurable, and undoubtedly terminal.
In any event, we shot little video but it didn’t matter. I had to spend Friday morning studying hard for our show, which I really never do. Then Brain and I went over it for 30 minutes before shooting. EVTV just got much more complicated to shoot.
[jwplayer file=”news081712-iPhone.m4v” hd.file=”news081712-1280.mov” image=”http://media3.ev-tv.me/news081712.jpg” streamer=”rtmp://s3einxnpkaij93.cloudfront.net/cfx/st/” provider=”rtmp” width=”950″ height=”584″ html5_file=”http://media3.ev-tv.me/news081712-1280.mov”]
But its all for a good cause. I’ve reached this transition point once before, with Boardwatch Magazine, and it is an exciting transition. This time I know what it means.
I started a little newsletter in 1986 about electronic bulletin boards, online services, and a network that actually did not have a name at the time and there was a very active debate as to what to call it. John Quartermain was doing very strange graphs of this network of networks and we began calling it the “internet” while very eagerly watching to see what it was finally going to be called. Oh well.
But in the early days, I wrote every word in the newsletter. It’s actually a bit of work. But I soldiered on.
At some point, some of our readers caught on to what I was doing, and wanted to do it too. But they couldn’t really see how to start a magazine and make it pay, especially since I was pretty much starving to death at the time. So I taught a couple of readers how to write an article and what was important and what wasn’t. They picked up on it pretty good, and I had additional material for the publication. It wasn’t as good as what I wrote, bot sometimes it dug up things I might have missed. And I kept after it.
At some point, there was a change in the publication. We started getting articles, that while short on literary merit and unlikely to win an award, were nonetheless from people who WOULD know and about things we were interested in and did matter. Some of the stuff was much BETTER than what I was doing. One of my greatest thrills was hiring my journalistic hero, John Dvorak. He had literally taken an interest in Boardwatch and mentioned it in several publication giants of the time which gave us much exposure. So at one point I approached him to do the back page of Boardwatch just as he had once done when I read him in every issue of Inforworld. Incredibly he agreed. At one event he participated in a Chili Cook-off and we learned that he was something of a gourmet cook. So I asked him for his chili recipe and dropped it in with his column in the next issue. It got so much response, he started including a recipe with every column. We were the only computer related magazine on earth with a regular cooking recipe included.
By the time I sold the publication to Mecklermedia and eventually to Penton Media, we had 22 regular columnists and two copy editors. And you basically couldn’t hook up a coffee put cam to the Intenet without it literally showing up in Boardwatch.
Why the trip down memory lane. We’ve published other people’s video walk arounds for some time in EVTV. But we’ve hit kind of a transition in quality and quantity. We can’t possibly run all we’re getting in, and the quality as exhibited in Marcus REddish’s submission on the SUNJAG in this episode is just phenomenal. We also ran some GOPRO footage of Monster Tajima’s run up Pikes Peak as well as some from Michael Bream of EVWest, who DID finish in is series DC powered BMW M3.
Robin Wainright does not have to worry about a video career, but it was very informative to see what he’s done with a very light and inexpensive A123 pack and I know a number of viewers are dabbling in that area. Take care, we show the results of inattention to these cells in this episode as well and not to be outdone, we have a wee bit of film of a Fisker Fry which uses the same cells, although the batteries don’t seem to be the culprit in this one.
The result is kind of a different EVTV – more a survey overview of our viewers and what they are doing out there. As there are hundreds of projects going on at any one time, this can only ever be a representative slice. But I happen to know if there is a ruthless editor in play, these things can be selected to fairly represent the state of an infant industry without entirely being one guys view of it.
And so in this episode, we kind of show what EVTV will hopefully become, a mirror in which you see yourselves, or better yet, each other, in a useful way. It has been my experience that such a mirror can dramatically accelerate the growth and maturation of such an infant industry, by fostering connections that would otherwise be left to a more random chance. I do think the Internet itself has made it easier to et in touch with each other. But the noise level is discouraging and the vastness of it kind of returns the same problem only louder. There has to be some asshole in charge of picking through the noise and editorially selecting that which might be interesting for a particular audience. IN all modesty, I’m pretty good at both of the important factors therein.
The very exciting and fascinating thing about the Boardwatch ride was that in building this “lens” to view the industry through, we found that we, and by extension our readers, had a framework for imagining the future that was uncannily accurate. With a few singular exceptions, it didn’t have a very long range – six months, a year perhaps. And indeed some of our prognostications are just now coming into fruition. But from one issue to the next it was just ENOUGH of an edge that in seeking an entrepreneurial slot in Internet space, our readers did pretty well, and in some cases VERY well, even if Obama thinks they just got lucky while using his roads.
It can take years of effort to grind a mirror for a large telescope – it can take months for it just to cool after pouring it. But when you get done, you can see further than you could before.
For most of the world, we are a tiny band of people going into their own garages and making an electric car for our own personal use. I’m ok with that view because it makes us kind of nonthreatening and not something to be made safe from – like airports and airlines. We’re under the radar.
But I happen to know that almost everyone who’s built an electric car has in the back of their mind started to puzzle over the concept that there are others like them, and that they might be able to parlay their hard won knowledge into just a little sideline entrepreneurial thing to pick up some pin money with. Just to offset some battery costs – you know. I cannot tell you how exciting this is for someone old enough and with knees and eyes bad enough to know what it means. It is America’s wealth engine at work. The bubbling petri dish of innovation that cannot be stopped by anyone or anything on the planet. You can pour cold water on it in the form of taxes and regulations but Americans have no compunction about avoiding taxes and ignoring regulations until it is too late. The Genie gets out of the bottle.
The early days of bullet boards and the Internet was exactly like that. ANd the participants weren’t really sure it was “ok” to make money from their sacred hobby. A guy named Phil Becker and I stood up together and at a Fidocon meeting told them HELL YES iT’S OK – it is AMERICA and it was like setting off a bomb. They were all standing around wanting to ask someone for permission. If you need it, you have my permission. I happen to know you don’t need it. But you have it anyway. it is quite alright to start a itty bitty business selling bumper stickers for EV’s if you like, and wind up building more cars than Tesla in the out years. You need feel no guilt. You can invent devices, services, and all manner of related products with no ones permission and either wind up bankrupt or a multimillionaire many times over. And some of you will do each, and a FEW of you will actually do BOTH along the way. In either order. I cannot promise outcomes – only opportunity.
If it comes to civil disobedience in places, I’m frankly a revolutionary. The treason is in Washington DC. YOU are the very best of America. And our only hope for continuing to lead the world in this manner. It is further ok for our many viewers OUT of the U.S. to adopt the same behavior. It only works 100% of the time.
We remain early in the process. Charlie Rickman is doing T-shirts. David Hrivnak is trying to get two J1772 connectors to marry and have offspring. (They’re both male plugs David – I’m sorry to say – state sanctioned electrical mating is reserved for ONE male and ONE female connector). But I was there when Cisco Systehms was three guys and Apple Computer was two guys hoping to recoup their circuit board costs and indeed when Microsoft was two guys selling a small roll of punched paper tape in baggies. It is not part of the corporate gestaldt to remind you of humble beginnings. But they all were comically humble. The innovators all had one thing in common, that wasn’t what they were trying to do and they never dreamed it would take off like it did.
So I see things you may dismiss as overly grandiose. Particularly in a world of General Motors. You weren’t there for AT&T. Talk about a monolith with a lockup on everything. It was the LAW. GM and Nissan are bumbling comedy pieces by contrast. Further, there was little perceived need of an Internet. ALmost everyone in the world is vaguely aware that we need to DO something with regards to alternatives in energy. You have widespread built-in acceptance this time, they are just heroically vague on what it should look like.
Should we succeed in building our lens, in erecting our telescope, and the mass of engaged viewers do start unleashing their entrepreneurial potential, this very early stage movement will grow. Slowly at first, and then more rapidly. Until like our viewer submissions in articles or video, it eclipse our efforts and simply runs away with us and becomes a thing too vast to do. Along the way THOUSANDS of things will be invented and ultimately abandoned. But some will grow phenomenally and become part of the American legend of innovation and wealth creation.
For me, it is almost unfathomable. To get a SECOND ride in a SECOND rodeo, and to be so much better equipped for this one, albeit with some physical limitations and fragilities that come too with age. Nabile Henke doesn’t have a pot to piss in or a window to throw it out of, but he’s off and started doing a conversion for someone else in Iowa. I listened to Matt Hauber explain to me for SIX MONTHS why he COULDN’T go to San Diego and start a conversion shop. He didn’t have any money and he didn’t have this and he didn’t have that. I literally had to scold him into submission and send him there myself. I did give him $1000 for a broken down Siemens motor. He was supposed to send me the pinout on the encoder. He never did.
This month a car he built with his own hands, pretty much on his own and with absolutely NO background in electronics or electricity at all, climbed 12.42 miles and a mile vertical up Pikes Peak in 11:58 leaving the driver surrounded by veterans of DECADES of competition all poking him for how he did that. And the man who had won 11 previous times standing there with a badly burnt up $3 million dollar electric car with 27 on-site engineers and the very latest EVO motors and Rhinehardt controllers – generously congratulating him on his success. These guys haven’t a CLUE what they’ve done. And don’t anybody tell them. They need to do more of it and if they DID know they would be paralyzed like deer in the headlights from then on.
With a cunningly modified fork lift motor and a controller produced out of thin air six months ago by some guys who’ve only met on an Internet forum. Of which, did I mention, 10 working copies of exist in the world. They THINK it will work. Lest we put this in context…. And yes, I know they are going to build 10 more…if they can get orders for them….
We are at an inflection point – a moment in history. The birth of an entire industry. It goes quite beyond Tesla and OEM electric cars. I would posit that the electric car has not even been invented yet. We have pieces of electric cars. And the birth of an entirely new industry of batteries and components and products and services – a whole rich panoply of innovation and disruptive technology that will eventually be of enormous size and scope. And among our viewers RIGHT NOW are the giants of this enormous world wide movement. Yeah, I’m a little excited.
I know a secret. And I can tell it entirely openly with calm assurance it will remain a secret because no one will believe it. And anyway it will be drowned out in the noise of the MORONIC PLAGUE that decimated America in the early years of the 21st century. But it is very exciting for me personally.
Send video. Make BETTER video. Like Sunjag. Show us what you do. Be fearless with your secrets. If you’ll help me build a lens, I’ll show you how to know what’s happening 4 to 6 months in the future. That’s all the edge you’ll need…
And yes, it’s ok to get excited and it’s ok to make money and its even ok to compete in the Pikes Peak Run. And you don’t need permission…
89 thoughts on “THE INFLECTION POINT – YOU DON’T NEED PERMISSION”
I converted the car in 2001 into lead acid and the car has been running maximum 50 km when I was in luck and only short trips. A trip over 15 km and I would not get home even at full charge.
It is like Jack described a science project (without a happy ending)
If it hasn’t been for Jack I would not be using my electric any anymore, he opened my eyes for the new lithium cell, and I am overwhelmed and Ecstatic happy with the way the cell works I am able to run over 180 km and I still got 12 AH left.
I have 42 Cells in serie 160 AH Thunder Skys. The cells are balance out at 2.65 Volt per cell. I think that it is the 6 time i am charting the cells, and i Know that they are increasing in AH the first 50 times, they have now reach 192.4 AH.
Question : When I am discharging my cells and I am approaching the cliff at a voltage of 121.8 V is that measures at “full” load 160 Amp or is measures at standby ?
As it is for now I am using the cars original charger it giving out 12 – 13 Amp. It takes approximately 14 to 15 hour to chart.
I bought a charter do-it-yourself kit at EMW. But have some trouble with the documentation or let’s say lack of documentation. But I’m sure that we going to solve the problem in the near future.
I’m looking forward to reduce the charging time with 10 I hour.
Jack and Brian I wish you the best !!
PS : what about a Fire Alarm ?
But thanks for showing what happens when everything goes wrong !
Best Regards Allan
The voltage should be measured as you say in “standby” – while at a stoplight for instance with no load on the batteries. If it is below about 125 volts, you are pretty much done for the day.
We recommend counting Ampere Hours or kiloWatt Hours as a more accurate way of measuring your usage.
2nd the question about the fire alarm. I would consider that essential for a lab manned by two mad scientists with too many amps on their hands.
Just think how close you came to a bad Karma event…
Sorry but I’m fuzzy around exactly what happened to your 12V A123 battery. I read from what you say you overcharged it but disconnected it 6 days before it went thermal?
If so what Voltage(s) was it taken to and what was this “BMS” thing?
Nobuhiro ‘Monster’ Tajima is very well know and liked in New Zealand and is notably famous here for his domination of the “Race to the Sky’ event in Queenstown, NZ. This event was run for 10 years from 1998 – 2007 on the gravel road from Cardrona Valley to the peak, a 9 mile, 3500 ft climb of 135 turns. Tajima won 8 of the 10 events but sadly this event no longer runs due to lack of funding and organiser enthusiasm.
Nobuhiro is also a member and supporter of APEV which is an organisation based Japan and New Zealand formed to promote and champion electric vehicle uptake, with a leaning towards commercialisation of conversions rather than sales of OEM vehicles.
I am sure that, had he made it to the top, Monster would have whipped them all in an electric car, but for now congratulations to Rhys Millen from NEW ZEALAND reclaiming the glory for his dad Rod Millen who held the record from 1994 – 2007!
Very neat idea about cooling the commutator with the shaft. Tool and die makers often use beryllium copper inserts for thin areas in water-cooled plastic injection mold tools to avoid hot spots. The stuff is very strong and conducts heat far better than steel. I wonder if it would be feasible to either make an entire motor shaft out of this material, or at least splice in a piece at the back end.
For race use what’s wrong with simply spraying distilled water on the commutator/brush/permanent magnet area? Don’t forget, TWO isolating contactors.
Did that A123 battery self short? I’m worried.
Motenergy just advised me not to have water spraying on the ME1003 motor. I’ll have to build my own BlowMeTM solution.
They would do. Can’t blame them whether they know what they are talking about or not.. This is something I would do on racing. *Not* road use even though it will suck in rain water anyway. You need air blowing through.. A lot of air!
I would recommend you have your rotor turned to match your bearing axle to ensure full width brush contact.
The problem with spraying water on the comm is that it washes away the film you worked so hard to deposit. That film acts as a lubricant for the brushes. I got to watch a few RC car races. In one of the race classes at that time the motor was provided as part of what you got for your entry fee. You have to race with that motor. This was to try to keep people from out spending the competition and making the races more a matter of skill. Anyway, after soldering on a connector to the motor almost everyone grabbed a battery and ran to the bathroom to seat the brushes. They did this by spinning up the motor and dipping it in the toilet for perhaps as much as a minute. And I got one guy to show me a before and after and yes the brushes were well seated after and just a V cut end before. The water in the bowl turned black in just a few seconds.
So I am guessing you could water cool your comm briefly but you might have to change brushes after every race. And I have no idea what it would do to the comm surface.
Not for road use. Will work well for racing.
Don’t forget, they use an extremely soft brush for racing.
Thanks for all the Pikes Peak footage. I didn’t realize Tajima-san was running Evo axial flux motors, though I suspected it was some kind of permanent magnet rig when it failed.
Interestingly, the winning electric from Toyota Motorsport also used EVO motors.
According to Hauber, EVOs were everywhere. They were the ONLY car with a series DC motor. They had way more motor than they needed, and needed a little bit more cooling. They were really only running about 1100 amps out of their 3000 amp Shiva. But were pegged at 350 degrees F through the entire second half of the climb.
We’re going to make a run at a “chill plate” for the Netgain Warp 11. Lucien should be in today to talk about it.
You might look at something like this placed in the motor shaft:
I agree. This would be a good way of bringing the heat out to the end of the shaft where the cooling “pump” has water passing over the shaft. We use these for cooling control panels and they are amazing in how fast they can transfer heat. At trade shows they typically have a demonstration with a 18″ rod sitting in hot/warm water and then ask you to grab hold and then pull it out of the hot water and place it in the ice water bath. Your hand is a foot from the ice and yet the end of the rod in your hand goes ice cold in just a split second.
This assumes that the shaft is capable of having a hole bored without sacrificing strength.
I don’t follow. For the heat to get from the commutator segments to the shaft it’s got to flow through the insulating material isolating the segments from one another and from the shaft. Is the insulating material really that good at transferring heat? Most aren’t.
Heat pipes might be a good way of dumping heat from an outrunner. Take several of them and embed one end in the windings and poke the other finned end out into the airstream or use a fan like some of the impressive heatsinks they sell for overclocked CPUs.
Did you tell Heyoka Merrifield of the dangers of his BMS? I am new to EV’s. I found your site while looking for parts for my conversion. I am converting a 1986 Toyota Celica. A friend of mine worked up a spread sheet of expected performance in Excel if you would like to see it. I also have an idea about having a separate current limiter on the stator field to keep the motor from going into saturation. I was wondering if you knew anyone who tried that or if it would even help. Seems to me like one could get a much greater range on a Series wound DC motor by controlling the stator and armature separately.
We have about 140 videos up, many of which deal with the topic. Mr. Merrifield seems curiously satisfied with his car and as it is HIS car I’m not going to tell him anything.
Some lessons are only acquired first hand. I try to share mine so others need not. But it isn’t a religion. Again, it is YOUR car and YOUR house, burn both to the ground at whim. It’s all good. You can build a NEW house and a NEW car. Both will probably be better at that.
“by controlling the stator and armature separately”. That’s called a Sepex motor. Useful if you want regen and other little things.
Motor current is limited by setting your controller. I’m not sure anyone has ever reached saturation limit because of the back EMF.
Re battery fires, I’m planning to use enclosed stainless steel battery boxes. I haven’t the slightest idea whether they would contain, or even slow down, a battery fire: but if the main culprit is burning ethylene carbonate, starving it of oxygen may help. Alumin(i)um [melting point ~660 degrees C] doesn’t really cut it as a firewall. Does anyone have any experience relevant to this?
ethylene carbonate is anti-freeze and coca cola. It’s not even dangerous.
The issue is when the cathode becomes hot enough to go into thermal runaway and the point at which it begins to release its own oxygen – feeding the fire.
The solvents are potentiallly flammable at a very low level. But about like an alcohol burner.
Stainless holds most of it’s strength up to about 1200-1500F (Depending on the alloy) So your box would do better than aluminum. However, I doubt it could contain a full on battery fire.
A fire suppression chemical dump would probably work better. I have used these on race cars to dump the chemicals on the fuel cell and engine compartment. They could probably be modified to work inside the battery boxes…
The OEM EV has $2,500-7,500 tax credit from the federal government and other tax incentives from state governments. How can we get a level play field for aftermarket EV conversion industry?
You may want to check your state. In the state of GA you can get a $5000.00 credit for a conversion. There are some rules like it must be able to go 55mph, but they are fairly simple…
Simple. Just buy a few senators. That’s all it takes to direct the handouts your way. I’m sure GM has more than most. Maybe you could ask them for some tips?
I’m flattered to be mentioned Jack. I just wanted to insist that no matter the amount of work I put forth, I know there is much more to do. Indeed, last night we confirmed with our client that we are going to move forward; my dad is planning the trip down to Morgan Oak this weekend to pick up the cells. You mention low hanging fruit, there is none lower than a garage with a dead lead sled parked into the garage by a Leaf! With that in mind, we consider it an uphill journey to our next several clients.
I recorded the entire convention last year on my phone and have repeatedly listened to the wisdom shared by Eric Kriss and George Hamstra, in particular. Two key points: inevitability and trust. Inevitably ev drive trains will be price competitive on the sales lot. And trust in your own perseverance to see your vision through.
The development of the Internet was not at all inevitable. It was driven by desire and the picture of a global network had sufficient allure that many were attracted to the notion and it became desire “in common” among a growing number of adherents.
I see this as very similar. The concept of simply making need for energy disappear by using the same energy six times more efficiently to get from A to B is enormously attractive. If you can do it not only with no suffering, but with exhilaration, it becomes more alluring yet.
As more are attracted to the notion, the desire in common grows and the industry and technology will do whatever necessary to satisfy that desire.
People forget that the early days of such networking involved icky bulletin boards, and that the operators could easily wind up with $20,000 tied up in a little 8 line BBS that by today’s standards really didn’t do very much and it didn’t do it very well. It was a painful and expensive journey. But they WANTED it.
All that we have today by way of Internet is the result of that desire and those expenditures. AT&T, Microsoft, and COMCAST did not do this. Individuals did and they did it in SPITE of those entities who at that time very very violently did NOT want it done at all. In 1996 Bill Gates emphatically declared that the Internet was not sufficiently robust for business applications and Microsoft would never participate. Six months later he had a grand developers conference with everyone who was anyone on the Internet coming to kiss the ring and declare for the Microsoft vision of the Internet, which all in all was pretty lame.
That led to our BILLGATUS OF BORG Cover issue in May of 1996 – reprinted as a poster to several hundred thousand copies at $19.95 . Thanks Bill.
If I have to make a video like Marcus REddish I do not think I’ll make the cut. That was a true visual/audio treat.
I have really been working hard on the EVThing lately. I installed the 3:44 Trans and assembled the motor today (after fixing the flywheel run out). I am building the rear battery boxes. I did not video the work as I was by my self and I am desperately trying to get this done by EVCCON….
As for a Pikes Peak hill climber, that very light wide body speedster would be nearly ideal…. ( Way to go Matt!!)
P.S. Any word on the Chargers? I hope that I will be needing one soon…
The SUNJAG was of course professionally edited afterwords. Most of our submissions are not. We do some basic editing but don’t make any attempt at art.
I think the important thing is to see what you are doing. Whether it is easy or difficult and how the result turns out. Others can then gage their plans on their assessment of their ability to perform such acts of gravity defiance.
Jack I thoroughly enjoyed the video of the Pikes Peak Hill Climb along with all your other incredible and entertaining info. When I was in the Air Force, I was stationed at Peterson AFB and traveled a number of times to the top of Pikes Peak. Pikes Peak was nicely framed in my apartment living room window. Even though I lived there for about 3 years, I never adjusted to the altitude being from Florida and each time I went to the top of Pikes Peak I got altitude sickness,LOL. But I enjoyed seeing how I could get to the top with my parents riding along yelling in my year,”slow down, you’re going too fast”. Thanks for the memories Jack & Brian. Never a dull moment at EVTV!
I was scared to death driving up the hill at 30 mph. It was all loose gravel then though.
Yea Jack, it was loose gravel much of the way for me too then. I hope to take my daughter back to Colorado Springs,but IF I do, not so sure I’d make the trip back up to the top again,LOL. Loved my time there back then.
An observation came up elsewhere that if you have a large enough pack and you charge it to say 3.5 volts per cell you end up with several volts of headroom for a cell to go way over it’s proper voltage. Let’s say a 40 series pack you’d be charging to 140 volts. (3.5*40)-(3.4*40)= 4 volts. A single cell could shoot up to 7.4 volts before charger cuts off couldn’t it? That might be enough for it to get out of hand.
A solution I thought of would be to use twin chargers for charging half a pack each. Could this work or would it be an extremely bad idea? It would also allow to use two plugs behind separate fuses fo faster charging. On a bad side you’d need to wait both chargers to get to full charge before cutting off or otherwise you wouldn’t know if your pack was balanced or not.
And thinking about for a bit more you still wouldn’t know if was balanced or not because the other half might have better cells than the other one. The first question is still valid I think.
Actually no. Why did this cell just magically shoot to 7.4 volts?
We balance them at the bottom. And then charge them in a sting. Each cell gets exactly the same current. They are in series. Now how did one of them go to 7.4 volts?
The BMS guys always get this extreme concept. The batteries just don’t vary that much. They START to vary a bit at about 3.5v. Bu t it is in hundredths or perhaps tenths of a volt. 3.65 at the top and 3.45 at the bottom. That would be an EXTREME example.
Under WHAT scenario do you get a cell to 7.4V?
You will never see it. But even conceptually, what is the THEORY you are proposing.
The worst I have seen even with my old Hi-Powers was when I was charging to 3.65 and I had a cell that would go just into the 3.8 volt range at the end of charge. All settled to 3.33/3.34 volts per cell.
Sounds like the question; How do you know the thermos flask is intelligent and runs on perpetual motion? It knows the temperature and maintains it there, that’s why!
Jarko, this wild theory never happened on any one of John Hardy’s 350’odd charge cycles.
Can you imagine a series string of capacitors being charged, all the same rating. The voltage does not divide down equally so the last one has more voltage across it? Same with resistors and the Voltage drop. It simply does not happen.
Seems to be a business ruse, (fraud) from some BMS proponent to sell his wares to some mug. Check him out. Find if I’m right.
Indeed this was proposed to me as a fault in bottom balancing system. It was submitted by a BMS advocate. He claimed that this could happen. Not sure if magic was involved.
Yes I didn’t see any voltage swings of that order of magnitude until cell 6 failed. Here is the last 20 days of cell 6:
This was 8 cells in series charged to 3.65 volts at anything up to 2C. Perhaps you might ask your informant for data to back up his 7.4 volt assertions
Sorry, but that sounds terrible. Using two chargers will only guarantee that your pack CANNOT stay perfectly balanced because the two halves are being charged separately, and not identically. They may be really close, but it can’t be exactly the same amount of power as they would get with a single charger.
If you just want to charge faster you would want to parallel your chargers, so all the cells would get the same amount, whether from 1 or 2 chargers.
He was actually suggesting 12 volt cell voltage. The 7.4 V was mine and as such even lower than he claimed could happen. I don’t think he actually has had anything like it happen which usually is the case.
For anyone interested in watching paint dry or batteries tested, 8 more Headways should be with me next week. I plan to shove them onto the rig as an 8 cell pack, all bottom balanced but with cell balancers on half the pack, then monitor voltages as before.
8 CALB 40’s destined for the rig are on their way from China by sea along with a set of bigger cells to go in a car
I just installed a 240V circuit in the garage last weekend and have been watching the end of charge behavior of my pack the last two charge cycles just to convince myself everything is still behaving as expected after switching from charging at 120V to 240V. All is well but I did notice something interesting.
First some details: 36 CALB LiFePO 130 Ah pack,
ElCon 2500w 120/240v
Charge profile: CC to 124.1V (3.45v/cell) CV cutoff at 5A
BMS free, measurements taken with Agilent 50,000 count DMM
It’s been a couple of months since I checked voltages at the end of charge but I have been doing this every 2-3 months just to convince myself no batteries voltages are getting too high at the end of a charge. Anyway last time I did this I marked 8 batteries that climbed in voltage to around 3.55V or above before charging terminated. Actually these 8 batteries ranged from 3.54 to 3.58 V. In my book this is great. Plenty of batteries climbing the sharp end of the charge curve ends up protecting any one battery from getting overcharged.
One interesting note though is that my 36 batteries come from apparently 3 different manufacturing lots. One crate of 16 cells came marked “F”, one crate marked “B6”, and one box of 6 cells marked “B17”. All 8 cells I marked as “hot” at end of charge were from group “F”.
So while measuring voltage during a 240V charge just before the transition from CC to CV I noticed 3 cells climb above 3.5V. I of course starting watching these cells more closely and to my surprise after the transition to CV phase the voltage of these cells started slowly dropping. Now it only takes about 7 minutes from this transition point before the current falls to 5A and charging terminates but during that time these 3 cells all kept slowly dropping in voltage to around 3.53 when charging terminated. Interestingly the 8 marked cells all were around 3.45V at the transition from CC to CV phase and all had climbed to between 3.54 to 3.59V at charge termination. A final note is that the 3 cells where voltage was slightly decreasing during the CV phase were from the “B6” group.
I’m not sure what this all means but it just seems curious to me. Do you think it’s worth the time to investigate further? My thought is that the only thing I’m going to discover with more investigation is possibly a batch effect or simply capacity differences between these cells. That certainly wouldn’t be new news to anyone.
This is an excellent observation and is a normal behaviour of cell voltage under charge. Different cells due to manufacturing differences will behave differently.
There is more than one factor involved in the charging voltage you are seeing – and is why you cannot use an active BMS to top balance your pack.
Charging voltage != SOC.
All that matters is the static voltage after charge. The charging voltage (which everyone seems to get hanged up on) is almost meaningless.
Love to hear more about this topic from Jack or John – what I get from this is that you need to figure out a charging “profile” that will charge your pack to a specific static voltage. Where from testing you know this voltage represents the SOC you want to charge to (perhaps 95%). This might be CC 3.5v CV terminate C / 20. It might be something else.
You are actually all over it and in spades. I keep saying this but I have apparently NOT found a way to convey it in understandable form.
The CHARGE VOLTAGE is an ILLUSION. This is a RECIPE, a procedure, to get to a static open circuit voltage and SOC of somewhere around 3.34 or 3.35 volts to my desire, or 3.400000 volta maximum, 24 HOURS after you stop charging.
And the recipe is to charge to your selected charge voltage, and then hold that voltage until current drops to 0.05C. That last part is important. Part of the procedure. A less important part, but important. It is a way to approximate energy and SOC by what IS actually observable. But the charge voltage never was the voltage of the cell, or really accurately representative of SOC. It’ just a recipe to get there.
The divergence observed is a function of diffusion delay. As you start to decrease current across the entire pack, different cells have different diffusion coefficients. As ions find their rightful place, and at this point have to tunnel hard to find them, the voltage decreases. This is kind of a function of the individual “grain” or porosity of the anode material and the SEI layer.
After charging, you will not the voltage will fall. If you observe that, they will also fall at differing rates and I don’t really connect that to SOC either. It’s just diffusion. New cells will sometimes “stick” because of a young SEI layer. After a few charges, they improve in dropping to static voltage and it happens quicker.
Just the nature of the cells. But after you observe some of this for months or years, you can see why I view top balancing as a panacea for child minds. They are reading messages from god in cloud formations and are as likely to be unbalancing packs as balancing them. Rich Rudman notes that WITH his Rudman Regulators, he still bleeds off a lot of current with each charge requiring cooling and points to this as proof that they need to be balanced every time. He’s wearing his batteries out with all of this.
You don’t need to. Undercharge them slightly – just as described here. We charge 36 cells to 126.6 volts daily. But at 124, your cells will last LONGER. If you back it down to 122, they will last longer yet. But at some point, you’re not getting the range you paid for. So it’s a trade off.
Charge to 51% and discharge to 49% SOC and I’m convinced these cells would work for 1200 years.
I’ve noticed the same behavior when I _used_ to charge and balance to 4.00V. Thank you, Jack, for getting me straightened out and getting me to play with a few cells. I’ve been trying to get across to people that it is a charging PROCEDURE and nothing else. I see more and more people starting to say the same thing. I think the difficulty is that some people “know” how these cells work and are in the camp of “My mind is made up, don’t confuse me with the facts!”
For those of us with chargers without the smarts to be programmed to stop at 0.05C just lower the ending voltage. Right now I’m charging to 3.455V because my Zivan isn’t smart enough. I still may lower that a little after I get a little more testing done on a 24hr rest after charge.
Thank you for posting that. Nice work. It is good to have that confirmed in a real pack because I observed exactly the same behaviour repeatedly on the test rig – in the CV phase some cell voltages rose and some dropped. You can see this in the video segment that Jack kindly webcast a couple of weeks back.
I think what is happening is this: every cell needs a slightly different current profile to maintain constant voltage in the CV phase. The charger of course can only work on the cell averages so inevitably the current profile is too aggressive for some cells and not aggressive enough for others.
I agree with you that slapping simple top balancing onto this sort of phenomenon is a little naive
Oh and one other thing: the effect you describe was (from memory) more marked at higher charge rates which would explain why you noticed it after switching to a more powerful charger.
I don’t know how much it is to be concerned about. I saw voltage differences greater than you did, but still well under half a volt.
This batch number business may be highly relevant to A123’s self conflagration properties.
Totally agree John. This was a phenomenon not observed at lower current charge rates. It only was observerable when increasing the charge current (via increasing the charge power applied from 1500W to 2400W).
Jack and John thanks for speaking about diffusion rates. It is an interesting phenomenon to observe but I wholeheartedly agree is doesn’t mean anything in the greater scheme of things.
I like the fact that I have cells in my pack from different batches with different absolute capacities, internal resistances, processing differences, etc. THIS IS WHAT YOU SHOULD EXPECT IN ANY INDUSTRIAL PRODUCT and at the end of the day in any decent product all of those differences shouldn’t matter. It only exemplifies the robustness and reliability of this chemistry and the cells produced by CALB. It’s more confirmatory evidence that these cells are definately ready for “prime time”.
There are a lot of similarly interesting observations to be made about a battery pack in actual operation. All of this is good knowledge to have about the behavior and characteristics of your batteries that ultimately help you understand them better and hopefully translates into getting the most out of them be it performance, long life, or whatever is important to you.
I have always agreed that a prescribed charging profile is a formula (or recipe) and not an absolute law. This IS an important concept for people to understand and cater it may be a bit abstract for some folks to comprehend. Just talking about it and showing examples hopefully helps the point sink in.
My goals as far as my battery pack is concerned is simply long life. With the meager tools and equipment I have at my disposal (and keeping in mind I’m still pretty new to this) my plan (read my “battery management plan”) to ensure this is simply keep from overcharging or overdischarging my pack. To watch for overcharging it probably takes me all of 5-8 hours a year to verify cells voltages with a DMM near the end of charging every 3-4 months. This is probably overkill but I can always change my protocol as my comfort level increases. On the overdischarging side all I need to do is reset my trip odometer after each charge and keep an eye on it and my AH counter while driving so as not to take too much out of the pack.
In some sense you could say I do have a BMS in my car. It’s just that I am the BMS.
Hey Jack, I like the new layout. I just now had a chance to see last weeks show. Just wanted to let Brian know the driver’s name is Boris Said and he pronounces Said just like “He said she said”. Not really important I guess, just though him might like to know. I have been watching Boris race sports cars for years and the last several years he has made a few inroads to NASCAR. He is a very good driver and when you run into him at the track he is happy to talk to you and answer your questions. I also ran up on this video that is relative to him and the BMW in the show.
Another great show, thanks.
here goes an extreme example , of charging a pack bottom balance
my setup TCCharger (Elcon clone) 1,5kW , with a charge 116,8 V max voltage, 12A max current and cut-of current of 1A
This charger is not recomended for charging these cells to LOW cut-of
Batteries are 33 ThunderSky 160Ah , out of 120 cells that I´m testing, lowest cell capacity is 177Ah and highest is 189Ah I did not mach the cells I did the opposite, and the cut of of the charger is 1A, this is doing everything wrong !!
The result was the lowest cell capacities ( 177Ah and 177,4Ah) on the final stage of the CV profile where at 4 V , after charge terminate, after one hour or so they were at about 3,4V (open voltage)
This was to see if there were trouble in charging bottom balance cells at these extreme conditions, at 1A cut-of and a big difference in capacities.
The 1A cutoff is extremed. 0.05 of 160 amps is about 8-9 amps. I do not think you would see 4.00 volts in that case.
I just looked at the calendar and realized that there are less than 5 weeks to EVCCon. My car is not yet roadworthy and I am starting to get nervous that I won’t be done. I guess I have to stop partying like a rock star for the next 5 weeks and get it all painted black so it will disappear 😎
Actually I know I won’t be done. Are these things ever actually done? I probably won’t have the heater working and I know I won’t have the AC in. I probably wont have the vacuum assist hooked up. There will probably still be idiot lights for things like the oil pressure and low coolant. But I am pretty sure it will be driving around under its own power.
I am really looking forward to returning to EVMecca. Who knew this would be Cape Girardeau.
So you bringing yours finished or not? I will not have some things done but I will have a running vehicle. My tow along house and garage is pretty much ready for the trip. New tranny should be in in the next few days as well as much of the brake work.
Oh yeah, I will bring it in whatever state it is in. I am really proud of a lot of the work I have done. Things like the carbon fiber motor mount and the fiberglass insulated battery boxes. I hope I can share some of the things I managed to figure out. Last year I looked at other peoples builds without enough background to even know what to ask. Now that I know what I needed to ask and look at last year I will be looking at peoples builds in a completely different way. I am sure I will be surprised at the ideas and approaches other people have taken to solving the same problems.
I find it interesting that no OEM’s are using the Lipo4 that you are selling. Correct me if I am wrong. Is there an OEM with a battery pack that you like?
I hadn’t thought about it, but no there is no OEM with a battery pack I like. Truly I detest the little cobalt oxide cells used by Tesla. The A123’s in the Fisker are interesting but the packaging is just a horror. I think the OEM’s have over engineered their packs with the wrong chemistry cells.
That said, they carry my undercharge and under discharge scenario to extremes, and really remove the issue of running out of gas from the consumers grasp. Chevy, in the Volt, provides a 16.3 kWh battery pack while allowing total SCO use of 65% by grossly undercharging the pack and then cutting in the generator quite automatically at very conservative levels. The net is a 10.3 kWh usable out of 16.3 kWh paid for.
All OEM packs are very blatantly proprietary and in most cases gratuitously so – just for the sake of BEING proprietary. On our cars, of course, we can upgrade at whim with ease from a variety of competing vendors. Until OEM’s provide that advantage, I doubt I would ever be happy with their pack designed.
I started looking at the OEM packs and found an interesting fact. My LiFePo4 pack has higher wh/kg than the Leaf and Volt pack even though the cells they use are higher. Here are the numbers:
My RX-7 pack: 99.88 wh/kg
Nissan Leaf: 80.81 wh/kg
Chevy Volt: 50.58 wh/kg
Tesla Roadster: 117.8 wh/kg
The Volt limits the use at the ends to increase the life of the pack. If they didn’t do this it would be 80.93 wh/kg. Very conservative. The Tesla pack is still better than my LiFePo4 pack but I am still astounded that they chose to do it the way they did. They must be paying very little for the cells they are using. The Tesla figure does not include the weight of the required environmentals to heat and cool the pack so is still an inflated number. I doubt that it would push it below the LiFePo4 number for my pack even if taken into account.
I thought the iMiEV had LiFePO4 batteries
The Fiskar Karma uses the A123 pouch cells so it definitely uses LiFePo4. The Wikipedia page for the iMEV says it uses Lithium Titanate Oxide made by GS Yuasa and developed by Toshiba. So not LiFePo4 in that case.
Whatever it uses I liked it a lot. It’s a nice little car. I drove the Citroën branded version, the C-Zero. Driving an ICE car seems to old fashioned and cumbersome now with all the gears and the power bands. Unfortunately the C-Zero and it’s siblings are 32.000 € here in Finland.
I don’t do the CV portion of the charge. For me the few extra miles of range aren’t worth the trouble. On my daily charges, I set the current to about 20A into the 36-cell, 100Ah pack, let it charge to 124.5V, and have the PFC-30 set to cut off immediately when the voltage is reached. Sometimes on the weekends, I used come out and turn the charger on a second at about 8A and let the pack charge to 124.5V again, still with no CV portion. I got very little extra Ah into the pack, so I stopped bothering.
Also, I can echo Jack’s warning about the 0.05 cutoff. One times, I set the charger to about 4A instead of 8A. When I went back into the garage an hour later to check things out, 124.5V still hadn’t been reached, but two cells had shot up to 4.19V! 24 hours later, they had still read close to 3.5V. So I’ll never do that again.
That’s very interesting Bill: I’d like to explore that more. I tried that approach (no CV phase) when fast charging on my test rig and not only did the maximum voltage variance drop hugely, but I was only leaving about 5% of the cell capacity on the table
Grrr, “turn the charger on a second ‘time’ at 8A” I meant to say.
I have 32 x 200 amp hour Sinopoly cells, I bottom balanced the pack to 2.8 volts then charged them with my TC charger set to complete at 3.65 volts per cell, I had the pack connected to my EV pro gauge which measured 211 amp hours to 116.8 volts, I measured all cells shortly after the charge and 3 cells went to 4 volts, this was a month ago, I have now measured the resting voltage and the 3 cells are all at 3.7 – 3.8 volts with most of the pack being 3.4, what would you suggest, I think I need to add 2 cells to bring the average down or use the alarm on the EV pro to turn the charger off earlier
I would lower the charge voltage to 3.5 volts per cell or add more cells. If you add one cell the effective termination voltage would be 3.54 volts per cell. If you add two cells the effective termination voltage would be 3.44 volts per cell. I am guessing you cant easily change the voltage of the charger which is why you posit adding cells. Adding one cell would probably be enough. Adding two cells you might not get a fully charged pack depending on how the charger behaves. Not fully charging the pack is perfectly acceptable if you get adequate range without going below 20% SOC on a daily basis.
The three cells you mention are either your weakest cells or they were not quite as bottom balanced as the others. In either case you are slightly overcharging them which over time will weaken them further. If you choose to add cells it would be best to take the pack down to the bottom balance point and add in the cell at that bottom balance point. Taking the pack down would let you check the balance of those three cells at the same time.
My EV background is that I have Mazda MX5 that was converted to an electric drive in 2008 and have now driven 48,000 electric Km’s. It uses the Azure motor and DMOC and originally had 73 Thundersky 90amphr batteries, which I replaced with 114 Sinopoly 60Amphr B batteries. The performance now has exceeded all my expectations.
I have always used a BMS system of some sort. The Thundersky batteries had a simple shunting cell board on each battery (no BMS computer) that shunted at 3.6V, and the charging was stopped when the first battery reached 3.8V. This worked a treat and the batteries were like new when I swapped them. For my Sinopoly batteries I used a BMS system that I designed for someone else. This system has a computer and works by charging at maximum current until the 1st battery reaches 3.6V and then starts throttling back the current to keep that battery at 3.65V until the last battery reaches 3.65V. Each battery shunts about 0.8Amps as they hit 3.65V. The time it takes for all the batteries to get to 3.65V is about 10min from the time the throttling starts.
The saftey features are, that the charger max volts is 416V (3.65V/cell), the software has a WDT that switches the charging off if it times out, and there is a hardware WDT which times out if communications with the cells is interupted at which point it stops charging.
The problem I see with the bottom balancing is that the small change in voltage at the end of a charge can be cumulative. You ask how the batteries can be different if the current is the same in a series configuration. Well this happens because the internal resistance of the batteries are slighly different, so each battery has a slightly different temperature when charging and discharging. Also the approximate 3% loss per month is different for each battery.
These of cause are small variations and take a while to show up, so when the BMS people show panic, because ev owners have lost batteries by not having a BMS, they fail to realise that these owners not only don’t have BMS’s but they also haven’t bottom balanced.
After reading a number of comments about bottom balancing, I still think, at the very least there should be monitoring of the batteries so that any OC or UC of the batteries can be rectified.
Just like you Jack, I have spent hours testing batteries.
By the way, the Sinopoly 60Amphr B batteries are 104wh/kg and I swapped to them because they reduced my car weight by 30Kg and gave me about 20% extra range. I will now use my Thundersky to store my solar panel power and charge the car free.
It’s your car and it is your opinion. The problem with all that is that not a single lithium ion moves based on opinion. Your understanding of internal resistance is flawed and so your conclusions are as well. You’re wearing your batteries out manipulating them.
But if it works for you – drive on.
Thankyou for your opinion of my opinion. I was trying to be helpful based on my experience and tests, but I see its hard to help someone who isn’t flawed. I hope all goes well with all you open system battery charging people.
Walter – your experience is interesting but I would have to agree with Jack that your theorising on the effect of internal resistance is dubious. It certainly doesn’t match my experience: I saw no evidence of cell drift in 500 + instrumented charge/discharge cycles.
Incidentally for my next test series I plan to put the simple cell shunts you describe on half the pack. My expectation is that they will do little harm but little good either – but I’ll keep EVTV posted on the data.
Now you have me interested. I’m not sure what you mean by cell drift. If you mean the 3% loss of charge per month, then that is specified by the battery manufacturers, and its easy to check by fully charging a battery, leaving it for a month and then check how many amp hours you need to fully charge it again. If you mean the relative difference between the batteries, while charging, it is obvious when you watch 114 batteries get charged and see that it is always the same battery that reaches the fully charged state first, and same goes for the battery that is always the last to be charged. The shunts mearly make up for the difference. In my case the difference is about .8amps for 10 mins, which is about 0.13 amphr. Sounds like a small amount, but left unchecked that would reach 1.3amphr after 10 charges and 13amp hr after 100 charges. I would love to see the data of your experience.
If you would like, I could video the last 30min of my charge and you can see what happens.
We kind of know what happens, and as both John and I have tried to explain, it is not what you think. But thanks for your opinion, of my opinion, of your opinion.
If we mean the 3% loss of charge per month, as you note specified by the battery manufacturers, then why with batteries that have been sitting for four years do we not have to put ANY amp hours in them at all? The reason is they do not self discharge at all, and indeed there is no such mechanism in a lithium ion cell to support a THEORY of this self discharge. This is simply lead acid thinking carried forward.
Purchases REQUIRE this information and the Chinese are all to willing to comply. But what it says is LESS than 3% discharge per month. Which seems to satisfy the entirely ignorant inquiry they kept getting from people wanting to know their discharge rate. We have tested the same batteries at TWO years, THREE years and FOUR years from the same shipment. No discharge of any kind. No measurable level. Nada.
This is what I am referring to. You, along with most of the BMS crowd, think you know something of these cells and have an entire lore wrapped up around it. But you are mostly reading messages from God in cloud formations.
The voltages you are reading are not even cell voltages. They are charge voltages and they are affected not by the magic sauce “internal resistance” that you allude to, but by diffusion delay.
Cells of different capacity will indeed reach full charge at different points. And without fail the one with the least capacity will reach fully charged first. If your shunt balancing did anything at all, that would be the end of it and they would all arrive at the same point on the next charge at the same time. As you dutifully note, they don’t. They don’t because you did not do what you think you did.
What you did do was a totally useless exercise in moving electrons around during the final moments of charge and you’re wearing the spots off your cells in doing so. To absolutely NO end but your own entertainment. It accomplishes precisely NOTHING. Other than to annoy both the cells and myself.
As I said, they are your cells and you are quite welcome to chant over them, take them out and arrange them on an alter at the precise moment of the full moon, or spit shine them with Nuvite. It’s all good. Because they are paid for.
That we don’t fall over (other than laughing) in amazement at your efforts and your “testing” should not be a source of irritation to you at all. We’re just religious agnostics and can’t probably be converted I’m afraid.
But you go girlfriend. Have a party. But don’t burn anything down.
Jack, Thankyou for your colourful comments. I fully charged my 73 batteries before I replaced them with the Sinopoly type in February and haven’t touched them since. I will attempt to charge a couple and see what amphrs they will take. If as you say, they immediately run up the fully charged curve I take my hat off to you, otherwise I’ll let you know the difference.
As far as bottom balancing without monitoring is concerned, it is an open loop system and as much as you promise nothing can go wrong, if something does go wrong you won’t know until it is too late. In my case once the first battery reaches 3.65V, it is merely held in CV mode until the last battery reaches 3.65V and then charging stops. I follow the manufacturers recommended charging pattern and it is interesting that everyone including the manufacturers are wrong and you are right.
I am sorry that the mere mention of a BMS irritates you so much that you get rude, but it is your forum and I appologise for taking up your time. (I suppose you feel that youv’e heard it all before)
John- a warning with testing half your pack with simple shunts. You need to top balance the batteries before it works, and if you decide it is not for you, then you will have to do the bottom balancing again.
Mention of a BMS does not irritate me, although a condescending tone can. If you perceive my written words as rude, then I was rude before your arrival hear and you need not fear that anything you said may have started it.
As to the open loop, it is of sufficient concern that we have expermented with an entire host of strategies over the years to monitor cells. What we’ve learned the hard way is that it is terribly easy to introduce parasitic loads that UNBALANCE your pack without intent.
For example, there is an entirely innocent little device called a Cell Log 8S that simply connects to 8 cells and provides an indication of voltage. Indeed you can alarm this various ways as well. But they are powered by the cells they are measuring.
Despite very low power requirements, if you do not have them on ALL the cells this power draw is cummulative and indeed causes problems. Worse, we find later that it draws power from the first six cells it is measuring and NOT the final two.
Additionally, all the wires come together in one connector. We had one of these soften and short two of the cells, burning the insulation off the wires and starting a small fire. Fortunately we were there.
We use a combination of voltage and AH measurement to “monitor” the pack. We used to measure it a lot by hand but over time the cells were just fine and the interest in measurement wanes after awhile. The cells perform admirably. The failures we have had have been within the first charge cycle or two. If they survive that, they run for years.
You are quite correct. It strikes me as strange as well that eveyrone, most especially the manufacturers, would be wrong and I in my little shop would be as you condescendingly put it “right”. That makes it neither so nor not so. It just is.
I have observed through several years of interacting with Thundersky, then Winston, Sky Energy, now CALB and a number of product introductions that the “manufacturers specs” are graven in jello. THier printed materials change regularly with no change in their batteries, and worse their batteries often change with no change in their specifications or printed marketing materials.
Thundersky very specirfically has gradually brought down their CC/CV point to very nearly match mine and it really isn’t that much of a coincidence. If you notice, their logo is on the Elescalade and indeed they provided the cells for that project gratis.
They originally recommnended charging serial strings of cells to 4.2v per cell and indeed at one point Winston claimed it would hurt the cells to undercharge them.
If you will simply graph the “charge” voltage of a single cell taking measurements each 30 seconds from 3.45 volts up to 4.2volts it will become readily apparent what nonsense this has to be. The curve turns essentially vertical with very little energy going into the cells at all.
About that time, we bought a Thundersky charger for 24 cells. It was entirely UN adjustable – a totally fixed voltage charger and it was set for 3.65 volts – very close to the 3.60 volts we were then recommending and indeed what has since emerged as a nearly universal charge voltage for LiFePo4 cells, including CALB, A123, Sinopoly, and Winston.
You will be further surprised to learn that the relationship between the battery company you are talking about and the factory where they are actually manufactured is not precisely what you think. And the branded specifications and marketing materials how they are derived and produced. And the odd to Americans relationship between sales, management, design, and science exists there.
At some point, you will lose your innocent faith in what appears from the manufacturer. But you’ll also learn that none of it is intended in the typical American swindle fashion. The cells always hold MORE energy than the spec sheet says, And just as you have disconvered in this case, the cells indeed DO exhibit LESS THAN 3% self discharge. Of course, at some point in studying these cells you will learn that the chemistry of lio-ion “rocking chair” cells simply does not support self discharge at all. There is no mechanism for self discharge. There is no internal shuttle mechanism such as you may be accustomed to in Pb cells for example. It has no analog here. But you must admit, they told the truth in fact. The self discharge level IS less than 3%. Obviously MUCH less. Like zero.
But you have to picture why this is. ALL battery companies list their self discharge level on their spec sheet. If I as a Chinese salesman, am trying to sell batteries to Americans, in a second language of course (Chinglish), and they request this data from me, I can reply that it doesn’t apply. The customer thinks I’m avoiding the question. If I get it 300 times, I take it and show it to management, who in many cases doesn’t speak English whatsoever, and explain it to them.
And then they take it to someone at the factory, which is actually run by the city and entirely separate from the battery company you are buying cells from, and after much arm waving, they assure the manager that it is less than those examples. Much less.
And so ultimately, as this traces back, they proudly survey all the spec sheets, and note that their self discharge is less than all those. Indeed less than 3%. Some say less than 1%. There is no difference between the two claims or the two batteries.
They have truthfully answered the question they were repeatedly getting, have they not?
We have a new booklet we have received from CALB with their new CA series and you should have a copy. You will find it fascinating and very encouraging. They not only insist on the use of a BMS, but also have repeated every myth I’ve ever heard uttered about these cells into one small book. It is just total nonsense from the front page to the back, and included with every order.
They also make some interesting claims about better low temperature performance, better power output, and better cycle life. We have verified most of those already – indeed to be BEYOND what they claimed in every case, and we’re working on others. Both John and I are planning power output tests of their more manageable 40Ah cells. We’ve completed the -10C temperature tests. And I strongly suspect that John will also find the cycle life promises equally true.
So we are left with both the fortunate and the unfortunate. The cells are actually BETTER than the company selling them claims, and almost everything on the spec sheet has to be questioned at the same time, as best we can by experimentation, test and measurement. And if our meters don’t match the given theory, we have to account for the meters.
Your procedure appears to me to be based on reading your meters, and then rationilzing how that fits with a bunch of crap dog shit you read on DIY from morons with overactive keyboarding skills. I can’t help you with that.
At the same time, you seem ammenable to testing and measuring. We would very much welcome your participation in the kinds of things Damien Maguire, myself, and John Hardy do and report here. But you have to be prepared to have RESULTS that often don’t HAVE immediate explanations, and some explanations that are still awaiting results. It’s a much uglier process not nearly so neat as you may be accustomed to, but over time, we can usually work out what it means and sometimes what it cannot mean, among us.
And yes, we’ve all become ENORMOUSLY mistrustful of know it alls and conventional wisdom and spec sheets.
The “recommendations” regarding charge votlage points and bottom balancing are a reaction to an insistence among our viewers to put this in terms they can use to make their packs last longer and avoid some of the mistakes of the past. And they generally come to us AFTER ruining a pack or part of apace, paid for with cash money hard come by, as a result of some newbie discovering the exciting world of electric vehicles, and sincerely wanting to participate, immediately goes out and designs the ultimate battery management system, with NO underlying understanding of these batteries whatsoever.
All cell death so far observed has come from overcharging or over discharging. For overcharging, we charge the series of cells up to thepoint where the majority have turned UP the steep part of the curve. The variations we observe from the greatest capacity to the least are trivial at that point – a few hundredths of a volt. Not wild swings like 3.4 to 4.2. More like 3.45 to 3.7 in extreme cases and 3.5 to 3.65 in most cases at the very end of the charge cycle.
We bottom balance to avoid cell reversal which is how you kill a cell in discharge. The reasons are many, but the short form is you can discharge at much higher currents in a car than you can charge with a charger. And the damage occurs much more quickly in the scenario where some cells have energy, and others do not. If they ALL run out of energy at the same time, they don’t seem to harm each other as much.
I do not think of myself as rude. I think of myself as direct. If you start from the premise that I’m wrong and you are going to show me, I’m actually all ears up to the point where you haven’t shown me. At that point, it may be true that I suffer fools poorly. That has certainly been noted in the past. I have no particular defense of the remark. But also little motivation to cure the problem.
So my normal reaction would be that it is ok to be offended and go away. But if you are sincere in your quest to test and measure, I’m loathe to do so. I might suggest you have a chat with John Hardy and develop some questions that might be interesting to investigate and work together on some strategies to test that. That could go two ways, you each do something different and sum them. Or you each do the same and confirm each others results – perhaps with variations on the methodology but aimed at the same question.
THAT sort of thing we have endless patience with, and indeed continue to take the congressional privilege to revise and extend my remarks later based on new data.
Welcome to my world.
My hat goes off to you. I just got 2 of my thundersky’s at random as I said I would, put them on a 4 amp charge starting with battery 1 at 3.36 volts and battery 2 at 3.44V. They both immediately headed north to 3.65V, battery 1 taking about 3 min longer than the other.
Ok I am now curious. In October I will have time to give your bottom balancing a try, but I don’t like the idea of an open loop system, so I will keep the BMS for monitoring. I will simply disable the shunts.
Hi Walter. Jack ran a segment on my battery testing a few weeks back but in response to your request for detail I’ve posted a couple of new graphs at http://tovey-books.co.uk/testing.php for cycle 500: this was after a month or so of testing. I’ve been meaning to post all the raw data but time is a little tight and loading files in my clunky website builder is not as slick as it might be. I think looking at the curves you will appreciate why I am reluctant to top balance: imagine all the cell voltages shifted so that they lined up at the top.
I agree about the open loop issue . You might want to look into the Lee-Hart battery bridge as a way of detecting pack anomalies without introducing hundreds of potential points of failure in the form of cell-level devices. I think the other thing to say about shunt balancers is that they don’t appear to have any benign failure modes. Open circuit silently stops them working. Short circuit would be obviously nasty. Any bleed above a few microamps will unbalance the pack over the course of years.
Walter – I take my hat off to you too: your last post was a generous response. Come to EVCCON if you can – we are all in this together
There is no magic in bottom balancing. If you simply avoid over discharge you will avoid cell damage and bottom balancing is entirely unnecessary. We have reached the point where we do perform this routinely, but with a good AH meter and an odometer, you really do not need to do it.
But let’s take a look at it anyway. The cells vary in capacity. Fortunately LESS so than previously. THe consistency is improving, but still varies. Our new CA 180Ah cells run from 187 to 201 as the highest and lowest we have seen. 90% fall into 196, 197, or 198 amp hours.
The cells arrive nominally charged to about 60% SOC and we assume at the same SOC from one to the next. If we had a 187 AH cell and a 201 AH cell, the extremes we would expect that difference, 14 AH to exhibit with a 7Ah difference at the top and a 7 Ah difference at the bottom.
IF you top balance, and if top balancing DID actually bring them to the same SOC at the top, you have exacerbated the difference by moving it all to the bottom – 14AH. As you discharge the cells, once they have both reached 187AH expended, you would find the static voltage of the two cells in two different places. The 187Ah cell woudl be at 2.50 volts and teh 201Ah cell is still up on the level plain of the discharge curve at 3.10v.
That’s a little bit of a problem. In a car, you can still drive and make the car move, and you move current THROUGH the 2.50v cell that is on the very steep part of the discharge curve – driving it to zero volts and then reversing through the cell. These cells are NOT recoverable. Meanwhile the other cells are still up at 2.90 or 3.00 volts, just starting down the slope, and the analogy I use is that they turn like a pack of wolves on their weakest member and eat him. Sharks would probably be more accurate.
Of course, we charge at 15 amps. But we now have cars that can discharge at 1000 amps. And so what you observe charging can take minutes. The same thing on the other end can take seconds.
IF we align all the cells at 2.50v initially, and charge them as a string, now the 14 AH is exacerbated by concentrating it at the TOP of the charge. So let’s don’t go there. Let’s charge the pack to 3.50v.
You might, experimentally, charge your pack to 3.50v per cell, let it rest an hour. And then charge it to 3.65volts. You’ll find it is a truly trivial amount of energy forfeited this way, and so range. And you would find the economics of adding two cells to make up that range, much more attractive than a top balancing BMS cost wise.
Now when we go drive, the cells are at all differing SOC, but they converge at the bottom. So at the point where one goes over the cliff, all the others do as well. So no one cell can really gain an advantage over any of the others. THeir individual ability to produce current diminishes as does their ability as a whole pack.
So the performance of your car noticeablly degrades to the point where it won’t roll at all. We have actually rolled a car to a complete stop and subsequently recharged that pack with absolutely NO cell loss at all. Indeed, all cells in that pack are still in use today.
So bottom balancing gives your cells a chance for survival in the extreme situation where you do in fact over discharge your pack.
We avoid overdischarge generally with a capable AH meter.
We avoid overcharge by a capable charger.
We have learned to observe our pack by pack voltage, current, and ampere hours, along with our odometer. After awhile, you develop a sense for what it should all look like in relationship. I’m rather confident that a signficant problem with any one cell would soon become obvious.
We would troubleshoot that back in the shop, not on the road. And if we had damaged cells, we would replace them. I can think of about two instances where that has occurred. We had no problems detecting it.
Both the expense, and the problems caused by well intended Battery Management Systems, simply are not worth the effort. Oh, I guess I would love to have a display showing me every cell voltage and highlighting any that were out of line, particularly under load. But I wouldn’t risk a single 24 gage wire to the middle of the pack to get that information.
I thought at one time of using XBee/ZigBee type devices to report cell voltages, which they do inherenetly. But they also go to sleep and wake up in various ways and I’ve never convinced myself they would actually draw power totally equally from the cells – introducing yet another unbalanced parasitic load. So I’ve never pursued it. I have a bunch lying here in my bedroom floor threatening to be used.
My pack analogy:
Think of your battery pack as a series of inverted wine bottles of slightly different volumes. The openings at the bottom are all at the same level(altitude). As you charge the pack each bottle is filled with the same number of drops(electrons) as they approach the fully charged voltage all these different bottles show different psi(voltage) because the tops of the liquid are all different heights due to the volume differences. The charger stops when the sum of the bottles psi(voltage) reaches it’s set point. The key is having all the bottles run out at the same instant. As they empty if some bottles still have higher psi as they reach the empty point they overpower their neighbors and drive them to destruction because they have higher psi(voltage) and are able to do so. If they are all at the same psi when they each empty they simply have no force or strength to harm the other cells.
You can see that in such a system, measuring the voltage near fully charged and making them equal psi or voltage at the top actually ensures that as they reach empty they will all be uneven and cause the death of the cells with less capacity or volume. The closer to empty you can measure and adjust them to equal the better off you will be.
Warning! the BMS you monitoring the system could easily create any imbalance you see. Even 0.1 milliamps difference in standby load adds up when attached 24 hours a day for months or years.
Hi Jack and John,- understanding the concept of bottom balancing has never been a problem. Let me explain my curiosity and scepticism. Firstly I had always assumed that lithium batteries loose up to 3% of their capacity because of the specifications, so any loss I noticed I put down to this 3% . Jack said this is BS so I checked it and he was right. I now know that the loss is as a result of my cell boards across each battery. My curiosity is this:
I charge every 24hrs and as I have explained before, the difference between the first and last battery being top balance is about 0.13Ahr over a 24 hr period. This means that there appears to be a 5ma higher drain in the last battery over the 24hrs, so I will now be looking at the difference in load due to my cell boards. On the side, I wouldn’t think that diffusion delay is the reason, because each battery gets the same charge and the same time to diffuse, and by the time the first battery gets near 3.65V the charge current is very low.
The scepticism is that for the open loop system to work, all batteries must have exactly the same efficiency. That is, any loses within each battery when large currents are flowing (I like to call this the internal resistance, but you can call it what ever you like) are identical.
I am now retired, so as I said, in October I will have plenty of time to investigate this.
With Jack’s track record, I dare say that the difference in internal losses will be less significant than the difference in the load that my cell boards exhibit, but this will be discovered.
One final comment is that my amphr meter is highly accurate and is set to show that the batteries are empty when 54amphrs have been used. I have only once driven my car to fuel gauge empty and the lowest battery voltage at rest was still over 3.2V.
Good questions and I’m not sure. But question the assumptions first. Does each battery get the same charge? Hmmm. Each battery gets the same current. As we have observed, ALL of the voltages could be different, in theory with no two matching in fact.
I’m going out on a limb here and saying they do not all get the same charge – to some level of precision.
And we get into this a bit here Walter I will admit. A thing can be quite true without quite being relevant, it all being a matter of degree.
For exaimple, let’s say cells do drift. About 0.002% of charge per year. How many years before corrective action should be made?
Contrast this with nightly balancing.
I prefer the term equivalent series resistance and it is a function of changes in voltage measurable at the terminals under different loads. The reason I like to call it equivalent series resistance, beyond it being the proper term, is that otherwise people get it confused with resistance, and start trying to drop voltage across it as heat and calculating efficiency.
The act of charging and the act of discharging are both intercalation and deintercalation of lithium ions from a crystalline structure. Actually both happen whether charging or discharging, just in swapped locations.
I suppose it is true that some cells are better than others. We know this in regards to capacity, and so in the way they produce power and exhibit diffusion delays.
Extending that to cells wandering because of varying efficiences gets us back to degree. Cancelling errors perhaps. But I’ve observed nothing to cause me to be curious in this area.
Your very observant. I would rate 0.13 Ah on 54 Ah in the noise level. That’s what I refer to as reading messages from God in cloud formations.
But if it occurs consistently every time, why does all this balancing not make it go away?
Hi Jack, thankyou for your informative comment, it was very useful.
My suspicion as to the answer for “But if it occurs consistently every time, why does all this balancing not make it go away?” is that my cell boards are loading my batteries differently, which of cause will be investigated. The designed load is about 2ma when the car is off and an average of 7.5ma when the ignition is on or I’m charging, but this was never actually checked.
I ike seeing my batteries bouncing about while I drive or charge so I will always use them, but one use that even you may consider as good, is that I will be able to bottom balance all my 114 batteries in less than a day.
A few comments ago, John ask if I would like to come to EVCCon. I would love to but its a bit to far from Australia. I will be in Washington DC next June to visit my Daughter so if anything is going on then, let me know.
It may be a bit too far from Australia, but again, where have you been? We had half a dozen guys from Australia and New Zealand LAST EVCCCON and I think we have a pretty good contingent again this year.
Of course, they may have been from parts of Australia and New Zealand that were a lot closer to us than you…..
Australia is a pretty big place….
Jack, I watched your EVCCON video and am impressed with the whole concept. Unfortunately I can’t make it this year or next year but I’ll watch from a distance. Perhaps I could make a short video of my car for next year, because it appears that the most have DC conversions, and mine is an AC conversion. For example, one thing that I hear from DC converters is that regen is useless and from a cost point of view this is probably right, but I find there is a significant return.
Another point I noticed is that you had a low resistance to chasis in your Mini Cooper. I developed (again for someone else) a device that checks the isolation when I turn my ignition on and displays it. The standards here, and I presume in the US is that the isolation reistance needs to be > 10K per 100V. What does everybody else use?
Hi John Hardy, I finally had a good look at you test results that you offered. They are so interesting that I can’t wait to do my own trials.
There are three points that seem obvious to me in analysing this data
1. Provided you don’t over charge the lower AmpHr batteries, bottom balancing will work. (I can just imagine Jack’s thought right now. ‘der’)
2. For this point you must remember the concept that the energy available from a battery pack is limited by the lowest Amp Hr battery. So if you fully charge (ie including its CV phase) the lowest amphr battery, then the pack can be considered as fully charged, and you do not loose 5% of the packs energy.
3. The reason that during the CV phase, some batteries peak while others fall is that the batteries that peak are the lowest amphr batteries and the voltage is running away. In doing so, the current falls, and when the current falls on a battery that isn’t fully charged, the voltage drops accordingly, so all the batteries that haven’t been fully charged will have their voltage drop a little.
You could confirm this by checking if the peaking batteries are the lower amphr batteries.
You did say that Jack featured your test in one of his videos. If you let me know which one, I’ll watch it.
We don’t do “I told you so” or gloat when knowledge is gained. Welcome to the club. And you’ve hit on, with no prompting one of the two main tenets of this approach to batteries. That is, you never were going to get a single AH out of your battery pack beyond that held in the cell with the LEAST capacity anyway.
The other thing that you have to develop a visceral understanding of is that ALL you see when charging is part of a recipe – a procedure. The voltage of a LiFePo4 cell is never more than 3.4000v. Kind of built into the chemistry. But unlike other batteries, it can take up to 24 hours to observe it. It can take that long after charging for the cell to reach it’s true open circuit voltage. Because we cannot see this without excessive delay, we are kind of blindly pumping energy into the cell with the hopes of a good outcome. And a good outcome can be reached by following a PROCEDURE of things we CAN observe if we do it carefully. That procedure is to
charge until the APPARANT voltage reaches a certain point, and then hold that charge until the current drops to a specified current, usually C/20. Unfortunately, this procedure is ALSO calibrated by RATE – usually C/3, and varies modestly at lower or higher rates. Because of this, the voltages you observe at the end of charge are not quite what they appear. We get used to looking at them and start to think they are real. They really aren’t. They are just observable indications of a complex process going on in these cells and there is delay. As I said, ti can take 24 hours for the true open circuit cell voltage to stabilize.
Try this yourself. Charge a single cell as carefully as you might to the exact manufacturers specs whatever cell you are using. To exactly the voltage, diminish to exactly the current, and charge at the specified rate. Then let the cell sit for 24 hours. If you do a very good job, you will probably see 3.36 or 3.37 volts.
Hook up your charge equipment and start again. Measure how many additional AH you can add to this cell using EXACTLY the same procedure. If it was fully charged, it shouldnt’ take any. Side bet?
Would you believe you can do it AGAIN?