PDA

View Full Version : What stinks and sizzles?



Richard230
23 October 2010, 1441
So today I rode my GPR-S around for about 15 miles before replacing the dead cell in my battery pack. When I arrived home I noticed a strong pungent odor that smelled something like lacquer thinner and heard a sizzling sound coming from my bike. I immediately thought it might be on fire and removed all of the fairing covers as rapidly as possible. By the time I had access, the sizzling sound had stopped and the odor was slowly dissipating. I checked all of the batteries that I could see and nothing seemed burnt or melted and I could not locate the source of the bad smell.

I then pulled out the bum battery and replaced it with the one that I received from Harlan. The old shorted-out battery was swollen slightly and had to be pulled out a little at a time as it was kind of tight against its neighbors. The replacement battery dropped right in and I reinstalled the BMS/conductor board. I checked the at rest voltage and it had risen from 74.6 volts (after using 20 Ah) to 77.8 volts. I then charged the battery pack and it is now performing its balancing act. Everything appears normal, the BMS seems to be operating normally and I do not smell any more odors, nor do I hear any more sizzling. Needless to say I am a little worried as I have never smelled this odor before and the sizzling noise was kind of scary.

Does anyone have any thoughts about what might have caused this odor and noise? I am worried that another of my batteries is cooking itself to death, but I don't know what the symptoms are of a battery going bad or how to check for a failure. Or could it be something else?

electrician
23 October 2010, 1841
Stop riding in the rain, lol.

Richard230
24 October 2010, 0710
It was cool and dry when I went out for my ride yesterday. I am pretty sure that the sizzle was not caused by boiling water. I just hope it wasn't boiling Lithium.

cycleguy
24 October 2010, 0740
It's quite possible that you were overheating your bad cell. I wouldn't even operate a bike with a known bad cell,it was either internally shorted or had very high internal resistance. Since it's in series with the rest of them, it still has to pass current through the cell, which would cause a damaged cell to overheat. You're lucky it didn't damage any neighboring cells.

Richard230
24 October 2010, 1426
The cell that I removed has apparently been dead since the day I bought the GP-S, 1500 miles ago, based upon the CA's voltage readings. It showed no voltage whatsoever when I finally got around to checking each cell, so I don't know how it managed to conduct power, but it did. I really hope that it was what was stinking and sizzling and not another cell that might be in a location that I can not reach. My pack is fully charged today. It is showing 83 volts right now after sitting around with the charger unplugged for the past 6 hours. Normally, I would be seeing 78 volts under those conditions and that has been typical since the day I bought the bike.

I plan to ride it around tomorrow afternoon, assuming that the weather drys out by then and hope that the stink and sizzle doesn't appear again. But, I am not going to push the bike like I did yesterday. I ran it wide open for most of my ride.

frodus
25 October 2010, 0813
Why on earth were you riding with a known bad cell?


Thats like driving a car with no oil in the engine. Its just not something you do until its fixed.

You probably started to reverse the cell and extreme overdischarge it. That would cause the cell to heat up and it probably vented through the tops of the cell.

seanece
25 October 2010, 0929
Sizzle.... smell.... swollen...... Sounds like that cell wanted to vent (or did)!!!! Good times. Glad you got it replaced ASAP.

Richard230
25 October 2010, 1555
I didn't realize that I had a shorted cell, not having any experience with a 24-cell pack. The voltage was always 76.8 volts when fully charged since the day I bought the bike (as I have mentioned a number of times before during the past year on El Moto), so I assumed that was the normal voltage for a fully charged pack (after all, the GPR-S is described as having a 72 volt system). I only discovered the bad cell when just for fun on a rainy day last month I took the fairing off the bike and started checking each cell with a volt meter. All the cells showed 3.35 volts, more or less, except for that one cell that showed 0.00 volts. I had no idea that a cell could be that dead and the bike would still run OK for 1500 miles. Now that I replaced the dead cell, I am getting over 80 volts, but after riding around today, the at rest voltage sure seemed to drop more quickly than before. Also, bulk charging is now intermittent. It would be really weird if another cell in the pack went bad (after 1500 miles) the very day, and just before, I replaced the dead cell.

I'll keep an eye on the performance, but I have a bad feeling about this. Any other cells boxes look really hard to remove from the frame.

BaldBruce
25 October 2010, 1921
Most LiFePO4 packs follow this profile.
24 X 3.6 = 86.4V for a absolute maximum charge. Probably set for slightly less. Almost no gain in capacity by going over 84V no matter what TS web site says.
24 X 3.35 = 80.4V when the surface charge has been removed.
24 X 3.2 = 76.8V when operating with a reasonable load. This is the voltage you should see the most when riding around.
24 X 3.0 = 72V when disharged heavily or getting low in capacity.
24 X 2.5V = 60V when cold and really heavy current or when the pack is falling off the cliff of capacity. Time to get off the throttle or get of the bike.
24 X 2.0V = 48V Push the bike home, your done.

The blue brand cell voltage looks like this graph from Jack: (but they are all similar.)
506

Richard230
26 October 2010, 0820
Thanks Bruce. That information is really useful. I wish I had known that 9 months ago. At the moment, with my replacement battery, I am getting 84 volts fully charged, 80.4 volts after the pack sits around for a day, so that is looking good. However, my pack still drops to 52 volts when going up a hill and yesterday at-rest voltage dropped to 72 volts after riding a few miles. I need to go out today and check the performance again.

seanece mentioned that one of my batteries might have been out-gassing. That would certainly explain the stink. If this happened would it occur through the sealed hole at the top of the battery, located between the terminals? I wonder if this hole is a vent. Could I tell which battery might have vented by inspecting this hole? See attached photo of the "dead" cell that I removed from my pack. The hole on this battery looks OK to me, although the sides of the battery are slightly bulging.

Also, would anyone want to hazard a guess what the numbers on the sticker that is on the side of the battery indicate?

frodus
26 October 2010, 0837
I only discovered the bad cell when just for fun on a rainy day last month I took the fairing off the bike and started checking each cell with a volt meter.

So you knew about it for a month and still rode it anyway?

No wonder it vented.



Here's a TS vent:
http://26.media.tumblr.com/tumblr_kyium4K3P01qa2swjo1_500.jpg

Richard230
26 October 2010, 0921
After I found out that I had a bad cell I rode my bike one time in order to reduce the voltage of the rest of the cells in the pack so that they would conform to the voltage of the replacement cell. I didn't think anything would happen, as I had been unknowingly riding with the dead cell since buying the bike new and after all that riding it seemed like one more ride wouldn't hurt anything. In any case, I don't think it was the dead cell that vented. I hope it was, but I fear that coincidentally, another cell went south on that last ride. What I am trying to determine is how I can locate a bad cell, should there be another one in my pack. When full charged, all the cells read about the same voltage, but it might be different under a load. I was hoping that by visually inspecting my batteries I could tell if one had vented. If that happened, would the vent cap have blown off?

seanece
26 October 2010, 0953
I would be flabbergasted if the one dead cell (that has been dead for eons) did not finally vent or crack.

However, if another cell vented or went bad, check the voltage on allllll cells. Check for odd fluid streaks and definitely look for swollen cells. Those poor Hi-Powers are notorious for swelling before and during their death. Yell at me if you need more cells, I actually have a few good ones laying I would sell for a great price.

Good luck

cycleguy
26 October 2010, 1047
Richard,

I'm assuming you don't have a BMS. I would highly recommend one. It will tell you when a cell drops below minimum voltage while riding, as well as balancing and HVC functions for charging. If you get a centralized system, with wiring going from the BMS to each individual cell, you can simply unplug the harness and check your individual cell voltages manually without taking the whole bike apart.

Mini BMS

http://www.cleanpowerauto.com/MiniBMS.html

frodus
26 October 2010, 1323
Walter,

He's got a GPRS, its got a BMS onboard. Just doesn't seem to be doing its job.

Richard230
26 October 2010, 1450
Cycleguy, I do have a BMS, as shown in the attached photo. As near as I can tell, the BMS is working perfectly. It uses a board on top of each set of 4 batteries and indicates when each cell has reached its maximum cut-off voltage by showing a green light. When balancing is completed, all four lights turn green and the charger stops charging for a while but if a couple of lights go out it will start charging again. The balancing charge rate is around 1.5 amps, instead of the bulk charging of 7.5 amps. However, it took me a long time to understand how these lights worked - being as how they are not visible unless you remove all of the fairing covers and I was too busy riding to do that. However, once I finally checked the voltage of each cell, I noticed that the green light over the dead cell never came on, although all of the other three lights in that pack did. So, it appears that my BMS is working OK, as after charging cuts off and balancing is complete, all of the cells are close to the same voltage, between 3.4 and 3.35 volts.

Having said that, I just returned from an easy 10-mile ride. When I started off I was seeing 80.2 volts (after the bike had sat around for 20 hours without being plugged in). Before replacing the bad cell I would have seen 76.8 volts. However, after a mile of riding, the voltage had dropped to 76.2 volts, where it stayed for 5 miles. It then dropped to 74. 3 volts until I arrived home. After stopping for a minute, the voltage rose to 75.2 volts. All of these readings were at rest with the throttle off, but with a 6 amp draw from the sepex motor, I assume. This leads me to think that there is at least another cell not pulling its weight. When I inspected the cells after the apparent venting episode, there seemed to be no obvious damage to any of the cells, but because they all fit tightly in their retaining boxes, swelling can't be seen unless they are pulled out. What I did not look for was to check if any of the vent holes had blown open (if that is what they do when they vent).

Unlike yesterday, when my charger was acting intermittently, the pack is charging normally and the charger is operating the way it always has. All I need to do now is to try to determine which cell is not carrying its load, if that is what is happening. My question would be, how do you do that?

Thanks seanece, I might take you up on your offer to sell one (or more) of your old batteries - as soon as I figure out what is going on and which cell is bad, if any.

I really appreciate all of your comments and suggestions.

BaldBruce
26 October 2010, 1539
Two generic things to look for witha pack that has lower than expected voltage under load. One is increased resistance from a bad connection, the other is a cell that has less capacity than all the others. You initially reported going all the way down to 52V under load and that sounds like a bad connection. Your ride today however looked more like the second problem where everything is fine until you start getting one of your cells near it's cliff. I'd charge it up, check all cells,then drive it till the voltage starts to drop quickly. We are looking fo rthe point where one of the cells is causing a quick drop in pack voltage. Stop and measure the cells to see if one or more is giving up before the others. (A BMS that actually tracks and records voltage would save you a lot of grief in troubleshooting this one...)

Good luck and keep us posted.

cycleguy
26 October 2010, 1607
Sorry for being ignorant of your setup Richard, I'll try to pay closer attention next time.
One way of identifying a potentially weak cell is to monitor your BMS cell boards while the battery is being charged, to determine which cell is the first to reach a full charge or first to start shunting (if equipped). This assumes that all your cells are balanced to begin with.
The basic idea is that the cell with the lowest capacity will reach a full charge first. It will also be the first to reach it's low voltage threshold as well.
You may want to operate the bike until the battery nears it's minimum voltage and then check the cell voltages before recharging, a faulty or low capacity cell will most likely show a lower voltage than the rest.

frodus
26 October 2010, 1628
Richard,

Your BMS doesn't cut off if a cell voltage goes below a certain point? I realize it cuts off charging if the voltage gets to HVC, but it should cut off the load at some point if a cell goes too low.

That IMHO is a MAJOR design flaw.

Richard230
26 October 2010, 1810
Frodus, my batteries have always had a limited power draw when the voltage drops to 51 volts under a load and then the controller (I assume) keeps it at that voltage while the current is reduced. It has operated this way since I bought the GPR-S. The only thing different is that the current draw under a load has dropped over time. Originally the CA would show 175 amps at the 52 V cutoff, and now it shows around 135 amps at the 52 V cutoff. The current bounces around and so does the performance when that happens. Since I have nothing to compare my bike's to (not knowing anyone nearby who has a similar machine), I do not know what performance and charging characteristics are to expected from the bike's equipment. So most of the time I am guessing.

Cycleguy, my BMS, as has been reported by others, seems to be set up to balance one group of four batteries at a time, starting with the ones furthest away and ending up with the four cells just under the "tank". At least that is how the green lights seem to light up. The box directly under the tank is the one where the dead cell was located and is the last one to be balanced. I don't know what the BMS does when the bike is running, but it does seem to operate properly when it is charging - at least that is my feeling.

I think I will give Bruce's idea a try. I'll ride the bike around then take it home, take off the fairings and check each battery's voltage in a partially discharged state. Perhaps that will uncover the tired battery. I'll also check for any vent that seems to have popped off and report back.

frodus
27 October 2010, 0909
It doesn't matter, a BMS that does not sense a FAILED cell (below 2V or adjustable) and cut off the pack, is a piece of crap in my opinion. Yes your controller senses PACK voltage and limits accordingly, but what you absolutely cannot live without is a monitoring system of some sort that senses the voltage on a cell level. The reason is, if ANY cell gets too low (like your dead cell), it should cut the pack off so you can take a look and service the pack.

Any BMS that does not protect from both overvoltage and undervoltage shouldn't even be installed. I'd rather have a system that doesn't shunt but senses cell voltages.

electrician
27 October 2010, 0934
@Richard - why don't you email EMS and see what they say about your problem. They may even exchange your battery for you. Also, asked them if your BMS is operating the way it is supposed to. Perhaps yours is faulty? Anyway, just a suggestion.

Richard230
29 October 2010, 1102
Well, here is the latest news about my battery problems. Yesterday I took my bike out for a ride and it was not performing well. Power draw was limited to only 5.5 kW, whereas when the bike was new I could get a maximum of 7.5 kW out of the pack before the voltage dropped to 52 volts and the controller killed the fun. (I might add that my old 60V GPR-S would show a maximum of 11.5 kW draw and 9.5 kW, continuous. It must have had a happy set of 50 Ah Hi Power batteries.)

So I took Bruce's advice and returned home, after using about 10 Ah, and removed all of the bike's fairings. I have not recharged the pack after the ride. This morning I checked the voltage on every battery and found that most cells showed 3.18 volts, with a low of 3.02 volts and a high of 3.21 volts. The resting voltage of the entire pack, after sitting over night was 76.1 volts. However, one cell was completely dead and showed no voltage whatsoever - just like the last cell that I replaced. Oddly it was the cell immediately adjacent to the cell that I just replaced. At the time, this cell was showing typical voltage and no swelling or any other indication that it had vented or otherwise was failing. I am really mystified, but I have ordered another replacement cell from Harlan. At least the cell is located in an accessible box, right on top of the chassis and is easy to replace.

I don't think EMS would be able to help me much. Not only do they only want to deal with their retailers, but the cells are guaranteed by Hi Power - which isn't much help as the US distributor says that there are none to be found in the US and they don't know when another shipment will arrive. So there is one weak link when owning an EV, if the battery manufacturer decides not to ship batteries to the US, you are out of luck if you have a failure and are stuck with using a used battery. Not something that is going to endear owners to electric vehicles. I know I am getting a bit tired of this whole thing. That is why my next electric motorcycle will be a Brammo. At least they have a track record of providing after sales service and repairing problems without the owner having to figure things out for themselves. But I sure would like to get my GPR-S running well so that I can give it to my daughter. I know for sure that she would not be able to deal with a malfunction and would just keep riding it until it came to a compete stop.

cycleguy
29 October 2010, 1713
Richard, Frodus is absolutely correct, a BMS at the very least, needs to monitor each cell and shut off the pack, or at least warn you when a cell drops below minimum voltage.
This may explain why you have another dead cell. Any time a LiFePO4 cell drops below roughly 2 volts at rest, (not under load) the cell sustains damage. Even though it recharges back to it's full voltage, it doesn't necessarily mean it it has full capacity.
Once the battery pack gets discharged while ridding, the damaged cell with lower capacity drops below minimum which causes further damage. If you don't have a good BMS to alert you and prevent overdischarge, you'll most likely have more failures in the future.

Richard230
29 October 2010, 1803
Cycleguy, unfortunately, I know little about what makes a good or bad BMS, but it appears that I am stuck with what came with the bike. The old spaghetti-type BMS that was on my old GPR-S worked completely differently than this current one does. It is a "Modolis Engineering, 4-cell, revision 2" BMS, photo attached. I don't know if this is a good BMS or not. I was told by EMS that it was a big improvement over their previous BMS, which had a tendency to burn up - as mine and a few others did. At least this one has not melted and it appears to do an OK job of balancing the batteries when they are being charged. The BMS in my 2008 GPR-S never did any obvious balancing. It just allowed the charger to run full tilt and then cut it off. The Modolis BMS charges at 7.5 amps, then cuts back to 1.5 amps and cycles on and off for several hours until it finally shuts off. But it has never cut off while I have been riding. Only once did I stress the batteries by running them down to 40 Ah (80% DOD). At that time the pack voltage dropped to 68 volts at rest and performance was reduced due to lower battery voltage and not any electrical cut-off. To my knowledge, the batteries have never dropped below 2 volts when under a load and the 24-cell pack has never dropped below 68 volts when at rest. Most of the time it has stayed above 72 volts at rest.

If the BMS is poorly designed or not functioning properly, I don't have a clue what to do about it. Replacing batteries stretches the limits of my electrical skills.

cycleguy
29 October 2010, 1942
Richard, with your BMS, you have no way of knowing if a cell dropped below 2 Volts. Just because your pack voltage reads 68V doesn't mean all your cells are at 2.83V. You can have one cell at 1.8V and all the others at 2.9. That is why you need a better BMS.

The BMS I recommended earlier is one of the most reliable, simple and most cost effective BMS's available. It provides a low voltage warning buzzer and LED light output as well as the option to chop the throttle. It also has shunting ability for charging and pack balance as well as high voltage cut-off to shut off your charger when the first cell reaches max voltage.

I would suggest using the centralized option which locates all the cell boards into one common board which can be located anywhere on the bike. This requires running individual wires from each positive battery terminal as well as one negative wire from the first cell. This will allow you to check individual cell voltages without having to remove you body work. But most of all, it will protect your batteries and your investment.

Richard230
30 October 2010, 0748
Thanks for the advice, Cycleguy. I am getting the feeling that you are right about the BMS. I am going to try replacing the latest dead cell one more time. If I get another failure in that particular battery box, then I will be pretty sure that the problem is a failed or malfunctioning BMS board located over that box of 4 cells. If that is the case, then I am going down to my dealer and ask what they can do to fix the problem, as it would seem that sort of malfunction would be covered under the vehicle's warranty.

Frankly, I would have no problem paying someone to fix my motorcycle, by installing a new BMS, or whatever is needed. But, while I have 48 years of riding and maintaining IC motorcycles, I never was qualified, knowledgeable, or skilled enough to work on their electrical systems and messing around with an EV is even more beyond my experience or knowledge, unless very detailed and specific instructions were provided, such as you might see in a Clymer, Haynes or Chilton service manual. As far as I am aware, there is no such manual for the GPR-S. I have the money, but not the skills or confidence needed to fix things like this.

electrician
30 October 2010, 0941
Doesn't EMS provide manuals (or a packet of information) with their kits? I thought I read that on their web site. Maybe you could ask them for a copy of their stuff, for a price of course.

Richard230
30 October 2010, 1322
Electrician, I asked EMS several times last year for a GPR-S manual and chassis torque information. I never received a response. Fortunately Chuck (as I recall) sent me an electronic copy of the 2009 GPR-S owner's manual, but the information provided in that document mostly relates to the 2008 GPR-S, with its 60V system and original BMS, which is not used in the current vehicles. There is no information how to perform repairs or normal servicing or what to do if thing go wrong. Even torque values for such critical items as the axle nuts and caliper fasteners is not available. Also, no written warranty information is provided in the document, nor was it provided by my dealer. If it wasn't for this forum, I wouldn't have a clue what was going on and no doubt I would not have ridden the GPR-S as far as I have.

electrician
30 October 2010, 1548
Sorry to here that. Now you have me taking a second look about EMS. Next year I was thinking about buying one of their kits. Maybe it is better just to buy the complete bike from Brammo?

Richard230
30 October 2010, 1807
Sorry to here that. Now you have me taking a second look about EMS. Next year I was thinking about buying one of their kits. Maybe it is better just to buy the complete bike from Brammo?

I don't think there is anything wrong with the basic chassis or motors - other than they are not purpose-made for EV use. The weak link, after owning two of their factory DOT approved production bikes, is the BMS that they have selected for installation in the GPR-S. The first one had a tendency to melt down and now the new one seems to have issues, also. I believe I am not the only GPR-S owner with BMS and/or battery problems. The only other complaint that I have with the performance of the GPR-S is the jerky throttle response. Both bikes felt like the controller was modulating the power very coarsely, even though they had different controllers. Also, the charging system takes a long time to fully charge the batteries, compared with the Brammo or Zero. I think you would be OK if you bought a kit, selected your own BMS and batteries and assembled the bike yourself. Just don't expect any customer service from EMS. They have very few employees and are involved in a lot of different things that don't seem to leave them much time for dealing with customer questions.

electrician
30 October 2010, 1943
Thanks for the input. I did visit EMS and had a chat with Martin for awhile. They all seemed very friendly and knowledgeable and said that they would work with me, so that sold me on the basis GPR-S. I do believe you are right about their BMS and battery set up. I might just buy their Developers chassis that doesn't come with: motor, controller, contactor, throttle, batteries and charger. and just figure out my own configuration. I have been getting a lot of real useful and helpful ideas reading what all of the experience builders have to say on this forum.

chef
31 October 2010, 0211
I wouldn't trust the Modalis BMS either. My 24-cell TS pack has some weak cells (weak being 50% capacity of the other cells). Could be that they were bad to begin with, can't say for sure one way or the other. The common link between Richard and I is the Modalis BMS.

electrician
31 October 2010, 0600
This BMS looks promising to me:
http://www.cleanpowerauto.com/MiniBMS.html

Richard230
31 October 2010, 0723
I am going to replace my dead battery again and switch my batteries around in their box. If one or both of the batteries on the left side of the box crap out again, then I can be pretty sure that the problem is the BMS and not the batteries. No doubt this story will continue - until I, or someone else, solves the problem.

electrician
01 November 2010, 0846
Good luck :)

Richard230
01 November 2010, 1507
I did visit a Chinese friend today and asked him what was written on the tag on my Hi Power battery - the one that swelled up and crapped-out. Photo attached.

The first writing is some sort of factory production code. The 3.30 below it is the battery voltage when tested at the factory, the 2.1 represents the measured internal resistance, I think he said the 52.99 is the weight of the battery and my friend was not too sure what the 41.70 was. The two items at the bottom are supposed to be the name of the inspector, who used a stamp instead of signing his name and the writing looks like the date the battery was tested for conformance with Hi Power's standards. It is all Greek to me.

frodus
02 November 2010, 0934
I don't think these cells aren't the best suited for the bike either. They're 1-2C max at best.

Not many of the higher cell-count BMS units protect from overcurrent. Most just look at cell voltage and cut off power. 2V under load is the lowest a cell should go. If its at 2V at rest, that would mean it went way lower than 2V under load. It doesn't sound like the BMS did its job in protecting from overdischarge (and potentially overcharge). I'd suggest replacing it.

Richard230
02 November 2010, 1508
I don't think these cells aren't the best suited for the bike either. They're 1-2C max at best.

Not many of the higher cell-count BMS units protect from overcurrent. Most just look at cell voltage and cut off power. 2V under load is the lowest a cell should go. If its at 2V at rest, that would mean it went way lower than 2V under load. It doesn't sound like the BMS did its job in protecting from overdischarge (and potentially overcharge). I'd suggest replacing it.

I think that is very good advice - if only I had the skills and knowledge needed to remove the existing BMS boards and replace them with a better system. Unfortunately, I don't.

Still I am willing to give the BMS one last chance. I have a replacement battery that I bought from Harlan on its way. I'll use it to replace the latest failed cell and if that one goes out, then I will know that the only real option for keeping my GPR-S on the road is to replace the bad BMS.

HighlanderMWC
02 November 2010, 1518
IIRC there's Justin that posts on here periodically. He's over in Oakland. When I first looked at either building or buying he offered to build a bike for me. I think he even had ties with the EMS so might be pretty familiar with your GPRS. You might want to contact him.

Richard230
02 November 2010, 1818
That is good advice, Highlander. I recall Justin saying that he helped develop the GPR-S. If nothing else works, I'll get in contact with him.

guity
16 January 2011, 1449
Hey Richard, how did all this end up? I just got done replacing my TS 60AH battery that pooped out on my GPRS a couple months ago. Harlan wants me to install a bms so I don't have so many battery problems!

Richard230
16 January 2011, 1628
Hi Guilty. My problem was the Modalis BMS. It was sucking the life out of the batteries under the master board. I removed the entire Modalis BMS on my bike and replaced it with 24 miniBMS modules and a head-end board. The work was finished yesterday. The new BMS seems to be working properly, but my bike has developed a random stalling problem and we are trying to sort that out. I plan to write up my story sometime this week, although I will not have the stalling problem fixed by then. Also, I blew up one of the miniBMS boards when I installed it and I was not steady enough to prevent a spark when connecting the last power wire. So I need to buy replacement. But the bike runs fine without it. Except when I ride it and it stalls. When my friend rode it it didn't stall. Anyway I am working on a story about the bike repairs. When I get a chance I need to ask advice about the stalling problem. In the meantime, attached are some photos that I took during the reconstruction.

One other thing I discovered was that the High Power batteries can come back to life, through recharging them. We were able to recover several batteries that had dropped to 0.2 volts and after being charged up, they tested out OK. When they swell, they can also be squeezed back into shape and they will keep on working. Very strange.

chef
16 January 2011, 1742
Nice work. I plan to do similarly with the pack in mine. I found someone locally who has battery testing equipment so will wait for the results before deciding what path to take (if too many cells are beyond hope, replacing the entire pack would make more sense).

Is the stalling random or only during high current draw? The MiniBMS's LVC may be getting triggered if a cell sags too much. Did you have them set the LVC & HVC to match the HiPower's specs?

guity
16 January 2011, 1857
Sounds like quite a journey! I tend to go the other way with bad cells -- rather than try to resurrect them I would prefer to root them out. I think I still have 3 or 4 questionable cells to go (according to my cell logger), at a cost of about $100 per cell and a day's work to replace each cell. I guess the BMS is a lot of work to install or remove, so it wouldn't make sense to try an experiment to see if you stall with no BMS?

Richard230
16 January 2011, 1916
Deleted duplicate post

Richard230
16 January 2011, 1921
My stalling occurred after riding about a mile. I stopped for a traffic light and the bike would not start up. I tried to reboot the system about 10 times and after a few minutes the CA showed a draw and I was off again. This happened again at the next stop light, as I waited for traffic to clear. Then after a few minutes, I was able to ride off again and had no more problems until I returned to the shop, when it died again. The miniBMS boards are set to Hi Power specifications by the manufacturer.

Now the weird thing is, when my friend took the bike out (and I followed on his 1978 Honda CT90 - :O), he was able to ride it for 5 miles with no problems. Much of that was at full throttle, with the pack voltage sagging from 80 volts to 50 volts (under load), according to the CA. (The voltage returned to 78 volts at rest at the end of the ride.) The bike ran great for him. So after returning to his shop, he handed off the bike to me and it immediately died and we were not able to get it going again, so I took off for home. He told me this morning that when he returned to his shop, the bike started up normally. Both of us are scratching our heads over this behavior.

When it stalls, it acts as if the kill switch has been activated. All of the 12-volt systems work fine, including the CA, but whereas the CA usually shows a slight power draw when the bike is ready to go, the CA shows nothing but 0.00 in the kW screen. When it decides it is ready to move, the CA will show 0.3 kW. I keep thinking that some wire that connects the kill switch or the brake levers that activate the regen feature, has an intermittent short. My friend thinks there may be a problem with the controller. All I can say is that when it is running, it runs better than it ever did. On the other hand, I never had this type of stalling before we pulled the batteries out and installed the new BMS system (which I am convinced is not the problem). I may have one battery that is suspect as it seems to have a high internal resistance and will reach the cut-off voltage quickly

If anyone has any ideas how to solve this problem, I would be grateful for the advice.

chef
16 January 2011, 2108
Probably worth investigating the kill switch circuit. Check all the inter-cell connections. Could be that one of the connections isn't good and the expansion/contraction of the metal is making it intermittent. Corrosion on the contacts can sometimes be difficult to spot. A single flakey connection will bring down the series string.

Man 80v to 50v under load is a huge sag, or a drop from 3.33v to 2.08v avg per cell! You might want to reprogram the controller to reduce the max current draw. That should reduce the sag and help prolong the life of the cells.

Richard230
17 January 2011, 0853
My friend would love to reprogram the Sevcon controller, but you need a rare $350 device to access the program and I think the batteries have had too much of this action to make much difference in their overall life. I would replace all of the batteries with new ones and reprogram the controller if I was going to keep the bike, but since I have a new Brammo Empulse 10.0 on order, what is the point? I'll be happy to keep it running until the summer and then sell it to someone who wants a project to work on. The real problem with the bike is probably the motor. I just wants a lot more power than the batteries can deliver. If I read the power curve of the D&D motor correctly, it wants to run at 4 to 500 amps and the batteries will only supply about 150 amps before gasping for breath.

Speaking of batteries, I really want to see how Guilty manages to pack 24 60 Ah TS batteries into his GPR-S. That would be my ideal choice, but I don't want to put that kind of money into the bike and I don't think there is enough room for both that many batteries that large and the D&D motor, too.

guity
17 January 2011, 0859
The problem is slightly different than the problem I had for a while - but my problem was fixed by replacing the controller. Fortunately my bike would not actually stall, but often when it was turned off it would not start up again. There was a little reset switch on the controller that, if I removed a side plate, I could reach in and hit, and then the bike would be good to go again. God what a pain, but not as bad or dangerous as just stalling out in the middle of traffic...

harlan
17 January 2011, 0933
If I remember correctly, the problem was not your controller but the way your precharge circuit was originally configured which was preventing the relay from consistently activating which then required you to manually activate it.

chef
17 January 2011, 1113
My friend would love to reprogram the Sevcon controller, but you need a rare $350 device to access the program and I think the batteries have had too much of this action to make much difference in their overall life.
...
Harlan rents the Sevcon programmer at a reasonable price. The controller determines how much is drawn from the cells, not the motor (well not directly). Continuing to tax the cells by drawing too much will degrade the cells further. I recommend setting it closer to 2c (100A) instead of 3c.

Watch for my post about the LiPo GPR-S. I imagine it's comparable to the performance of the low end Empulse

Richard230
17 January 2011, 1536
Thanks for that information, Chef. Right now the controller is set to adjust the power so that the voltage stays at 50 volts for the entire pack. Originally, it would limit the output to 175 amps for a few seconds and then drop it down to 150 amps (3C), and then drop to around 120 amps, where it would stay for a while under full throttle. As the pack deteriorated, maximum amperage dropped to around 130 amps when the voltage dropped to 50 volts, but that was with a dead battery, or two. Unfortunately, the motor needs at least 150 amps to get me up the local hills at a safe speed. We have a lot of fast moving traffic around here. The D&D motor is just too large for the batteries to feed. My first GPR-S, with its smaller Mars motor, performed better. BTW,.I note that EMS no longer has my sepex motor listed on their web site (it is ghosted on the site). I am not sure what that means.

guity
17 January 2011, 2024
If I remember correctly, the problem was not your controller but the way your precharge circuit was originally configured which was preventing the relay from consistently activating which then required you to manually activate it.

Ah, OK, I guess after my first accident you guys had to fix a lot of stuff, and didn't have time to elaborate on everything. This whole time I have been under the (mis)impression that the controller was replaced...

guity
19 January 2011, 1003
Ah, OK, I guess after my first accident you guys had to fix a lot of stuff, and didn't have time to elaborate on everything. This whole time I have been under the (mis)impression that the controller was replaced...

Just checked my posts from a year+ ago on the electric Motorcycle forum. I guess the stalling problem was fixed later than the repairs for the first accident. The stalling problem would have been fixed after I drove the bike too far and trashed some batteries. At least a couple of batteries were bloated up so big and stuck so tightly into the cases that Hollywood Electrics couldn't get them out without possibly damaging something. (I never mind damaging things, so I went ahead and split the metal cases at the corners.) I think the stalling problem was fixed during that time the bike was in the shop, and I was so focused on the battery problem that I didn't look into what the final resolution was for the stalling problem...Sorry.

Richard230
19 January 2011, 1605
My friend, who is trying to fix my GPR-S, told me that he discovered that Hi Power sells a tool that will squeeze bloated batteries back into shape. He says that they claim that bloating does not hurt the batteries. :confused: He showed me how you could take a big clamp and squeeze a bloated battery back into shape and it would then drop right into the cell pack. Frankly, I was amazed and also surprised that the bloated batteries that I had, which had not completely shorted out, still seemed to function fine - at least for the distance that have ridden the bike so far.

I was also surprised that I had no real problem removing bloaters from my pack. I sprayed silicon lube between the batteries, then I screwed in the bolts into the cell terminals. I then grabbed them with a pair of vice grips and wiggled the cell back and forth longitudinally a little at a time until it came out. Once one cell was removed, you could pull the rest out of the pack container easily.

guity
19 January 2011, 2305
Richard, there was actually more to my battery's story than bloating and stinking and leaking -- it was one of 4 batteries that my cell logger caught pooping out in the middle of my hill road home. These batteries can be really sneaky. Any time I measured voltage on them when the bike was standing still, they matched up with all the other batteries just fine. But the cell logger, recording the voltages every second during 20 miles of riding, showed that battery and 3 or 4 others taking a total nose dive way below the danger point when I made my last run up the hill, leaving the other batteries to do all the work. I probably should replace all those poop-out batteries, but at $100 a pop I'm biding my time. Do you have any way to log/record your individual cell voltages as you are riding? It could possibly shed some light on your stalling issue if you do...

Richard230
20 January 2011, 0913
Hi guilty. No I do not have a cell logger. Perhaps you can let me know what you are using and how to buy one. That sounds like something that would be useful - or maybe it might show me something about my cells that I don't really want to know about.

I am pretty sure that some of my batteries were performing the same way right before I finally gave up and took things apart and ditched the original BMS. The CA would show 80 volts (after replacing the cell that had apparently been shorted out since the first day that I bought my bike) and then drop down to 76 volts after riding only a mile or two. So I figured that one or two cells had crapped out by then and were likely doing the same thing as yours were. Unfortunately, I had to go a mile uphill to get back home and I was not about to turn the bike off and push it home - so the batteries just had to suffer. That was why I was surprised that some of the cells could be brought back to useful condition, although I got them mixed up and I couldn't tell which ones were the ones that were bad at that time and which ones I had removed because they seemed sick earlier. I just checked out all of the cells that I had and installed the ones that tested out the best when reassembling my packs.

That was why I was both happy and upset when I went for a ride after installing the miniBMS. While it ran, the CA was showing that the total battery pack was functioning normally. The pack would start out at 81 volts, then it would drop low under full throttle, but would rebound quickly to 78 volts once the throttle was closed, even after 5 miles of riding. That voltage level was an improvement compared to when the GPR-S was brand new. When the GPR-S was new, the CA would show a maximum voltage of 76 volts after a overnight charge. So I was really happy with its performance (I left the CT90 chase bike in the dust), until the bike stalled at a long traffic light stop and wouldn't get going again for a while, even though the CA was showing 80 volts at the time. That was depressing.

But my friend is doing a lot of research and plans to talk directly to a Sevcon expert tomorrow and he hopes to learn more about the controller. He also seems to believe that there is some way to access the controller and alter the programming using a PC. He might be right. It could be that you only need the rare programmer if you are in the field and can't bring the equipment near a lap top. He also tells me that the Sevcon controller is used for all sorts of electric motor-driven industrial equipment and that operating motor vehicles was not really in their marketing plan when designing the device.

DaveAK
20 January 2011, 0934
But my friend is doing a lot of research and plans to talk directly to a Sevcon expert tomorrow and he hopes to learn more about the controller. He also seems to believe that there is some way to access the controller and alter the programming using a PC. He might be right. It could be that you only need the rare programmer if you are in the field and can't bring the equipment near a lap top. He also tells me that the Sevcon controller is used for all sorts of electric motor-driven industrial equipment and that operating motor vehicles was not really in their marketing plan when designing the device.
Hey Richard, I've got a favor to ask. Could your friend ask the Sevcon expert about the CAN protocol used in the Powerpak controller, and if there's a specification that he could send me? I'm hoping to use the information to build a real time display.

As for hooking the Sevcon up to the PC, from what I understand is that this is available for most Sevcon controllers, except for the Powerpak. However, the interface and software required would likely cost as much as the hand held calibrator if it were available.

I should have my calibrator today or tomorrow, but right now my batteries are out of the bike again. Hopefully I can get it hooked up this weekend and I'll tell you what I find out about my settings.

Richard230
20 January 2011, 1544
I'll pass your request along to my friend, Dave.

chef
20 January 2011, 2056
Richard - I recall you dismaying about how your bike would "surge" going uphill. What's happening is the current draw gets too high, the cells sag below 50v, the controller's pack LVC triggers, reduces power, the cell voltages rebound, the pack voltage rises and the controller briefly allows full power again before the pack hits LVC, etc. It's a vicious cycle that should be avoided to prevent cell damage. It's currently set to something like 180A (?) which is 3.6C, more than your HiPower cells are designed to provide continuously (such as when climbing long hills). What you want to do is set the max current to a level where the voltage sags but stays just above the LVC. That will prevent the surging effect and you'll end up with a better riding experience. In fact I'd expect you would have more climbing power at a lower max current setting by avoiding the sag/surge cycles. I think that counter-intuitiveness may be what you're getting hung up on in regards to reprogramming the controller.

You seem to have a misperception about how the controller is doing its job. You mentioned that the controller keeps the pack voltage at 50v during acceleration. The controller isn't trying to keep the pack at that voltage, the cells have limited discharge capacity and sag to that level as more current is drawn. The controller cuts back power at LVC (50v in your case) to protect the pack (but not protecting individual cells). On the flats you don't notice much of the surging effect, but hills demand more current draw and your controller is allowing the cells to discharge at a higher rate than they're capable of (though only for brief moments until they sag too far).

Richard230
21 January 2011, 0912
Chef, that is not quite what was happening with my GPR-S. Both of my bikes, including the 2008 one with the 60V system and its other brand of controller, had a "surging" problem. But it was only slight when riding at full throttle. The surging was very noticeable at low throttle openings, such as when trying to ride around in a parking lot. Both bikes would seem to turn off and on at about one second intervals when I would try to ride slow.

On both bikes, the maximum battery pack power output seemed to vary while the voltage was held at a specific level, such as in the case of my current bike. I believe this is the controller's work as the low pack voltage would always drop to exactly 51 volts, plus or minus .5 volt under full throttle. I don't see that type of continual precision would be the result of the batteries' chemical reactions. While the voltage would remain steady, the amperage would vary up and down. When new it would go as high as 175 amps for a few seconds and the drop to 150 (3C) amps at 50 volts. When the upper pack's batteries started going south, the voltage would still remain at 50 volts, but the power output would show around 130 amps max. There would still be a small amount of surging, but it was still much worse at low throttle openings.

So, based upon my observations while riding my two examples of the GPR-S with two different controllers, I would disagree with your analysis. I think the surging is the result of the limitations of the equipment on the vehicles.

harlan
21 January 2011, 1422
Richard,

I believe Chef is correct. The surging on your 2008 GPR-S was a different problem altogether and a result of the crude control the Alltrax controller would allow. The Sevcon you have in your new GPR-S has more advanced capabilities that make it a smoother ride unless you keep on hitting LVC as Chef describes.

Richard230
21 January 2011, 1552
Richard,

I believe Chef is correct. The surging on your 2008 GPR-S was a different problem altogether and a result of the crude control the Alltrax controller would allow. The Sevcon you have in your new GPR-S has more advanced capabilities that make it a smoother ride unless you keep on hitting LVC as Chef describes.

Unfortunately, the Sevcon controller works no better than the Alltrax did, as far as low-speed surging goes. Holding the throttle grip as steady as possible (a skill I learned from years of riding BMWs) at 30 mph, the power draw would vary all by itself from 20 to 40 amps. It actually is more steady under full throttle, when the variation is much less, although the power will fall after a few seconds to a point where it will stabilize anywhere from 150 to 100 amps, depending upon how the batteries are feeling at the time.

chef
21 January 2011, 2125
Your bike's low speed surging seems unusual. If it's not the controller as Harlan suggested, maybe it's a bad throttle (dirty contacts?). Try the volume knob on old cheap computer speakers. When the pot contacts corrode, it sounds very static-ky when you change the volume, especially at the lower volumes. Another possibility might be a loose chain.


On both bikes, the maximum battery pack power output seemed to vary while the voltage was held at a specific level, such as in the case of my current bike. I believe this is the controller's work as the low pack voltage would always drop to exactly 51 volts, plus or minus .5 volt under full throttle. I don't see that type of continual precision would be the result of the batteries' chemical reactions. While the voltage would remain steady, the amperage would vary up and down. When new it would go as high as 175 amps for a few seconds and the drop to 150 (3C) amps at 50 volts. When the upper pack's batteries started going south, the voltage would still remain at 50 volts, but the power output would show around 130 amps max. There would still be a small amount of surging, but it was still much worse at low throttle openings.
That's now how a controller works. The controller will pull as much current as it's designed or programmed to (varying dependent on the throttle of course). The batteries will try to deliver that current, but if the draw is too much, the voltage sags significantly. The controller doesn't try to "hold" the voltage at any particular level, though the LVC is mildly related to what you believe to be observing. It could be that the HiPower sag is right around 2v/cell for the higher current draws your bike is set to pull.

You can prove that to yourself very easily. Run the throttle at 1C or less (<50A). The voltage should not sag much. The lower the current, the less the sag.

Richard230
22 January 2011, 0843
Chef, I have tried to hold my throttle steady many times, while riding on the 30 mph speed limit street, and looking at the CA at the same time. The voltage will just not hold steady. A 50 amp draw will get me going about 40 mph, but it varies about 50% up and down, all by itself. I always assumed that the problem was with the throttle control mechanism, since both of my GPR-S bikes would do the same thing. When I bought my latest model, my dealer told me that EMS was recalling the throttle assembly and a replacement was forthcoming - naturally it never arrived, although I kept bugging my dealer about it - until he went out of business. A new or properly serviced throttle might fix that problem, but I doubt it is going to help get the bike running when the power shuts off and that is where I am focusing my attention right now. Too many problems and too little time.

Richard230
22 January 2011, 1923
Hey Richard, I've got a favor to ask. Could your friend ask the Sevcon expert about the CAN protocol used in the Powerpak controller, and if there's a specification that he could send me? I'm hoping to use the information to build a real time display.

As for hooking the Sevcon up to the PC, from what I understand is that this is available for most Sevcon controllers, except for the Powerpak. However, the interface and software required would likely cost as much as the hand held calibrator if it were available.

I should have my calibrator today or tomorrow, but right now my batteries are out of the bike again. Hopefully I can get it hooked up this weekend and I'll tell you what I find out about my settings.

Dave, I found this PDF for the user manual for the Sevcon PCpaK on the EMS web site: http://www.electricmotorsport.com/store/pdf-downloads/sevcon_pcpak_manual.pdf

I notice that it mentions CAN stuff in the manual.

Every time I hear the phrase CANbus, I am reminded that my two BMW motorcycles use that system and that you have to bypass the internal wiring system when you add any after-market accessories to get them to work on the bikes.

chef
23 January 2011, 0828
Richard - Have you ridden in the rain/wet on either GPR-S? Could be that moisture corroded the throttle contacts. No telling whether EMS re-used the throttle from your first bike.

Down the road you might consider switching it out with a hall effect throttle. No direct electrical contacts so no corrosion concerns (well, as long as the throttle is a sealed design). Harlan hacked one together if you want to see how it works
http://www.elmoto.net/showthread.php?490-Magura-throttle-hall-effect-mod

Richard230
23 January 2011, 1606
Richard - Have you ridden in the rain/wet on either GPR-S? Could be that moisture corroded the throttle contacts. No telling whether EMS re-used the throttle from your first bike.

Down the road you might consider switching it out with a hall effect throttle. No direct electrical contacts so no corrosion concerns (well, as long as the throttle is a sealed design). Harlan hacked one together if you want to see how it works
http://www.elmoto.net/showthread.php?490-Magura-throttle-hall-effect-mod

I have avoided riding in the rain like the plague. I think I rode through a couple of small puddles and down a damp street a couple of times, but I do live near the coast and that can certainly corrode electrical contacts and such.

I spoke to my friend who is stuck with my bike until he fixes it. He tells me that it runs just fine as long as you don't close the throttle for more than a few seconds - at which point it dies. He is going to try disconnecting the regen circuit and see if that makes any difference and he has a couple of other ideas. His problem is a lack of time. He has two jobs, an airplane, a sailboat, a couple of cars and 19 motorcycles. Frankly, I feel sorry for him. I could never deal with all of that stuff, but he seems to like to work on Sparky. It is now a puzzle and he apparently enjoys the challenge. I am really glad that he has been willing to take on this task.

He thinks my "surging" problem is due to over aggressive throttle control programmed into the controller by EMS. He also told me that the "kill switch" seems to be hooked up to the controller, which is designed as a "dead man" switch for use on devices like fork lifts, where you would have a switch under the operator's seat and he would have to sit on the seat before he could operate the device. This would be a safety feature to keep the operator from trying to run the fork lift from an unsafe position (walking alongside, that sort of thing).

That hall effect throttle idea seems like a good one. I'll put it on my list of things to do once Sparky gets sorted out.

NonPolluter
11 February 2011, 1115
Several people have taken LIFEPO4 batteries apart, or drilled into them. However, the Thundersky MSDS says that there are health warnings about the fumes and stuff coming out.

Here, a GPR-S owner was forced to drill into his swollen batteries:

guity's gpr-s experience (http://electricmotorcycleforum.com/boards/index.php?topic=512.0) 1 (http://electricmotorcycleforum.com/boards/index.php?topic=512.0) 2 (http://electricmotorcycleforum.com/boards/index.php?topic=512.15) ... 13 (http://electricmotorcycleforum.com/boards/index.php?topic=512.180) 14 (http://electricmotorcycleforum.com/boards/index.php?topic=512.195)

Richard230
11 February 2011, 1607
When I spoke with the Brammo staff, I was told that the fumes that come off of lithium batteries are very hazardous. Now I wish I hadn't sniffed around my bike when my batteries were sizzling, puffing and venting. I don't think I'll be making any lithium-smell cologne. And here I was just about ready to submit my patent to the government patent office for that odor. :O

I think I will pass on drilling holes in my batteries too.

guity
12 February 2011, 0912
Heh -- well that solves the mystery of why tying my shoes has become so mentally challenging lately... Not sure there is much I could have done about whatever damage my brain sustained. By the time I was drilling holes in the batteries, they weren't leaking or smelling any more and probably weren't all that dangerous. Most of the leaks and smells came during/after rides, though I do remember at times kind of sniffing around the assembled bike to try to get a more precise idea of where the bad batteries were. I don't think it's a good idea to put those batteries in fixed metal cases at any rate, given that they are going to expand like that. Bolted metal bands seems like a much better idea.

Having watched videos of Jack Rickard dismantling TS batteries to show how they are constructed, I wasn't too terrified of doing it. Drilling the holes, however, was only helpful in getting leverage for pulling them out -- no matter how much leverage I had, I couldn't pull them out until I cracked open the metal cases.

Now I realize that the main problem with my TS batteries is that they are ridged at the sides, probably for heat dissipation purposes. The ridges on the broad sides of the batteries are vertical, so I don't think they cause any problems with removal. But the ridges on the narrow sides of the batteries are horizontal, and when they interlock with the neighboring batteries' ridges I think it becomes just about impossible to take one battery out by itself, without doing something to increase the amount of space between the interlocked ridges...

chef
12 February 2011, 1709
Someone on the TS group posted the chemical makeup of TS electrolyte:


David Dymaxion wrote:

How hot does a surface have to be before it'll ignite the electolyte steam from a venting cell? How hot does it have to be for a spark or flame to ignite it? The thought I had was could the shunt bypass resistor get hot enough to ignite the vented steam?

Thundersky electrolyte has:
Ethylene Carbonate (EC) + DiEthyl Carbonate (DMC) + DiEthyl Carbonate (DEC) + Ethyl Acetate (EA)
http://www.everspring.net/product-battery-msds.htm

I'll give the conclusions here first, and then the supporting data. It looks like a lithium cell will vent at as low as 77 Celsuis (EA) and DMC will vent at 90 Celsius. The flash point is room temperature for DMC, DEC, and EA, and auto-ignition occurs at 426 Celsius to 465 Celsius. If a cell vents, it appears any spark (like from a wire that melted off its insulation, or a relay) could ignite a venting cell. The shunt resistor (or any surrounding surface) would need to get to around 426+ Celsius for ignition. EA looks like the most dangerous component, as it has the lowest boiling point, lowest ignition temperatures, makes an explosive mixture with air, can creep considerable distances to an ignition source, and flash back to the source.

So there's a theory for some of the recent car fires: Perhaps some cells vented EA, and it crept out of the car to an external ignition source, and flashed back to inside the car. That might be stretching, but worth thinking about.

My question for the experts is, what is the hottest a shunt resistor can get? Is it possible for one to get to 426+ Celsius in a runaway situation (like where the charger fails to shut off)?

It sounds like you do not want open relays near lithium batteries, if your BMS uses relays you'd want to be sure they are the kind that are sealed and don't have open sparks.

Maybe it is more important to monitor temperature than Voltage? Perhaps an independent system that monitors temperature and shuts down if anything gets hot would be a great idea. I like someone else's idea (sorry, forgot who) that said to put a smoke detector in each battery box, and shut down in the presence of smoke.

Now for the data:

Boiling points:
EA: 77 ° C
DMC: 90 ° C
DEC: 127 ° C
EC: 243 ° C

Ethyl Acetate (EA):
Auto-Ignition Temperature: 426.67°C (800°F)
Flash Points: CLOSED CUP: -4.4°C (24.1°F). (TAG) OPEN CUP: 7.2°C (45°F) (Cleveland).
Special Remarks on Fire Hazards:
Vapor may travel considerable distance to source of ignition and flash back. When heated to decomposition it emits acrid smoke and irritating fumes.
Special Remarks on Explosion Hazards:
The liquid produces a vapor that forms explosive mixtures with air at normal temperatures. Explosive reaction with lithium tetrahydroaluminate.
http://www.sciencelab.com/msds.php?msdsId=9927165

DiEthyl Carbonate (DMC):
Auto-ignition temp: 445 Celsius
Flash point: 25 Celsius
Special Remarks on Fire Hazards: May form explosive mixtures with air.
Special Remarks on Explosion Hazards: Vapors may form explosive mixtures with air.
http://www.sciencelab.com/msds.php?msdsId=9923746

DiEthyl Carbonate (DEC):
Auto-Ignition Temperature: 445°C (833°F)
Flash Points: CLOSED CUP: 25°C (77°F).
Special Remarks on Fire Hazards: May form explosive mixtures with air.
Special Remarks on Explosion Hazards: Vapors may form explosive mixtures with air.
http://www.sciencelab.com/msds.php?msdsId=9923746

Ethylene Carbonate (EC):
Auto-ignition temp: 465 Celsius
Flash point: 143 Celsius
http://www.sciencelab.com/msds.php?msdsId=9923969

guity
12 February 2011, 1744
So Chef, thanks. I'm getting the impression form your post that the danger from these batteries is not from toxins but from explosion -- is that right?

Richard230
12 February 2011, 1836
None of those chemical compounds seem like anything that I would want to breath. They kind of remind me of chlorine - which was used as a poison gas during WWI. :O

chef
12 February 2011, 1920
So Chef, thanks. I'm getting the impression form your post that the danger from these batteries is not from toxins but from explosion -- is that right?
Personally I don't think the risk of ignition is significant, but I don't have enough information to say one way or the other. The recent topic of vapors was discussed because of the LincVolt fire. Jack's camp blames the BMS (not surprisingly), others say it's the charger, and others speculate that vaporized electrolyte caught fire. What's germane to the discussion here is the relatively low boiling points of the electrolyte -- 77 & 90 deg C for two of the compounds. The batteries will swell from the loss of electrolyte and suffer permanent capacity loss. It only takes an overcharge of a few tenths of a volt for the electrolyte to boil. Drawing too much current could also lead to excessive heat and boiling. EMS has a nasty habit of doing both.


None of those chemical compounds seem like anything that I would want to breath. They kind of remind me of chlorine - which was used as a poison gas during WWI. :O
Don't be afraid of a little ethyl acetate... puts hair on your chest ;)

BaldBruce
12 February 2011, 2013
The organic vapours boiled off an overcharged battery are flammable as all get out!!!! This is not a suposition, but a fact I have unfortunately witnessed. LiFePO4 batteries just have the significant advantage that those vapours are not self-igniting as in a cobalt or maganese cathode types. The organic solvents used in all the Li family batteries are very different based on which flavour and manufactirer you are discussing but all come from the same category of materials. I would agree that these organics as a class are not a prticularly nasty from a MSDS perspective, but I still would treat them with respect since we really don't know the long term exposure risks. (Organic in this context means that they are carbon based compounds, not good for you to eat Ted.)

chef
12 February 2011, 2101
Bruce - how did the vapors ignite? A spark mayhaps? I've been curious as to how concentrated those vapors are and what levels are typically emitted. A tiny bit of flammable material doesn't have much punch, I imagine like a lighter that gets blown out by the wind. Or is the danger from the vapors building up in a semi-enclosed space?

BaldBruce
12 February 2011, 2116
One particular example was a closed space with open relay INSIDE of space.(Theory that either the relay or an overheated wire caused the ignition. My money is on the relay) Computer controlled battery cycler lost it's brains and overcharged the individual cells on test. Nice puddle of batteries after the fire. These were smaller prismatics, but can happen to any size or shape.
Unfortunately, I have seen the 18650 types (Cobalt types) gp off like Roman candles when overcharged way too many times. We put a 1 inch thick piece of plexiglass on the test station because the 1/2 inch one actually failed when several rockets went off at once. (Again a case when a power interruption screwed up the cycler and cause a massive overcharge event.....)

chef
12 February 2011, 2157
"Pictures or it didn't happen" :D j/k

That sounds scary fun. Thanks for burning up cells so that we don't have to!

BaldBruce
13 February 2011, 1940
Sorry, no pictures, so it never really happened...:). I work in a totally different area and company now. Doing the LED power supply thing now. The previous job was with a battery manufacturer and charger company. Cameras were strickly verboten in the lab areas and I can sort of understand why. No company would want to be tied to some of the fires that happen in the development phase......:O

teddillard
14 February 2011, 0434
(Organic in this context means that they are carbon based compounds, not good for you to eat Ted.)

awww, hell, back in the '70s we started the day with a little DiEthyl Carbonate on our toast. never could see what the big deal was... that Ethyl Acetate, though, man... THAT ****'ll kill ya.

>>twitch<<

>>twitch<<

(nice move sidestepping mention of blackouts and such resulting from said experiments btw :D)

fixitsan
27 February 2011, 1623
Hi,
I'm building a cell voltage monitor, specifically for large lithium cells. I have started a yahoogroup....
http://autos.groups.yahoo.com/group/CellMonitor/?yguid=314528416

I just wanted to point out that ethyl acetate is far from deadly. it is sometimes used in food as a flavour enhancer and in fragrances, as well as used in beauty salons as nail varnish remover.

What is important to note in terms of the explosion risk are the parameters known as 'Explosion levels'

The upper explosive level (UEL) for ethyl acetate is 9% by weight mixed with air and the lower level, LEL, is about 2% by weight.

Compare that to hydrogen which has an LEL of about 5% and a UEL of 75% and you can see that any precautions which you apply to the charging of lead acid cells, in terms of preventing a hydrogen buildup, also apply to lithium cells. You need take no more precautions other than bear in mind that where hydrogen rises, EA sinks, so vents in the bottom of any battery box and a fan will keep you safe.

Chris

chef
27 February 2011, 1646
Hello fixitsan / Chris -- Glad you dropped by from the TS group. That place has been nutso lately though it seems to have settled down somewhat. You might ask Mike (the forum owner) and see if he'd be willing create a sub-forum for you. Alot easier to manage threads on a vBulletin type of system than Yahoo Groups.

fixitsan
27 February 2011, 1649
Thanks for the welcome.
They say birds of a feather stick together. I'm also a trained chef, motorcyclist, EV fan....small world isn't it !

NonPolluter
24 March 2011, 0853
A different kind of odor: Our LIFEPO4 packs gave off a fruity smell on a bench. A process of elimination determined that the BMS was the source of the smell. That BMS was removed and replaced. Batteries were charged again, and no more fruit odors.

Weeks later, we got an acrid smell from Thundersky/Winston LIFEPO4 cells under load conditions of less than 3 minutes, with voltage sag down to below 2.5v/cell.