We struggled to get a show out this week.
Part of the problem was that I spent entirely too much time playing with RechargeCar’s AutoBlock AMP. For some reason, I keep having a brain freeze and calling it the AMPBLOCKER. I have no idea where that came from.
The REchargeGuys chose PROCESSING to develop the software for this device and it requires a Jave Runtime kernel. On Windows7, either the USB drivers or the Java slow this program to a crawl. It takes 3 to 5 minutes to LOAD any of their widgets. And when you do, they are largely non-responsive.
We added a MacPro laptop to the game and got the software to come up. But there is no documentation currently and so we were guessing as to functions. That said, it’s a good Beta team to be on. This device was announced over a year ago and is still not quite “available.” But it is very promising. It uses a hall effect sensor to measure current flows into and out of your battery pack, and can so measure instantaneous current as well as totalize ampere-hours – the only good way we’ve found of estimating state of charge for the pack.
But the philosophy and concept behind this product is very good. First, it addresses existing tachometers and fuel gages. I don’t think mastering a digital counter of amp hours is too high a price to pay to learn to drive an electric vehicle. It is a LITTLE different from a gas gage, but in a way, it is more the gage we always wished we had. It’s a number. It’s finite. You simply observe it until it reaches a certain number. You have to know what the number is. 180. Or 100. Or whatever. Easily enough labelled. But it gives you a pretty good count of amp hours.
[jwplayer file=”news110212-iPhone.m4v” hd.file=”news110212-1280.mov” image=”http://media3.ev-tv.me/news110212.jpg” streamer=”rtmp://s3einxnpkaij93.cloudfront.net/cfx/st/” provider=”rtmp” width=”950″ height=”584″ html5_file=”http://media3.ev-tv.me/news110212-1280.mov”]
But our viewers modify a variety of vehicles – some going back to the 1930s. Often the 1960’s. And they already had fuel gages. If we can’t make them work, they don’t really go away. They just sit there dead. And that doesn’t look like a very complete conversion when you have gages on the dash that no longer do anything.
The Autoblock AMP will drive your tachometer to display instantaneous current in hundreds of amps in place of rpm in thousands of rpm. We use a small switch to toggle between the two.
The fuel gage output is of course used to drive the fuel gage. And this can be quite flexibly tuned to simulate the fuel sender on a lot of cars using the GAGE program. This program allows you to tune the output of the fuel gage driver, the relationship to SOC, how many amp hours are IN full, and how it decreases. It also lets you set the pulses per turn for the tachometer. Better, you can enter values for current and SOC and watch the gages move to new positions to calibrate them.
This is infinitely better than the Ziva device we used before with blind trim resistors that were very difficult to tell what they did.
The Autoblock AMP has one final feature. It will output a 5v signal when a low SOC warning goes off, and again you can program that for anything you like. Since the device runs on 12v, we don’t quite get it. I would much prefer a 12v output sufficient to run a relay. But they envision a warning light LED obviously. It might take a transistor or opamp to actually drive a relay but you could use this to set a limp mode, or disable the controller, or anything else.
Playing with this rather set us back this week. We’re including some Speedster Nippon footage from the previous week that didn’t make it into last week’s show. We have the battery pack of 36 CALB SE180AH cells installed, thoroughly bottom balanced, and then charged as a unit pack to 120.5 vdc. That’s 3.35vdc and really rather fully charged. We’ve added a JLD404 to monitor all that and both J1772 and NEMA5-15 plugs to the car to do all that. We were able to devise a very attractive I think bedliner painted box in the front spare compartment without hacking up the fiberglass at all and it holds 14 cells. Each “saddlebag” box adjacent the motor in the rear holds 11 cells. And so we have really a very nice and secure mount for 36 cells of 180 Ah and I’ve never been so pleased with a battery pack in an electric car.
Unfortunately, our show was shot on November 2nd. And while I bleat piteously about the nefarious doings at Tesla missing their delivery marks, truth to tell we missed ours. November 1. This kind of assumed we could do this build in less than a month. It just didn’t happen.
We’re very close to the first roll. But that’s not quite a delivery. We have a lot of instrumentation issues, which is why I was working on the Autoblock AMP, which we really expected to receive weeks before. I don’t have the ELECTRIC emblem for the vehicle yet, though I’m told it will shilp this Wednesday – actually about a 100 of them. I had to buy 100 from China to put two on this car. Recurring theme? Methinds so. We’ll add them to the online store, though they are in rather a Porsche Speedster script.
We do need to test drive teh vehicle, and adjust the Curtis controller for proper feel on the regenerative braking and the accelerator. Shouldnt’ be too difficult actually. This final wiring and adjustment always seems like a big deal to me and everytime it kind of collapses to not much of a big deal at all at the end. Battery boxes are always 60% of the build effort. Drive train mounting is about another 20%. And that leaves 20% for EVERYTHING else.
But we do want to thoroughly test the car.
There is one thing I had not counted on. Speedster Duh was delivered to Tokyo the first week of August via Pomona. It was already licensed in Missouri. And we agreed to leave the plates on so they could test drive it around California before it went to Tokyo. I recall distinctly discussing this licensing issue and agreeing NOT to license the second car in Missouri in order to expedite shipment. We have obtained a dealer tag so we can test cars here, so no problem.
As it turns out, in licensing the vehicle in Tokyo, they made a kind of interesting discovery. Licensing kit cars is a bit different in each of our 50 states. Missouri has an interesting twist on this. First, you can license a vehicle that is 25 years old or older under a HISTORIC license plate. This is kind of cool. You license it ONCE and it’s good for LIFE – no re-registration every year. And that costs $25. Second, in a twist I don’t really understand myself, they title these kit cars as a 1957 PORSCHE REPRO. It is actually TITLED as a 1957 vehicle. And so even though it is brand new, it qualifies for the historical plate. This actually kind of makes sense. It IS a replica of a 1957. I guess this throws me because it is NOT a Porsche. But that is covered by the REPRO qualifier that it is a reproduction.
But in then moving the car to other states, or in this case countries, it IS a 1957 model because Missouri licensed it as such initially when assigning a VIN number. Kind of cool actually. Apparently this gave them some advantage in licensing in Tokyo.
So they tried to rewrite the history of our agreement to include the fact that we had agreed to license the car here. Didn’t happen. But can happen. So we’ll do it for them. The problem is, we have to apply for a title and wait for THAT and then drive it to Sikeston in the cold for a State Highway Patrol inspection and we have to have an appointment for that. This can all take weeks as I recall. Somehow, if we will do this Missouri Registration, they are ok with the delay.
This is all part of the tale of weaving in and out of governmental regulations in different jurisdictions. It is central to my philosophy of less government being better. This is because I see it in action, and can readily do the arithmetic to scale it to the millions. The cost of EVERYTHING is hugely higher than it needs to be, and the time to accomplish things is hugely longer than it needs to be, to satisfy what I view as simplistic child minds oozing good intentions but without sufficient savvy to get in out of the rain. They will put 99% of the population through eggregious hoops to solve a 1% problem, and they do it over and over and over. The cummulative drag on the economy, innovation, and the evolution of technology and our standard of living is NEVER accounted for. They are forever and BLISSFULLY unaware of the deep deep damage they cause. And I see our current economic malaise as simply being the sum total of too much of it over too long a time.
The public good? Look at all the thousands of lives that have been saved by safety inspections and these measures. Actually you can’t. They are the “blue elephant” for which the “blue elephant gun” exists. Certainly some level of advantage is gained somewhere. But the relative balance is just absurd.
A case in point is the UL1741 anti-islanding regulation. All inverters now have to be UL 1741. What this means is that they can only produce electricity when the grid is active. If you lose grid power, they shut down and do not produce power either. This purports to be so that utility workers are not injured trying to get the lines back up by power from individual backup units that are tied to the grid.
This makes PERFECT SENSE. Particularly if you are an eight-year old with Downes syndrome. But it implies that trained utility workers are unable to detect whether any particular line is live or not. And it flies in the face of generators that everyone hooks into their house anyway.
I have a Generac natural gas generator on my house – a 12kW version i believe. I installed it myself. With no inspections. I’m a lawbreaker.
I did however install the code requisite servcice disconnect, despite it’s $430 cost. This is a largish relay that DISCONNECTS my generator using the grid power on the line to keep it energized. In the even of grid failure, this relay DISCONNECTS the grid from my house circuit, and connects my generator. Actually it even STARTS the generator. At ANY time that grid power is restored, it again energizes, disconnecting my generator and reconnecting my house to the grid. It is entirely fool proof. I don’t need to do ANYTHING to make this work this way. And the only way it can fail is if the relay fails, which DISCONNECTS me from the grid. Simple. Robust. Virtually indestructible. And required by code anyway. Whether or not I have a solar inverter that HAS UL 1741 or does NOT have UL1741.
So what is UL1741 FOR? It is to actively PREVENT you from creating power for your home. Why would anyone want to do that? Anyone but an electric utility. Who as it turns out wrote UL1741.
An added safety measure? Added to what. We’ve never HAD a blue elephant in our area. Proof positive that the UL1741 requirement is WORKING to prevent blue elephants.
It actually prevents you from independence from the local grid. In the wake of Hurricane Sandy, I would count this as high crimes and treason in the event of war – and subject to the punishment provided – hanging by the neck until dead.
It certainly makes it difficult to hook up your electric car and use it to power your house. I have found no suitable device to do this. I need a DC input range of 100=350vdc and an output of 240 vac and no anti-islanding. If you no of such…. let us know.
Meanwhile viewer Alan Levenson reports that after six days, he does indeed have electricity back, his car is fully charged and now he DOES have some advantage over those waiting four hours in line for gasonline, and electric cars are not so bad after all. But he said he IS still rethinking the electric car thing.
In this episode, we attack the description of how a lithium ion battery works. There are a lot of problems with this which is why I don’t do much of it. First, I see no value in discussing the grade school simplistic explanations of how such a thing works that at the end, leave you wondering how they REALLY work. Second, our viewership is so varied, from those with PHd’s in Electronic Engineering and Materials Science to people who’s dad taught them to work on hot rods and would like to get into this electrical thing. And so any presentation is going to either bore someone to death or scare someone to death. And finally, the shocker for most of you is that NOBODY REALLY KNOWS how it works. They were able to invent it. They are even able to improve it. But you would be amazed how little the best inds on the topic really know about what’s going on inside. The number of papers published attempting to measure and learn the specifics of any part of it, cathode intercalation, electrolytes, anode intercalation, Solid Electrode Interphase layer, diffusion, impedance, etc is just astounding. Easily 10 to 12 major papers per quarter – each announcing that what we knew 3 or 4 months ago apparently never was so.
It’s like an onion. The more layers you peel away, the more layers there are beneath. And I’ve been peeling for several years now – nearly every morning with coffee after addressing my electronic mail, this is what I read. And quality varies there as well.
The good news is that Google has just become phenomenal – a living breathing thing. It could very easily become self aware in the next year or so. The bad news is that it will vomit a lot more trash than it does gold, and you have to become very very savvy at sorting through it and rating/grading/considering the source. The concept of critical thinking doesn’t begin to cover the topic.
But the result is truly a remarkable firehose of information and investigation worldwide. The scientific INTEREST in the topic is clearly growing. And so inch by inch the knowledge.
I am fearless when it comes to producing long boring technical videos. We are not competing to go viral on youtube. We have our eyes set on a certain horizon of a certian number of viewers seriously restricted to a certain profile, and we are getting there hugely. But even I quaver at going to these depths and surviving the channel knob. If this is a bridge too far, let me know. I have to know it. But you probably don’t unless you just want to. You can build a very polished vehicle that does just what you want without ever considering charge transference as opposed to covalency. Or SEI temperature issues. Or Fe dendrites. And I really do NOT suffer from professorial desire to appear superior. Could care less what the world thinks about what I do or do not know. I focus on COMMUNICATING what might or might not be USEFUL to gain more enthusiasts for the electric car. Everyone thinks I was making some sort of inside joke in our very first video when I called for raising an army of 100,000 “guys” to go into their garage and build an electric car. I am actually quite literal in all things. And I can simply count on the fact that if they do so, they will also want to bring it OUT of the garage and demonstrate it. And it is my belief that is what is required to trip the entire planet into a mass movement to electric vehicles.
I have full faith in that. That IS the step. I don’t have to worry about technology or better batteries. I don’t have to worry about costs or economics. The price of everything. The value of nothing. The President. China. Or religious discussions. Or that lots of people don’t HAVE garages or for that matter electricity.
My hubris is actually limitless and I live in a self contained world. I have already determined that magnetic drive for personal transportation solves a LOT of problems globally and more or less permanently. Now how to accomplish that. And delightful as it is, because of some accidental happenstances of my earlier years, I know the answer. I need to find 100.000 guys that are not only like minded, but have probably already come to the SAME conclusion, and tell them it is alright to do just that, despite the askance of their wives, children, neighbors and the communal and confounding boobtube in their living room. They already know the answer. They need validation that it was truly a valid and appropriate question.
It is. Let it be so. Let it be written on all temple walls, obelisks and tablets throughout the realm. You are doing a good thing, and should do more of it. After that, we’re just doing guy talk in Tim Allen fashion on whose got the biggest tool. Ideally, but unlikely, …me.
93 thoughts on “CATIONS, ANIONS, ONIONS and JACKIONS. Through a glass darkly in the land of the Motor Verks”
Jack, you’ve really been outdoing yourself lately. Great shows and information about these batteries that are so essential to a viable electric car.
As more I understand it is getting more and more magic that it works.
Jack, loved the battery talk. They are as close to being alive, as anything vehicle makers have dealt with, since horses!
Also working with the Rechargecar guys on the Beta Autoblock Amp. Like you I’m fascinated by the idea that it will drive a fuel gauge and “Tach-Ammeter”. While I love the info the JLD-404 gives me, I didn’t want to put something so digital and modern on my vintage looking dash, so it’s mounted on a smoke colored piece of plastic (hacked from a desktop letter rack) hanging below the dash. I had a small hole between the Tach and Speedo, so I filled it with a 2 1/8″ Speedhut tachometer that I had them label “AMPS”. That way I’ll have a dedicated ammeter. I got kind of spooked with switching the tach on the eBugeye sometimes when I forgot how I had it set and the readout just looked wrong.
Still working with Rechargecar to get everything working right (the joys of beta testing), but I feel good that they’ll get it sorted out and have a really useful gauge driver when it’s done.
High Jack, I watched it all and I didn’t even need milk and cookies.
One minor point: at the 1 Hour mark when you are splaining to Jess the intricacies of 3rd grade math, and how to add 161 and 16.1 you give the answer as 176. I think you need to take back some Hubris!
Reminds me of my favorite maths quiz question. Add the following in your head… 1000 + 40 + 1000 + 30 + 1000 + 20 + 1000 + 10. Now check your answer on a calculator. This works best when administered verbally to your victim. Many people answer five thousand.
I never said I could add. But I ought to be able to hire someone who can.;-)
I stand corrected of course.
Back to the reason I love EVTV! I love the technical discussion and gain alot from it! Please keep these qualiy video installments coming!
All the best,
Thank you for a good show, when I watched the show, and I think the last one or the one before that, it made me think that I might have some good resources for you to look at. I have never my self used the stuff that I write about in this post, but it has been recommended to me in several different places, so here it goes:
Learn math, and get inspiration on instructional videos:
First up is http://www.khanacademy.org/ it looks really cool, if Jess wants to, and is motivated to, he might be able to learn math here, and as a bonus, you might get some inspiration on how to do instructional videos. I have never used the site but I am tempted, at least watch the TED talk that can be found embedded on the bottom half of the page.
Screen capture software:
The next resource is ScreenFlow, it is soposed to be the best screen capture software for the Mac, http://www.telestream.net/screenflow/overview.htm You should also be able to find it on the Mac App store.
For Windows the best screen capture software should be Camtasia: (http://www.techsmith.com/camtasia.html)
Sound quality when doing screen capture:
If you want to get the same quality sound on your screen capture videos, and maybe do a voiceover with the same sound quality as your recorded videos, I guess that you can use this device: http://www.centrance.com/products/mp/
It is a guess because I am guessing that you are using microphones with XLR connections, and this device can convert them into a USB microphone, so you can use these quality microphones together with your computer.
I liked your explanation of what is the current thought of what goes on in the LiFePO4 battery, for me it makes a good foundation for why bottom balancing it the best way to go.
Please do not out Jess on his math skills any more, only tell about the stuff that he is good at (you already do that) but only do that, and let him get even better at that, and if he finds out that improving his math skills will help him, and he is motivated to improve them, then help him, and maybe point him towards Khan Academy as I sugest.
Many thanks for the LFP theory. Fascinating. When you began your discussion of temperature, in this case heat, I was on the edge of my seat, and in the future I hope you might extend your discussion of temperature into the cold (Dark) side. Here in Maine we don’t historically have heat problems, and I would love to have you expand your discussion to how cold effects the theoretical internal workings of the LFPs, and specifically why it is a problem to charge a CALB cell, or the like, below 0C, and why Thundersky (Winston) cells with yttrium or Flux Power cells are rated at use from -45C to 85C for both charging and discharging. I have had assurances from Winston Battery and Balqon’s CEO that these temperature figures have been tested both on the bench and in the field, although they did caution to gently (low amperage) start charging and start discharging in extreme cold. I plan on running my Thundersky 160Ah cells into the coming winter, but obviously I am still slightly hesitant to risk the investment. Many thanks.
I’m also interested in why low temperatures cause the issues it does, why charging below freezing is not recommended and how is it possible that the new CA series cells fare better in colder climates?
An educated guess… As matter loses heat energy, the molecules grow closer together. This would make it harder for the lithium ions to move around the cell. My guess is this affects the anode side more than the cathode side, thus charging a cell at lower temperatures would be harder than discharging it.
I don’t know the answer to your question. And I’ve asked it. The gain in cold weather discharge performance is incremental – what 7 or 8% as I recall, but very real and quite welcome. There is a kind of overarching phenomenon termed the Arrhenius effect (I think) that would seem to apply here, but I’m not sure it really covers your question. The electrolytes are all essentially antifreeze.
Richard East of AMP Electric Vehicles visited last week and said they had done some experiments charging as low as 0F. Said to give it a shot to warm it up, let it rest a few minutes and then charge it and he had no difficulties. I note a strongly rising voltage that doesnt’ quite match the energy going in, kind of like we see a strongly SAGGING voltage on cold discharge that seems simliarly out of place. So there is some kind of diffusion phenomenon going on here to my way of thinking. But I just don’t know the specifics. I’ll cast about and see if anyone has done a recent paper on it.
Here is an interesting paper from JPL. http://trs-new.jpl.nasa.gov/dspace/bitstream/2014/41485/1/08-0012.pdf
They imply that at low temperatures, anode potential is insufficient to support intercalation of lithium ions and that lithium plating on the SEI surface occurs instead. Some of this may be reversible during the tapering current phase, but to some degree is irreversible and causes capacity loss.
So we follow the admonition not to charge below 0C pretty faithfully.
I believe I found the answer in an OLD JPL document from 2000. http://trs-new.jpl.nasa.gov/dspace/bitstream/2014/13702/1/00-0068.pdf
It would appear that the entire problem with charging is that at low temperatures the carbon anode becomes virtually incapable of intercalating lithium ions, and so lithium plating occurs. The electrolytes and the SEI layer appear to have nothing to do with it. Coke intercalates much better than graphitic carbon and and 6 um coke particles substantially better than 25 um coke particles.
In all cases, charging intercalation of lithium cations into the anode were very poor. But things do go better with Coke.
We can kind of infer that CALB is using a more cokish carbon mixture with simply finer particle sizes to improve cold weather performance.
But there remains a very asymetric result between discharging (Lithium ion extraction or “deintercalation” from the anode and charging – intercalation INTO the anode. They are not the same.
And so charging below 0C is rather persuasively a no no.
beautiful show Jack!
I am also very interested in what happens in the battery when the temperature drops, especially with WB LYP.
At a temperature of 10 – 20 °C can my car brake with 160 to 170 amps and at a temperature of 4 °C car the can brake with 110 to 120 amps and after 20 km it can brake with 150 to 160 Amps. As I see it, it must be resistance inside the cell that rises or have I got it wrong?
If you have a battery box will the outer batteries be kolder than the inner, can it get the cells out of balance?
When I started on my car, I’d make a passive BMS system and attached wires to the cells which were least accessible, but never got farther as the batteries did the same thing every single time they were charged and discharged. I have not measured my cell after that.
Thanks for another great show. I have been thinking about getting city fleets converted. There are two key areas that need to be addressed that I would love to see you or some other EV expert cover more completely.
The first is converting automatic transmission vehicles in a way that minimizes the efficiency loss. Most fleets have automatic transmissions and actually most cars in the US are as well.
The second is dealing with late model instrumentation.The Autoblock may go a long way to simplifying this. Connecting and integrating into the Cambus world may be required for many of the conversion of newer cars(like your Mini project.
You certainly dealt with both of these issues in your Escalade conversion. Even understanding how universal the solutions are would be helpful.
I hope more on these issues can be covered by you or others who send in videos.
Thanks for all you have done to promote EVs.
Well, while you did not give me the initial idea to start a conversion you did give me the confidence that I can actually complete this project. The amount of distorted information running around on the internet about conversions makes this non-trivial for the beginner. You can count me as one of your hundred thousand or at lest you will be able to soon. The motors in and the battery box is in… so by your calculations I’m 80%. Maybe before Christmas.
Regarding UL 1741 – can’t you use a small Uninterruptible Power Supply (UPS) unit and connect it to the inverter input so that it thinks it still have the grid?
Not sure why couldn’t you so that.
It’s probably not a question of CAN one spoof the Intertie relay as it is SHOULD one. Here is an example of one such relay, http://www.beckwithelectric.com/products/m-3410.html
Great Show and Blog this week! You are right about kit cars and state regulations. Here is GA it is nearly impossible to register a kit car. You would think that a safety regulation would be the issue. It is NOT. In fact we do not even have state safety inspections any more. What we have is a clean air force. Any car that this less than 25 years old is exempt from emissions testing (you know the ones most likely to pollute) I find in ultimately ironic that the “Clean Air Force” regulations make is nearly impossible to register an Electric Kit Car (or Assembled Vehicle as it is called here). Now, on the other hand, if you convert a gas car to electric, you qualify for a $5,000.00 state tax credit….
P.S. I just purchased and set up a web site for my EVThing. The URL is EVThing.me I’ll add some pics, videos, etc soon. The current page is my first 10 minute design and it shows….
Your site is fine for a start. Nice looking motor compartment and nice looking boxes. Painted white and they disappear just fine. Nice Thing.
Looking good Jeff!
I bet Mrs. Southern can’t wait.
It may come to pass Jacks evthing will be the best grunge car his daughter is going to blag off him yet. Yellow in lieu of olive drab camo? I may be very wrong. 😉 hehe
Listening to Jacks; “less government being better”. Peeps ought to consider Gary Johnson, Libertarian. The main vote options have nothing between them; representing a non-vote.
Thanks for looking into the charging below freezing and why it’s a bad thing. It’s very useful to know that if you do charge in such conditions you will cause irreversible damage to the cell. Makes one inclined to work on logic that would prevent charging unless the measured temperature in battery boxes is between say zero to sixty Celsius. Perhaps there is a product opportunity here as well – a JLD (612 maybe?) temperature meter packaged with a couple of suitable thermal probes connected to a contactor which only closes if the conditions are met. The charger would be connected to the pack via this contactor. Alternatively the JLD could tell the PulsaR if it’s ok to charge.
Insulating a standard converted vehicle is a tough act.
No need to over complicate simple things. Consider a cheap thermostat or two for safety. One for heating, the other for charging.
Brilliant idea Jarrko – and while we are at it, get the device to inhibit the charger if the batteries are too hot (i.e. when something went wrong)
Indeed, the meter should disconnect the charge if temperature is below zero or above sixty Celsius. Another meter could monitor total pack voltage and cut off charging if total pack voltage goes above 3.6 times cell count to catch charger failure. The temperature meter should probably also alert the driver if the battery box temperature rises too high while driving for example due to a loose connector causing excess heat. That should just about be all the “BMS” we’ll ever need.
Good point Jarkko. I plan to have an isolated voltage input to my little PLC. It will kill the AC power to the charger if it sees the pack voltage getting too high. I am not a fan of relying on only one device to prevent an over charge situation…
Next time a vendor tells you they only sell to OEM’s tell them you are an OEM. You build Porsche Speedster replicas for export to Japan.
In your recent shows your temperature measurements while charging have been very informative. Would you do some of the same experiments showing the temperature changes when discharging the cells at amperages they see while driving the car.
Don – I’m about to start some long term testing on CALB 40’s (when I can get the rig to run in the new configuration without stuff kicking out due to voltage spikes and EMI). I am monitoring cell voltages, current, variance and amp hours. Maybe I should add pack temperature?
John – That would be great, Jack’s tests showing the temperature rise getting the last bit of charge into the cells shows how temperature is probably a very good indicator of healthy cell operation.
High Jack. Great show as usual. I was just wondering when you were going to redo the last bit of the fast charging test to see what the real cell heating is through the copper rather than the SS bolts.
This relates to Jack’s current topic.
Interesting Mark – the issue as I understand it with tin and silicon as anode materials is that they have far greater volume change during intercalation so are rapidly pulverised
Best wishes you all you left-of-the-pond guys with the new/old president.
Have not watched this weeks show yet so, this may be answered when I do. I thought there were issues with Hall effect accuracy for current measurements related to temp changes and that was one of the reasons for using a shunt driven device like the JLD404.
How is the Autoblock AMP overcoming the accuracy issue?
Or am I just confused?
Your not confused. And it isn’t that accurate. But it lets you drive a fuel gage that never was that accurate either. We will have a JLD404 in the Speedster Nippon, but also a gas gauge running off the Autoblock AMP.
Unexpected Results from Fast Charging a single A123 Pouch cell. I am using a MegaPac with a 5v 40 amp module for my power source. It puts out a solid 5 volts and 45 amps. I have my JDL meters connected to monitor the voltage, and AH’s into the cell. I also connected up my PowerLab 8 to monitor and log the voltage file. What happened was the voltage rose much higher than expected and much sooner than expected. I hit the 3.5 volt range at around 1 AH put in to the cell. The cell was discharged to 2.9 volts at 20 amps then terminated. The voltage rose up to the 3.2 volt range but it was empty enough for this test as shown by the graph and resultant knee. The graph is not unlike what Jack has for his test but the voltages are much higher and much sooner. It is a test of CC only and when I reached 10 AH I noticed that there was some gassing within the pouch. The pouch was very tight when we started and is a NEW cell. There was NO heating of the cell terminals, tabs or pouch. When I noticed the ever so slightly swelling I shut down the test and had only put in around 10 ah’s. Time was not long either which is noted on the video. I have a graph posted on my blog if you’d like to see it. Video is processing right now.
So with a CC fast charge I noticed a very high rate of voltage rise even early in the charge process and no heating but noticeable swelling. After sitting over night I also noted that the swelling was worse than when I terminated the test. So it continued to gas even after power was pulled.
This was strange and totally unexpected. I expected to get at least 15 AH and voltages around 3.5 volts after reaching 15 ah. But nope. I am thinking it has to do with the power source. Not sure.
The voltage at termination of the test was only 3.7 volts.
Pete – that is very interesting – thank you for posting it. What we don’t know about these cells could fill a few libraries.
Is it possible you had a dud cell? IIRC Jack had a few A123 cells that unexpectedly refused to take a charge
It is quite possible but I am concerned about the high voltage output way too soon in the charge. I noticed it with my Hi-Power 100ah cell too. Same setup and power output of the MegaPac. I will be trying another. Don’t really want to do that but I will. I will keep watch of the cell over the next few weeks to see if it goes thermal on its own. It will be kept in a safe place. I am also thinking of dissecting this pouch and do a show and tell video of its innards. Guess I should drain the cell first.
Here is the video of my test. I am having issues with output from FCPX so this will do until I get the issue resolved. The update mucked it up.
Link will also be on my Blog.
How many Ah did you log out of the battery when you discharged at 20A to 2.9V? Had it been fully charged at a lower C rate before you discharged it, or was it fresh out of the shipping container? I have yet to do any testing on the A123 cells I got in. Still clearing the work bench.
I had charged and discharged the cell with the PowerLab 8 prior for a couple cycles. The last discharge showed 12 ah pulled out. I am going to discharge the cell once again but do a CC/CV discharge to 2.5 volts. Then start again. I will first charge the cell to 3.6 volts using CC/CV with the PowerLab at a 20 amp charge/discharge. I will end on discharge and post the AH out then do another FAST charge. I think that will give a good indication if there are any further issues. I will not video the charge/discharge. I will video the FAST Charge.
I did another Fast Charge. I first charged up the cell using the PowerLab 8 then discharged it too. I used the CC/CV algorithm to C/5 then terminate. I was able to pull out a full 13.99 amp hours. From a cell that is touted to give 20 amp hours I am very disappointed. I did another Fast Charge to 3.65 and got only a bit under 10 amp hours. No excess heat at all. 78F on the terminals and tabs. Just no heating issues. I posted new graphs on my blog but I am not going to post video after video of paint drying but I did video the process. I did that just in case the cell went thermal so I could capture it. It did not. Since this cell is just junk I may do a video of the cell charging until the cell terminates on its own. So go check out the latest Graphs if you want.
Clearly the cell is faulty. Not even close to 18 amp hours into the cell. I might get some extra AH into and out if I do a slow charge at CC/CV. I might just do that to see how well much it holds under normal low current charging like at 8 amps.
Little more on the charging of the A123 pouch cell. I did a slow charge discharge and got 13 amp hours pretty much and that was what I got with the Fast Charge then a finish charge at 20 amps. My slow charge/discharge was done at 8 amps instead of 20 or 45 amps. I have kept the termination voltages to 2.5 volts on the discharges and no more than 3.65 with the Fast charge and 3.5 with the 20 amp and 8 amp charging with the PowerLab 8. Graphs are on my site. It is clear that this cell is damaged and most likely from the over voltage on the FAST 45 amp charge where I got some puffing of the cell. I will redo this with another A123 cell. I will do some slow charge/discharge cycles to determine the amp hour capacity then I will do another 45 amp discharge to 3.5 or 3.65 but not more than that. Clicking on my name will take you to my blog to view the graphs. I see it is pretty much empty at a static voltage of 3.0 volts and pretty much full at 3.5 volts. Any more than that on either end is pretty much a waste of time. The last slow discharge took over an hour to put in 2 amp hours. What a waste of time to push that much. I will be changing the termination voltage on the PowerLab to 3 volts then go into CV mode and hold until C/5 or something like that at 8 amps. I want to see how that affects the charge.
Any other requests for testing? I am limited to no more than 20 amps with the PowerLab because more than that really heats up the cables on the PowerLab. I am going to parallel two 5v 40 amp modules to see if I can get up to 80 plus amps for FAST charging the A123 Cell and that will be on video.
Let’s keep this simple.
1. New cell out of the box.
2. CC/CV to 3.65 volts at 0.5C (10 amps). Terminate at 2 amps.
3. Discharge by the book at 10 amps to 2.50v. No CC/CV discharge. Just 0.5C to 2.50v. Note capacity.
4. Fast charge at 3C. 60 amps to 3.60v
5. Let cell rest for 45 minutes.
6. Finish charge at CC/CV to 3.65 volts 10 amps to 2 amps.
7. Discharge at 0.5c, 10 amps to 2.50v. Note capacity.
What we are lookng for along the way is any swelling of course. And subsequently, any change in capacity.
You damaged the cell with fast charging on the first bout. Swelling ALWAYS indicates damage. That’s why your capacity is less.
This was Rich Rudman’s observation as well. I could not duplicate it but the only test I’ve run it on was the CA40FI cells. Thundersky or High Power cells may indeed have that characteristic. I don’t know.
Interesting to see how A123 fares.
Do me a video of this and we’ll run it.
But I need it today of course.
I can do a video but it will be sorta late but I can do it when I get home from work. I can charge it up at 10 amps with the PowerLab and and I will not to the CC/CV discharge to 2.5 volts at 10 amps and note capacity. I can’t do 3C but I can do 45 amps. I don’t have control of the amperage. If I did not have to work today I would be on it straight away. Not sure if I can get you the vid in time but I will attempt. If not the viewers can always go have a look on my blog.
Wrote a lengthy piece on Bottom Balancing:
If you guys spot any mistakes let me know… Thanks.
Very nice piece it is too Jarkko. I think you covered all the bases. I would point out that in top balancing, they are using a voltage that is more procedural than actual, and so suffers from another level of abstraction. In bottom balancing, you mostly ARE dealing with static open circuit cell voltages which are quite indicative of SOC directly.
I’ve also been busy writing. A brief 12 page literary masterpiece on battery care. http://media3.ev-tv.me/cellcare.pdf
Thanks, Jack! Your PDF on the CA-series looks like the Missing Manual they should come with. I’ll have to read it thoroughly..
A literary masterpiece indeed! Thanks again Jack for all that you do.
Jack, very nice document. You might check note 5 socket size for the bolts should be 13mm.
Jack and Jarkko, I have read both your bottom balance explanations and found them excellent at summarizing this important issue, but you still leave the overcharge issue a little vague, which gives top balancers a point to argue. The way that I have reconciled this issue is that each battery’s 100% full voltage is a little bit different, and to determine exactly what it is you would need to overcharge the battery and let it settle for 24hrs or so. This is clearly not acceptable, but is exactly what top balancing does.
To avoid this Jarkko recommends to stop charging once the CC phase has finished, and Jack suggests a CV phase of 3.6 or 3.5 volts until the current drops to C/20. The point here is that Jarrko will definately not get to 100% so cannot overcharge his batteries and the CC/CV concept has by experience/experiment shown to get close to 100% without overcharging. By only allowing a 3.5V CV phase, it is a little bit less like to overcharge.
A small point to mention, is that once any battery is overcharged in a string of bottom balanced batteries, the pack is no longer bottom balanced.
It is not a little bit less likely that any cell will be overcharged. None of them will. And if they are, it does not necessarily follow that anything is unbalanced.
That is with anycells we have any experience with – Thundersky, Sky Energy, and CALB. I don’t know about headway and High Power. But the usually bleating response is about cell variation. It exists, but not to the extent you portray. I would be astounded to find any cell over 3.65v when charging to 3.5v and really no love lost to 4.00v.
Bottom line is that the total variation you will see is 3.44 to about 3.65 or 3.68. And that’s on a not very well matched pack.
If you DID have a just horribly bad cell, it would show up pretty quickly. And the appropriate response would be to replace it. We just haven’t had that.
With the “less likely to overcharge”, I was suggesting, that if it doesn’t overcharge when you bring the volts up to 3.65V, stopping at C/20, which is known from your experience, then it is even less likely to at 3.5V.
With the pack losing balance: I was suggesting that if you have a B/Balanced series of batteries and you continue to charge them until one or more starts to overcharge, you would slowly begin to top balance, thus losing your bottom balance. It was meant to be a warning.
With regard to the variation of the battery voltages in a bottom balanced pack when they are in the CV mode, I have no experience. However, If you have a look at the 8 batteries that John Hardy used to do his bottom balanced cycle tests, then the variation is quite a bit different to what you have experienced.
An item that doesn’t exist that is needed and a possible product for the EVTV store is a battery box temperature system. This would probably be an Arduino powered relay driver able to handle a seperate termperature input for each battery box (4-6 inputs) and drive a relay that would not allow charging if any were below 32 degrees F or 0 degrees C. From the importance of temperature shown in your battery manual there is an awful lot to be lost if charging is done at overnight charging. This is when unattended charging is done and these low temperatures are going to happen.
Re Don’s suggestion, I’ve been eyeing up this device: http://www.ebay.co.uk/itm/New-Automatic-Digital-Temperature-Controller-Thermostat-12V-Control-Switch-EA-/221147036126?_trksid=p2045573.m2042&_trkparms=aid%3D111000%26algo%3DREC.CURRENT%26ao%3D1%26asc%3D27%26meid%3D3347971000523003464%26pid%3D100033%26prg%3D1011%26rk%3D1%26sd%3D221147036126%26. Not sure if would do the job
Regarding the setup to bypass anti-islanding, using a solar system inverter and inverting the DC from either the EV Pack or a Solar PV collector output Use a 12 v dc to 120 v ac 100 watt mini inverter to supply “Psuedo-Grid-Power” so the UL1741 is “Fooled” into thinking the grid is up…
Err. Does that mostly work when the sun is shining.?
Actually, the most popular mini inverter is ALSO UL 1741 and it won’t make any power unless it has a grid to synch to.
If you look at the Outback units, http://www.outbackpower.com/products/sinewave_inverter/radian/,
They have provisions for both grid tie and battery charging and usage. One would think that you could transfer dc current to the standard battery backup if nothing else and they maintain you can hook 10 of them up for 80kW?
“it certainly makes it difficult to hook up your electric car and use it to power your house. I have found no suitable device to do this. I need a DC input range of 100=350vdc and an output of 240 vac and no anti-islanding. If you no of such…. let us know.”
If you look at the Outback units, they are 48vdc input. I could use the John Deere Lawnmower maybe…
In reference to your cellcare.pdf I may have found a typo. In the line “There are several fast charging standards emerging, including ChaDEmo, and SAE J1773
Rev B Level I and II DC charging.” I believe you meant J1772. J1773 appears to deal with inductive charging.
This is a well thought out document.
On your (excellent) Cell care document, might it be worth putting in something about insulating tools? We use a long quarter inch drive extension – the ratchet is taped with insulating tape and the long extension covered in shrink tubing (long extension is to keep your hands away from the sparks and/or plasma arc if something goes wrong)
Good idea John.
New link. I had added quite a bit to the CELLCARE document but our cache servers haven’t cleared. So it continues to serve the previous version. You can get the latest at http://media3.evtv.me/cellcare.pdf
Excellent, excellent late night reading. . My new bedside favorite!
In case anyone was wondering where JLD comes from:
There must be a lot of Seinfeld fans in China.
I am not an electronic expert so there may be a big problem withthis, but if you put three of the Outback inverters in series on the battery side, you could use 144vdc as an input and have three 8kw 120v outputs? Not cheap, perhaps doable. It is a bit odd that the solar dc input is low, because on the grid tied only units they always push the 600VDC UL limit.
A big problem Wayne. No assurance we are drawing from the different 48v “banks” equally – unbalancing our pack most likely to an eggregious degree.
I need a 100-300v input. I can live with something different like 120-240 for example, it simply eliminates some cars.
48v doesnt’ work at all.
Jack & Walter,
I came up with this way to explain the charging with bottom balancing:
We don’t charge to reach a voltage.
We charge to get the cells full.
This is indicated by pack voltage rising to 3.5 V per cell on average.
That’s when we cut or taper the charge.
In other words voltage rise is the cause, not the purpose.
Jarkko, What you are saying is correct, but the point that seems to be illusive is that there is no way to determine if a battery is 100% full, while charging, with voltage and current monitoring alone. There is a third factor which is time, and because no one wants to wait, they come up with charging profiles that get the battery near to 100% full. The very successful profile of CC/CV works because all charging is stopped at C/20.
Top balancers try to force 100% full by getting the faster batteries to mark time while the slower ones catch up. This will cause the batteries that are “marking time” to deteriorate because they ARE being overcharged. The interesting thing here is that, as Jack says, the only reason the batteries that were top balanced lose their balance in the first place, is the cell boards. They do not all have exactly the same load effect, so they are parasitic loads.
Yes. Now you are on it. Not discussed, but always in my mind, is that the shunted cells held at that voltage, AND the cells still receiving energy, are dithering about on what MIGHT be diffusion differences, not SOC charge differences. Some of John Hardy’s data implies this with the diverging reaction at the SAME charge voltage. It might be capacity differences. But it also might be a time factored diffusion variance that may not even be connected to SOC.
Yes absolutely. The only way I can understand behaviour at top of charge is that voltage there is an unholy muddle of differences in state of charge, rate of diffusion and impedance.
Love the show, even though it tends to run long, and the blogs. I was wondering if you were considering doing anything on the rechargeable zinc air batteries that Eos Energy is working on? Aside from their website and press releases, I think it would be great if you could get an interview with them and share your thoughts on the possible future of zinc air technology.
Thanks for doing the show and keeping us, the unwashed masses, informed.
No plans to. Anything that would work to make zinc air rechargeable would probably work with much higher energy lithium as easily. We kind of keep an eye on other chemistries generally, but they tend to devolve to grid applications more than EVs.
Thank you, Jack. I guess I was just hopeful that zinc air might approach ‘holy grail’ status as half the weight and half the cost, of course if I were able to use zinc air for longer distance, I would most likely have to match it up with lifepo4 cells to get good performance, ie. Acceleration, torque.
There are new chemistries and paper launches all the time, but until there are actual products that move actual cars, they have rather limited appeal. Oh, and they must also sell them to the general public, which is not always the case either. That’s not to say a 1000Wh/kg battery capable of 1000A continous wouldn’t be just lovely. 😉
It’ll surely come: my current laptop runs twice as long on battery as my work laptop bought three years previously.
But is the battery in your old laptop really smaller than your new one or is it just that the new one sips so much less juice? Cars will always use as much as they do now, provided they stay about the same size and shape. There’s only about 10% max theoretical improvement in motors.
True: I’m also thinking of silicone or tin anodes or lithium sulphur
There is also scope for serious improvements in tyre rolling resistance which would be one automotive equivalent of more efficient processors
I think you mean silicon.
I agree. There is actually a LOT that can be done regarding efficiency. Most of it has to do with weight. Aluminum and carbon fiber have been avoided in automobiles because of the cost. As energy increases in cost, such measures look increasingly viable.
Ford Motor Company is preparing an update to their venerable Ford F-150 pickup to incorporate a great deal of aluminum in the 2014 model – cutting the weight by some 700 lbs.
As a rough rule of thumb, a mile of range for each 50 lbs you take off teh vehicle.
The electric drive train makes almost all of the other trade offs in automotive engineering require a revisit as well.
An F-150 truck may have a lot of room for improvements, but if you look at John’s Honda Civic you can see that the possibilities are much more limited. I believe just about any given component (outside the batteries) in a mass produced vehicle has room for improvement in the 10% range, which although significant is not quite comparable to leaps made in computer technology.
Well, you can believe that of course. But it remains that the evolution of the car, including Hondas, has been in a context of internal combustion power plants and inexpensive gasoline. Every piece in it is a tradeoff made in that context. If you change the context, then all those decisions have to be revisited. And they are quite cumulative. It is interesting that you reference the Honda. The earlier Honda Insights were a text book example. They employed aluminum, plastic, aerodynamics, low rolling resistance tires, and cummulatively it was just one of the most efficient cars built. That same model has actually porked up since to meet consumer desires. Consumer desires at $3 per gallon are somewhat at variance with those same consumer desires at $7 per gallon.
But the big gains to be had are in the Escalades, Ford F-150’s, and Suburban’s. Not in the Hondas. In converting a Honda, John will wind up with a very nice car. But he could make more of an imprint on the world doing an Escalade. The Escalade gets 10 mpg in town. It has a 26 gallon tank and the Tesla Model S has a greater range.
My choice is carbon fiber. And I can assure you you could take several hundred pounds off the Honda with it. It makes no economic sense at the moment. But if millions of cars used it, the price of carbon fiber would come down right smartly. This is the direction BMW is going. It makes no economic sense at the moment, but they are putting a bet down that it will soon.
All that said, the battery is an obvious area of improvement and it is my position that those improvements are more driven by battery demand and economics than they are on “scientific breakthroughs.” I see a lot of science laying around undeveloped because there is largely NO market for the product. We have two BANKRUPT American lithium battery companies laying on the ground after hundreds of millions of dollars – like nearly a billion, in public funds invested along with an even larger investment of private money. That does not augur well for future venture capital activity.
A lot of people drank the kool-aid. But the battery market didn’t show up. GM isn’t doing 120,000 Volts per year. Nissan isn’t doing 150,000 Nissan’s a year. Those were the numbers bandied about going in. Didn’t happen. Where did you hear it first?
Along Jacks comments:
“If no motorcycles had previously existed and we designed one from scratch, it wouldn’t look anything like what we’ve got now. Maybe the whole mechanical layout would be different”
Design Director – Ducati
I liked the tv this time, here is an interesting analogy; Top and Bottom balancing are like measuring a pint of beer in different places. I would say my pint is not full if the top of the head was on the one pint line on the glass, but I would if the bottom of the head was on the line. Top levelling is measuring the frothy head on top of the beer instead of the beer itself. Bottom balancing is taking the reference from an empty glass then adding a dose to it. You may still get a head on it, but you know how much you measured in before it frothed.
Also I work for Ford and can confirm that the gasoline, hybrid, plug-in hybrid and electric Focus and C-Max are built on the same production line in the same factory. Ford can then adjust the build mix according to the market. That is where they have an advantage. Also sharing common parts keeps the cost down.
Not a bad analogy. Mine is a little different. Let’s take 36 pint glasses and make sure they are all EMPTY to begin with. Then we put just shy of a pint in each one and let the head do what heads do. As long as we drink from each glass equally, we can then adjust the level any time we want, as long as we don’t put in more than we drank, and we do it to each glass equally, we don’t have to worry about the head.
And the serious damage appears to be if one glass gets empty and the others are still a quarter full. Broken glass.
The beer glasses being narrow at the bottom.
Very wide but parallel in the middle and narrowing very quickly indeed at the top.
None exactly the same capacity.
The exact shape and height of the beer glasses vary with temperature.
Clear as Guinness?
Talking tyres (mentioned earlier)
It’s already sorted for cars but motorcycles cannot. In fact looking around for any sensible resources on R/R on M/C tyres over widths, makes, diameters and types. Can anyone help?
Last crazy thought on how to use existing inverters. With an amperage limiting device on the output from you car battery, you should be able to replicate the DC input from a solar array into a inverter/charger that could then provide 240V AC. You would need the minimum 48v battery needed for the inverter charger.
I got an interesting offer in my mail today from our local Nissan dealer: Buy a new Armada, Maxima, Titan or Murano at MSRP–and get a free two-year Leaf lease (12,000 miles per year).