TORRID HEAT, BITTER COLD, and Gliding About in an Elescalade. Announcing EVCCON 2013

THis week we talk a bit about several small things. Heat seems to be a pretty simple topic. It is hardly in the ballpark with range issues – right?

Actually, at least for us, the range issue is almost entirely manufactured by the media. Heat in an electric car, on the other hand is very real. It becomes more real as the temperature indication drops. And indeed at about 20F all thoughts of range kind of go by the wayside.

In fact, one evening in the Mini Cooper Clubman, my wife and I went to dinner. I left the car “running” for an hour and a half while we dined with the heat on. This so the car would be warm when we left. Unfortunately, when the ignition was on, the running lights were lit as well. We had four different parties stop by our table to advise us we had left our lights on, concerned that we would run out of battery and be unable to start our car.

We had a 34 kWh battery, and we were dining about two miles from the house. The temperature was 22F. Range wasn’t the topic of the evening. Nor did I car if the battery “ran down” at about 8 amps and 360v. What I cared about was the car was warm when we came out. We could charge at home.

[jwplayer file=”news022213-iPhone.m4v” hd.file=”news022213-1280.mov” image=”http://media3.ev-tv.me/news022213.jpg” streamer=”rtmp://s3einxnpkaij93.cloudfront.net/cfx/st/” provider=”rtmp” width=”950″ height=”584″ html5_file=”http://media3.ev-tv.me/news022213-1280.mov”]

This week the Cadillac Escalade EXT. This has not been a problem free build. In fact, it is probably the hardest and most complicated we have tackled and we have yet to declare victory. But I have to tell you I am driving it daily and it is just a joy for me personally, even clouded by the vast amount of ducats it has eaten along the way. Most of the time it is very smooth, quite powerful, very quiet, and the six speed automatic transmission works beyond my fondest hopes. I recall a LOT of our viewers simply “advising” in their vast experience with electric Cadillacs that we would never get this transmission to work. Thanks for the vote of confidence guys.

We DO have ongoing issues. Something periodically upsets the ECU and it goes into a coughing fit at 3000 rpm any time we hit that magic mark. At other times it does not display this AT ALL. ANd it is really quite random. We have fixed it a dozen times. We change something, it goes away for two days, and then comes back. That’s because it is very intermittent anyway. Frustrating. Annoying. But not terribly important even when it is occurring.

Heat is something else. We had devised kind of a complicated system using a very high power whole house tankless water heater system. Great lesson in there. It was designed apparently to be in a house. Reminder once again that the automotive environment can be harsh. It lasted days. And with two heater elements, one blew up, and then the other.

In this episode, we go to a bit of a more robust heater system – a liquid heater from Kim HotStart that is designed for diesel trucks, caterpillar tractors, and diesel locomotives. They have been making this device since 1942. It’s basically a cast aluminum tube with a water heater element in it.

Our control system remains problematical. After the show was edited, we had a total failure of the temperature probe. It’s easy to replace, but it was supposed to be immersible and accurate. They didn’t mention the part about it lasting a few weeks. I’m hoping it was a fluke.

The system is both expensive and a bit complicated. Glycol/water heating has great attraction because you can leave your cabin environmentals entirely alone. You’re basically just spoofing the engine heat normally supplied. But you have to circulate it. You have to control the temperature of it. And it takes a lot of power. We are currently doing about 3700 watts in the Cadillac and it may not be enough. I’ve ordered a 4000 watt 240v element to replace one of the 2500 watt elements.

You also have to be very careful that it doesn’t run away with you. The penalties are severe. Unnoticed, if it somehow is left on it poses a dire threat to your battery pack as this is a significant load and if not detected, it will drain your pack to 0.000v overnight. Even a 76 kWh pack.

Second of course is the fire hazard. PLastic and high temperatures don’t mix. Fortunately, we have a solid cast aluminum tube that on full meltdown can’t really meltdown.

A number of viewers, ever helpful, have pointed us to the MES-DEA RM4 system. Actually we installed one of those in the Mini Cooper 3 years ago. It went entirely dark within a month. We had bought it from Victor at Metric Mind. WHile he ordinarily provides probably the poorest support and the greatest amount of excuses of any component supplier, in this case he replaced it immediately. In fact we had one before we could remove the old one and return it. While he nornally disclaims any knowledge of his products and definitely any responsibility for them, noting he’s just a middle man, in this case it went very easy. Turns out he WAS familiar with the problem and we werent’ the first replaced.

The replacement seems to work well and we’ve had no further difficulties. But one vehicle fire has been blamed on this device. An MEA-DES has discontinued it, promising a corrected version in the future. They were then absorbed into another company and nothing further. Victor is discounting his stock of some 60 remaining units at a very attractive price. We’re going to pass.

Our little Lang solar hot water system pump continues to annoy with it’s very high price, and amaze with its hardy willingness to move the liquids entirely silently. As you know by now, I have a weakness for good components and this one rates as EV jewelry.

Another area where we’ve just had famous experience with is Summit Racing’s black aluminum AN fittings. When used with the black nylon rubber hose this removes all my problems with leaky hoses and the cheap band hose clamps that plagued the vehicles of the sixties and seventies. This system is just leakproof. We have all of that in our cooling system kit. If I can work something out with HotStart, maybe we’ll do a HEATING system kit.

In this episode we also do a survey of the Azure Dynamics DMOC845. As you know, we purchased a large lot of Siemens 1PV5135 motors and AZD DMOC645s. We’re working on a generalized vehicle control unit to control it over CAN bus. Well, Collin Kidder is doing most of the work of course. And I’m doing what I do best, approving of all that work and cheering him on.

But AZD also received a $5.4million Department of ENergy grant to develop an “advanced traction drive” for the future. This is simply careful wording of a grant application. What they were doing was simply developing their next version of DMOC. You can really afford the very best components when you’re doing it with other people’s money and they spared no expense. Actually I’m since advised that Rinehart Motion Systems also uses the SKiM power module from Semikron. You can find these online from $1800 to $2300 onesey/twosey. I’m told that AZD had a stash of $1.8 million worth of these but I haven’t seen them surface.

In any event, modules like the 93mm SKiM and the Sprague SBE film capacitor, dramatically simplify the development of a three phase controller. It still needs some software and a multicontroller, but the controller used in the AZD DMOC845 is about on par with an Arduino Due. I think I see on the horizon the end of the $10,000 controller market. There should be something in the $4000-$5000 range for 150kW soon. With a GEVCU at certainly less than $500 somewhere, and a DMOC645 at $2500, I think we’ll look pretty much a bargain at $3000 for some time. At least until Curtis works out the concept that there is a market for a 150 kw controller. Man are those guys slow on the uptake. But the little AC-50 with Curtis controller is a relative bargain with the motor and controller both for less than $5000. The 75 kW restriction just eliminates too many cars, but we love that system for our Speedsters and so forth.

That’s about to get better as well. We’ve already sold our AC-75 prototype. We’re supposed to receive two production units this week. The AC-75 is a much larger motor physically – about the size and weight of the Siemens and really with a very similar power rating. While the controller is limited to the same 75 kW, the motor is rated for 180 ft lbs at 108volts. We run these at 120v and the torque has to come out very close to the 220 ft lbs the Siemens delivers at 300v and 400amps.

On the topic of temperature, the Argonne National Labs has set up an Advanced Powertrain Research Facility. And they are doing some great testing there. They recently put a Leaf on a dynamometer. This is a very good technique for checking range etc because you can program the system to drive the car to a defined route that just never varies. Actually it varies all the time but you can do da drive and then duplicate that exact drive. This is impossible in the real world as lights and traffic are very dynamic.

In this case, they put the Leaf on the dyno for several different drive profiles. The variable was the temperature. They ran it at 20F, 72F and 95F with a simulated 850 watts/square meter “sun” on the 95F run. The concept was to determine the effect on range at different temperatures.

I love this test. And I don’t. First, I’m very curious what would happen to the batteries themselves. BUt it is also interesting to see the effects of running the heater and air conditioner. This of course depends on how powerful a heater or air conditioner was installed on the car. But assuming it was adequate to maintain the 74 degree cabin temperature, it is very interesting.

LeafCold

LeafRange

LeafEC

Why all this temperature talk? Well, it’s February here in southeast Missouri. I intensely dislike the environment here during two months, February and July. In July, it is too hot and humid to bear. And in February, it is too wet and chilly. I would go for a ten month year here, but I fear the young socialist democrats on CNBC are the only ones hopelessly naive enough that I could get to take up such a notion.

I have to tell you I was a bit reluctant to do an Electric Vehicle Conversion Convention this year. We had a great convention last year but a number of key individuals from the previous year didn’t make it. It was a weird year with an election and a messed up economy and so forth so a number of guys I assumed would be with us just weren’t able to come. True, we nearly doubled in body count and vehicles and the ShowMe Center was a great venue for what we were doing. But these events are very stressful on a small organization. Brian left in August. And you have to work at it pretty much full time for six months, then all of it hinges on a few days of the event itself. Anything can wipe out your careful plans. Airline strike, terrorist attack – even if its somewhere else, weather, anything. And you’re still on the hook for all the expenses and of course all the work you did leading up to it. The vendor support is tepid. Indeed I wondered if I wasn’t just being taken advantage of.

Nick Smith was the first to sign up for the first EVCCON. No one suffers more difficult nor expensive logistics than Nick. But he’s got the EV fever pretty badly and is struggling to get a components business off the ground in Auckland New Zealand. He was ALSO first to register for EVCCON 2012 and showed up early bringing friends. He wants one for 2013.

So I decided to throw the vendors under the bus – they can come or not as they like and find convenient. We’re going to host an Electric Vehicle COnversion Convention for Nick Smith and in fact, I’ve invited him to be our keynote speaker. John Hardy of Surrey in the UK immediately signed up to speak as well. And so we are hoping to do it again.

I’ve selected a bit smaller interior space, the AC Brase Arena, with a large surrounding park. And an often hot time of year – August 6-11. I always say if we can make it to August 15 the heat is mostly over. Unfortunately August 6 is on the wrong side of August 15, so it might be a bit warmish.

Other differences. We are expanding playday from half a day to all day Friday. We already have the dynomometer guy eager to come back and do it again. He was just fascinated with the electric cars last year. We’re going to shorten the drag strip from the quarter mile to an eighth mile run to provide more rundown space. This should let the fast cars “let the dog out” without danger of winding up taking off on one of our runways.

I have already heard from several of the missing warriors from the first year who intend to return now. I don’t know what size it will be. But we are looking hard for ways to make it more enjoyable and meaningful for those who do attend.

One thing we’re going to do is recognize the guys who DO go through the heroics to bring their cars. In a sense, they are cohosting this event. And so indeed we are going to recognize this. No, they don’t have to pay a portion of the shortfall we seem to run every year financially, though that would be a tempting notion. But we’ll have some different shirts and caps etc for this group so the rest can easily identify them. Maybe even numbers for their cars and shirts so you can easily match them up for advice and questions. When you stop and think about it, they are and always were the stars of this event.

I’m always torn about this event. Having done trade shows in the past, I am naturally keyed to seeing it grow to success with 5000 attendees and a 100 vendors. Probably have to move to the Bay area to get that to happen. But the fun for me is 35 guys in the shop, all of us studiously trying to figure out what’s wrong with Jason Horak’s car up on the lift.

It remains a mystery…

Jack Rickard

68 thoughts on “TORRID HEAT, BITTER COLD, and Gliding About in an Elescalade. Announcing EVCCON 2013”

  1. Glad to hear you are going to have EVCCon2013. Sorry to say with the move to August, I’ll not likely be able to make it but I’ll try to see if I can. It’s processing season for my agriculture customers and they like me to be very close for emergency support. Hopefully EVCCon2014 in October.

    I also saw this… http://www.sinopolybattery.com/ClientResources/201302271823132.pdf

    I believe it says the Mr. Winston Chung has gone bankrupt. I’m not quite sure what it means about Winston Battery Co. Seems this saga has gone on for way too long.

  2. If You want to be happy with Your heating system for a long time, You should replace the contactor by
    an IGBT as even very strong contacts burn down very fast when switching DC-current.
    They easily withstand this current when closed, but tey cant switch it very often!
    (At voltages around 200V it is not easy to find some FET with enough power, so an IGBT does the job)
    If there is no PWM it is even possible to simply switch it on and off, but a little driver is the better solution.
    I did some tests with IXYS MII 400 – 12 E4 IGBTs and as driver i am using TC4451 or TC4452 and
    10 Ohm in serial to the gate. This can easily switch 200A in my homemade controller at 8 kHz
    Most of those MOSFET-drivers switch off at a voltage lower than 7.5V.

          1. I don’t really think that the SCR is better than the IGBT. I find the the IGBT is more efficient, but requires more protection in the circuit for noise. The SCR is a little more stable. Typically the SCR is a little less expensive as well…

            I guess i have typically used the SCR in resistance heating applications….. I am not really sure it is necessarily any better….

          2. Ok I thought about it more… I had a brain fart. My circuit only work on AC systems where you can shut the gate off as the 60/50hz AC wave crosses zero. This is what causes these to generate virtually no EMI issues as you effectively switch them off with no voltage applied…..

            So never mind….

  3. I am looking forward to EVCCON 2013. I think I will have some interesting “things” to show people this year. I am going to get the EVThing there even if I have to push it….

    I am in the wiring phase of my build and the PLC is turning out to be a good little brain for the car. I am doing things like an interior and external temp. I wire this into the control circuit for the charger. That way it will not charger the car below a setpoint temp. Believe it on not, it does often get below freezing here in Kennesaw GA, However, my car will be in an insulated (not heated) garage. However, when I park my gas guzzling F150 in the garage it’s radiator will raise the temp in the garage by 20-25 degrees allowing a short charging window. It will also monitor the pack voltage just to make sure the charger behaves itself…

    I have even decided to bring a couple of PLC’s to the show and give away as prizes (however jack sees fit…). I’ll make sure that they have the necessary I/O included and the programming software. I even have a new one with a GSM texting interface. This allows you to insert a standard GSM mini SIM card and send and receive text messages from the car. You can even use pre-paid plans from At&T and T-Moble. You can program the messages and responses. For example you program it to respond to “Heat on” or “Heat off” to start and stop the heater…. Almost anything you can think of…. I have one sitting on my desk now testing it….

    Since these PLC’s have PWM/PTO outputs, it is possible to get them to act as charging, heating and inverter controllers. Once I get the car running, my first plan is to get the PLC to actually control a buck style charger….

    P.S. I think we should start a pool on what Jason will break this year….

    1. I’d love to be able to bring my Civic to EVCCON – a few trillion tons of cold salt water in the way though.

      BTW minor correction – I’m from Worcestershire (we pronounce it “woostershire”), not Surrey. My fault for living in a place called Redditch (a characterless but efficient dormitory largely developed by Soviet era bureaucrats) which gets mistaken for (among others) Redhill in Surrey

  4. Also meant to say that I wholly agree with Jack on the naivety of attempting to top balance batteries on the basis of voltage. There are at least three factors (there might be more) affecting a cell’s voltage –

    1. state of charge,
    2. internal impedance or quasi impedance (behaves like a series resistance but is not fixed or wholly predictable in the way that ordinary resistance is)
    3. Diffusion delay (a decay or recovery of voltage that occurs in minutes or hours after charge or discharge ceases)

    I have seen cell voltages writhing around like a bundle of snakes at the top of charge – rising and sometimes falling at different rates in the constant voltage phase of charge. If you just leave the poor innocent things alone they all fall back into line when they are allowed to rest a while.

    1. John,

      That is why I am so grateful for your work. Your “Data” kind of proves beyond a reasonable doubt that most of what people “think” about these batteries is simply wrong.

      The scary part is that it also kind of means that Jack is right!

      1. Many thanks for the affirmation Jeff. To give credit where credit is due I would probably not have bothered to look (and just bought a BMS) if it had not been for Jack’s efforts in pushing back against the received wisdom of a few years ago, allied with the willingness of many in this community to listen to him

  5. This is still the ground floor people! The OEM’s may be developing electric cars but the pricepoint will still be a premium for years to come. NOTHING is more satisfying then converting a car yourself. With people like Jack and Richard at the helm its sure to be an exciting ride. This year needs to continue the fondness that I grew to love from EVCCON 2011 (requires Jack to drink)! I cannot wait for the years event. I wish it was tomorrow, or wait, no I dont, I still have batterys to relocate and tons of little stuff to do. But I love every minute of it! Thanks Jack for continung this show and providing us with your services.

    All the best,
    Aaron Lephart

  6. RE the heating issue: I’m currently playing around with infrared heating film for a flat that I am renovating and I am pretty impressed by the stuff. It’s basically a carbon resistance heating element, emitting far infrared radiation and it can be bought very cheaply from Chinese suppliers. Maybe it would be possible to use it under the door panels and the headliner as an alternative to the classic convection heating approach which is the obvious solution for ICE engines. I also think BMW is considering this for their not existing i series.

  7. Regarding the pool, I say “the sound barrier”! I met Jason at the first EVCCON, and he is one of the nicest guys in the EV world. I was about 50 ft away when his Daytona spit a CV joint out into the grass on the final right-hander of the TT course. I didn’t catch the fire of EVCCON 2012, but it seems that Jason has a flair for dramatic destruction. That being said, he also has a knack for figuring out how to fix his car quickly and get it up and running! Okay, I’ve nothing more to contribute, just wanted to give Jason some props….

  8. Hi Jack,

    Forgive me if this has already been brought up but I’m wondering if it is possible to combine your heating and cooling systems so they help each other out. You have a coolng system for your controllers where you are trying to get rid of heat and and a heating system where you are trying to make things hot. Is there any way of using the heat from the controllers to supplement the heat for the cabin heating system or is that too complicated?

    Like an old friend used to say to me “Carry on as if you were normal!”

    Cheers,

    John

    1. John one thing about solid state devices is, the cooler you can keep them the better. You would be running the controller at heater temperatures which could be as high as 200 Farenheit, since it would be run in series. I would think that the controller temps never rise high enough, to even add to that, and probably runs a lot cooler then the water heater.

      So far the electric water element is the best idea since you can use the existing heating system, unless one can convert to ceramic heat.

      Also if the voltages are close, one could make a water heating system with existing off the shelf parts. Home electric water heating units come in 120v and 240 volts at different wattages. Using water pipe with a screw in home water heating element is possible, if constructed properly. One thing that one needs to be aware of, is to be 100% sure that the heating element is covered with fliud, otherwise it will burn out very shortly.

      I dont know how many times I’ve seen someone change an element in a water heater unit, and turn on the electric before water….Ooops.

      Roy

      1. You would need another heat exchanger after the cabin heat exchanger in order to dump the heat that is not pulled into the cab in order to meet the inlet water temperature requirement of the motor/controller.

  9. Jack,

    While watching the trials and tribulations of the Escalade heating system it occured to me that since you are running a zero pressure system it could be of great advantage to run one of the waterless coolants. Evans is one brand that comes to mind. It boils at 350F (180C) unpressurized and would allow you to run your coolant temps over 200 degrees like you would find with the original system once warmed up. Your failed coil might have occured because of a localized hot spot that allowed the coolant to boil which leads to an almost immediate thermal runaway at that spot. And this kind of coolant doesn’t wear out so you don’t have to replace it every few years when the corrosion inhibitors are gone. Because the thermal transfer ratio is slightly lower than water/glycol I would probably not use this in your controller cooling loop but for heating the cabin and your batteries there are big advantages. It freezes at -40 which is the same as properly mixed water/glycol but it does not expand when it freezes so no ruptured components if that ever happened.

    Looking forward to registering for EVCCon 2013. Planning to bring the Mazda again only more sorted out this year. Just broke 1200 pure EV miles today.

  10. Here’s a good article on the Boeing battery ongoing saga. This one has some very telling details.

    An explosion in a lithium battery under development for use in the Boeing 787 Dreamliner resulted in a fire that destroyed the Tucson, Ariz., facility of manufacturer Securaplane in 2006.

    Following that incident, an employee filed a whistleblower lawsuit with the Labor Department alleging the company had produced an unsafe battery. Labor dismissed the suit on procedural grounds.

    Internal documents from the Pipeline and Hazardous Materials Administration show that Securaplane tested a battery from one manufacturer for use in the 787 but substituted another for use in production systems based on assertions that the new batteries were identical to those tested.

    Michael Leon, a Securaplane engineering technician, was working on the 787 battery charger unit when it exploded in 2006. He complained in January 2007 that the units did not conform to design schematics, according to a final decision and order Labor issued July 15, 2011.

    According to that 56-page document, Securaplane won a contract for the 787 battery charger unit in 2004. In November 2006, while Leon was running tests, the battery exploded, resulting in “a devastating fire that destroyed [Securaplane’s] labs and production building.”

    Despite his concerns, Leon said he was “pressured [by Securaplane] to sign off on Acceptance Test Procedures (ATPs) for noncompliant units and ship them to customers. He says he gave in to pressure to run the ATPs, knowing they would be re-run after he fixed the units, but balked at actually shipping noncompliant units,” according to the documents.

    Continued…..
    http://www.nextgov.com/emerging-tech/2013/01/2006-battery-fire-destroyed-boeing-787-suppliers-facility/60809/

    1. Mr. Dove, I’m receiving a blank page 🙁
      They eventually sacked Michael Leon for being a trouble causer and they stated the fire was his fault because he did not connect the BMS pigtail to the board. He was the guy who told the world about the issues after his sacking.
      A faulty board should “open circuit” the whole pack as de-facto. Not allow it to carry on charging until destruction.
      The whole business should of been taken to task over the design. Instead they who take responsibility signed it off as fit for aircraft use.
      Nobody is getting sacked.

        1. Thank you for the paper DRdove.
          HR gave a fair round-up of Leon’s severe character flaws. Doubts of the company actions by me were high. Not now.
          .
          Page 20 onwards states Leon was correct. The schematics did not match the (corrected) circuit boards. Any and all non-conformities must be straightened before any approval.
          However, Leon did not Tell Dr. Cheng of the error but made a holla and a whistle of it then went to the FAA when he saw the writing on the wall.
          .
          Nonetheless, all that hullabaloo was over some company badly re-inventing the square wheel.

          1. All a matter of how the data is viewed IMO. It said hardware and documentation was lost in the fire. He could not prove much of what happened and after the fire everyone was trying to cover their ass and he most likely felt like a scapegoat.

    2. Very interesting.

      I note:

      “Securaplane, along with Thales and GS Yuasa, conducted a two-year investigation and narrowed the cause of the fire to one of three possibilities: a defect in one corner of a battery cell; prior damage to the battery; and Leon’s failure to connect a harness that transmits signals between the battery and the charger unit during the test…”

      Or how about it WAS connected and the BMS (for that is it’s name) failed to shut off the charger……

        1. It hadn’t been running long. I wonder about a short in the rat’s nest of BMS wiring: “….The stainless steel sleeve and signal wires had damage consistent with excessive electrical current where they attached to the connector s at each end…” (Page 10) “…at 1021:37, the voltage of the APU battery began fluctuating, dropping from a full charge of 32 volts to 28 volts about 7 seconds before the shutdown (Page vii)…” Of course there it could be that these observations were the consequence of the fire not the cause of it

  11. An interesting aside to Broders test drive of the Tesla he flatlined. You know, *that* man who does not test cars for a living but speaks out for the power companies.
    New York Times deletes its environment desk.

  12. facebook_musicmjb

    Just have two things to say, did the companies which you mentioned which were not given the same monies to increase their R&D apply for a grant? If so your statement about unfair comptition has validity. Two Investing in Thomas Edison at the time would have been considered “Junk Status” is ths a proper measurment?

    1. Probably not, which actually has no impact on the validity of the point, and there were many who did invest in Thomas Edison, with very mixed and mostly negative results. He actually was not known as the financial wizard of Menlo Park.

      In the end and on balance he probably did make more than he lost, but it was a close call. He lost several fortunes along the way and would have truly been rated junk status. While he did have some government contracts, he was never given a grant or a DOE loan of any kind.

      Jack

    1. There is a basic disjuncture in most of these analysis by battery guys. They apparently don’t know what an APU is. The battery is not NORMALLY a power source. It is a starter battery for the APU and could conceivably be a backup power source. The APU is a small turbine engine that runs a generator.

      In the air, the normal jet engines have large generators and a number of them to provide electrical power. The problem is on the ground in the heat, we need air conditioning and so forth and at low idle RPMs the jet engine generators don’t work. So we start up a dedicated turbine/generator termed an Auxilliary Power Unit. The batteries start the APU. The APU makes the electricity.

      It is kind of clear to me that they are using the batteries to load level the 32v output of the generator.

      It is also clear that there was some SIGNIFICANT swelling of SOME of these cells and not others. That comes from overcharging guys. That’s the ONLY place that comes from. Now with TWO BMS boards (BMU) and a Battery Charging Unit (BCU) and two contractors, how in the hell did THAT happen.

      32v on 8 cells. Current IN to the batteries. Your cited analysis states that you can CHARGE to 4.2v. That would be 33.6 volts. Yet the FULLY charged voltage of these cells at 3.7v would be 29.6v.

      They are routinely charging these cells as if they were lead acid cells on a trickle charge. This is because they think the 33.6volts is inherently signficant. It is not. It is a procedure. IF you charge at any current level to 4.2 volts and then HOLD it there until the current diminishes to 0.05C and then TERMINATE. You will magically find these cells fully charged. You can’t measure this until later. Do NOT confuse that with anything else. The fully charged voltage of the cells would be 29.6 volts.

      If you put them on a 32 volt at low current levels continuously, you THINK the cells are fine because you haven’t overcharged them. Is THAT the recipe provided by the manufacturer?

      No it isn’t. It’s lead acid thinking run amuck. IF you want to hold them with a trickle charge, and I don’t, the voltage would be something LESS than 29.6v. That’s the fully charged voltage of the cell.

      The belief system is that they can hold them at 32 volts indefinitely with current because this is less than 4.2v per cell. That’s because they BELIEVE that the fully charged voltage of the cell is 4.2v. It is NOT. That is a TARGET voltage for a PROCEDURE that has other parts.

      We COOK our chocolate cake at 350F for 30 minutes. That’s not 350F as along as we find convenient because 350F is cake temperature. And no one wants to eat a cake that is 350F. It is just a PROCEDURE for making chocolate cake.

      That holding them at 32v and low currrent levels doesnt’ overcharge them VERY MUCH doesn’t help guys. The photo tells the tale. They are overcharged and swollen. And the BMU and BCU purport to not have failed. I believe they have not failed.

      I would think they have DESIGNED this to overcharge these cells. And I find this kind of thinking RAMPANT among otherwise intelligent engineers who are trying to deal with these batteries.

      Ironically, it is 32 volts NOT because they need 32 volts. Aircraft systems are nominally 28volts. If they operated the cells at 28volts, they would never fail.

      This lesson doesn’t matter much for electric cars. But we are gaining viewers from the solar community and it DOES matter for them. They are wanting to basically trickle charge LiFePo4 cells. They must do it below the cell voltage. In our case ,that is 3.38v. If you must trickle, be below that. Like 3.25-3.3v. You will reach equilibrium below the open circuit voltage of this chemistry.

      So I don’t think there was a signficant event. I think they were routinely holding these cells at 4 volts thinking that was well under the 4.2v CHARGING voltage. After a certain period of operating time, the cells were gradually but increasingly overcharged. They swelled, and they shorted. Who’da thunk it.

      Jack Rickard

      1. So for example a four-cell LiFePO4 pack as an SLI battery could be left permanently connected to (say) a “12 volt” system running at 13.2 volts but the usual 14.5 would break the bank? It would be an interesting experiment…

        1. And easy to verify John. Hook up a cell (one of your old headways will do) to a power supply at 3.625 volts. Remember the charge voltage is 3.65 so if their theory is good, it should be no problem right? Now leave it hooked up for two weeks and see what happens. These JAL cells had nearly 100,000 flight hours on them. But I think you should see smoke within two weeks continuously.

          Please do this in a metal container out on the patio or somewhere with no flammables and NOT in the house. Not even in your shed.

          There is not really a ready solution to this as an SLI in a regular automobile. In the case of the Boeing, the solution is trivial. Go to 9 cells. They would be at 3.55v then and their open circuit voltage is higher than ours at 3.7v.

          Ironically, you COULD make an SLI battery out of Lithium Cobalt or Manganese. 4x 3.7 = 14.8v.

          This is all basic math. The interesting thing here is that they are sneaking up on a fire in slow motion. This is why it doesn’t show up in their testing, other than the fact that they DID burn a factory to the ground along the way – clue in there somewhere.

          Charging 8 cells with an open circuit voltage of 29.6 volts at a 32 v rate doesn’t produce much charge current. But it does produce charge current. If over time your withdrawals from starting the APU do not exceed your runtime at 32v, you ARE going to eventually overcharge the cells. Evidence the photographs of the cells. But it won’t show up on the first flight. Indeed, slightly overcharged cells continue to operate for some time. This is a creeping situation where they gradually overcharge and gradually swell and continue right up to the point where they don’t. And then this failure. What would the Battery Monitor Units have noted. Remember they DESIGNED this for 32v on purpose. What would the charger do. It isn’t even charging. So how did they GET overcharged? About 2.4volts at a time.

          And absolutely you could create the same fire with Lithiums as an SLI in a car. Undoubtedly somebody will. If they use LiCo or LiMn they will be fine. ANd if they sell LiFePo4 they will NOT be fine. I’m not sure how the modern built in regulators to the alternators work these days. But if they are on continuously at 14.5v, they will burn up a LiFePo4 four cell group in time.

          The heart of this matter is that we CANNOT use voltage for SOC WHILE charging. We never could. It takes a day of rest to even be able to READ the true voltage of the cell. So I suppose we could put in 2 Ah, wait a day and measure, put in 2 more AH, wait a day and measure, and do this for a month or so until the cell was right at 3.8v. Cool.

          The procedure can be developed experimentally. BTW ALL and I do mean ALL the LiFepo4 battery manufacturers were advising charging to 4.2v initially. There is no magic in this number guys OR in the 3.65 volt number. 4.2 was originally chosen because it was safely under 4.5v. At about 4.5v we start to have some undesireable interactions between voltage and electrolytes used in these cells.

          So now your mission is to devise a charge procedure that will produce a full charge on the cell. It’s simple. Pick a voltage LESS than this 4.5v, charge it constant current to that voltage, and then charge constant current and HOLD that voltage until the current decreases. Pick a point. Till it decreases to half. Now shut it off and wait a day. Now measure the cell voltage. 3.3 volts is too low. 3.8v is 100% and we know this by knowing the negative potential of the anode and the positive potential of the cathode by chemistry, and adding the two of them together.

          If your resulting voltage is too low. Then do teh charge again but let the current taper further down. Let’s say to 25% of the original charge current. Wait a day and measure.

          You can develop your OWN charge sequence at ANY voltage from 3.8 to 4.5 but your termination point will move around. Instead of 0.05C, it might be 0.2C. Or 0.03C. Everyone dismisses the part about STOPPING because it is at a low rate. It is IMPORTANT TO STOP.

          Once you have this, you simply give it to your customers. FOLLOW THIS PROCEDURE and your battery will wind up at 100% SOC. (Pad it a little so they don’t overcharge). Thundersky originally claimed 4.2v. Then 4.0v. Then 3.8v. Then 3.65v. NOTHING CHANGED but the procedure. People were overcharging the cells.

          And they watch EVtv. of course.

          I caught them early in the game selling a FIXED charger for 24 cells – 3.65 volts. That’s where that number came from. Today, it is the number everyone uses. It’s just a number. Chocolate Cake. 350F for 30 minutes. 375 – 25 minutes. 325 – 45 minutes. But after you cook it, Chocolate cake is NOT inherently 350F.

          So IF you connect a cell who’s fully charged open circuit voltage is 3.38 volts to a 4.0v source, at what point will current STOP. Once it is at 4.0v, it should end. Will it? No, the current diminishes, but it will never reach zero before the fire.

          Jack RIckard

          1. Here’s another experiment I performed accidentally and at some expense. Lots of charge electronics in laptops of course. PLug one into the wall to recharge. DON’T use the laptop at all Leave it for 3 months. Come back and inspect the battery. Try to do so before the fire.

            Looks a lot like the old JIFFY POP commercial. And you do get to buy a new battery.

            Jack Rickard

          2. Yes four cells and an ordinary SLI charger would do it. I’ll buy a metal box from ebay and shove it in a flower bed with a LONG wire

          3. David Seabury

            So do most of the chargers out there have enough historesis to prevent restarting a charge after the voltage drops? Or do they just stop until reset? Just curious?

          4. David Seabury

            Let me try to clarify. If a charge is completed and the pak settles back to the nominal voltage is there any danger of a charger starting up again.

          5. David:

            In ALL cases of a proper lithium battery charger, once it terminates it is done. Future changes in the battery voltage won’t do ANYTHING.

            If you remove AC power, and then reapply it, this RESETS the charger and it starts again.

            Some multipurpose battery chargers not specifically designed for lithium DO have “float” settings that can seriously damage your pack. The Manzanita does this BUT it has a dip switch system you can set up to prevent that.

            Jack Rickard

          6. JOHN. Depends on the SLI charger but yes. If it charges to 14.5 volts it will have this effect. I have seen some that charge to lesser values like 13.2v. Again, open circuit voltage of 13.6. So you would want to determine the voltage the charger charges to. But a 14.5 alternator would certainly overcharge them. And a 13.2v charger certainly would NOT.

            As an experiment, I would use a single cell, and and adjust a 0-12v power supply for exactly 4.0v. It might be nice to start with a fully charged cell and a timer that would note the time from the start of the charge untl the fire becomes uncontrollable.

            Jack Rickard

  13. @ Dave Seabury – FWIW, here is a voltage and current trace on an eight-cell Headway pack that has had the charger restarted a couple of times. It isn’t exactly a standard scenario because the CV phase was virtually eliminated. You can see from the current trace that it enters the CV phase almost immediately and the amount of extra charge crammed in (the area under the current trace) is not great.

    http://tovey-books.co.uk/attachments/Image/multicharge.jpg

    1. John,
      My concern is with a car left on a charger for months that periodically restarts and adds just a little more charge each time until one day it bursts into flames.

    1. Apple MacPro. I agree. Any laptop should be able to sit forever plugged in. Try it. ALmost none of them can be.

      Usually you just wind up with a battery that has no capacity. But some will jiffy pop.

      Plug one in for 60 days and just don’t open it. Almost guaranteeed.

      Jack Rckard

    2. Heater again.

      I dont know how fast that solid state relay will switch but it should do 1 kHz easily. Dont use Arduino’s analog PWM for this, do your own PWM switching about 400 Hz to 1 kHz and tune the pulse width to derate the power feed to the heater so you can feed a 120V heater from 180V. If you can spare an Arduino for regulating the heat, dont use the main ECU for this, you could even do analog control the heat.

      The analog output PWM is way too fast for the solid state relay that would actually analog drive the relay and that would get you more heat in the transistors than in the heater.

      Failure mode – if the relay gets stuck the heaters will act as a fuse. You could actually put a fuse in series with the heaters.

  14. Happy….Happy…Joy…Joy our Brain is backkkkkkkk!

    Other than my original EV grin… I’ve never smiled a bigger EV grin until I saw that Brian had been found.

    Jack I’d like you to extend my thanks to the local milkman for his determination in finding our beloved Brain… 🙂

Leave a Comment