Battery Pack Issue

[cdb0ewm]

Not sure what I did but ....

I have been regularly charging my BM3 type battery pack - 13 groups of 6 Headway 12ah cells. Everythiung was fine until I recently tried to charge the pack again. The charger said there were only 7 battery groups, not 13. Check all balancing wiring via continuity tests and everything was fine. Check voltage of individual cells and found some that had less than 1 volt. I tried to individually charge these cells and some charged up to the 3.6 v level. Some charged up to 1.8 volts and that was it.

So, 2 questions

1. Any idea how this big variance in voltage within a group might have happened and how to I p[revent it from happening again?
2. Is there anyway to 'save' the cells that didn't charge up?

Appreciate the help

[expresso]

sorry to hear that - BM may be the best to answer you helping you with this - i am just guessing here - could be bad cells right from the start ? could be when you build the pack - did you let them sit for a few months and then recharge each cell to group them in near perfect balance for recharging ?

do you monitor the cells during use or someway so you dont over discharge them ? from what i read here - if they go below 2.5V - they may be worthless after and wont hold a charge - will bring the others down also over time and kill more cells

hope you get more better answers from others who can guide you thru this -

[Burgerman]

I have lots of lithium cells here. Some are many years old. Lipo, life, etc. Theres a spare headway or two on my bench from when they first came out. Still over 3V. Never been charged or used. Non of them have a high enough self discharge to go dead on their own.

So something must be draining them. Either left a long time connected to something, like a cell log, or a turned off powerchair, or something. You didn't elaborate.

But if you discharge a cell below 2.7V its fully discharged. 2.5V is still safe, but too low... 2V is the point where its permanently damaged. They don't do it alone if disconnected even over years. A BMS can do this, a powerchair controller can do this even if OFF. A charger can if left connected and not charging indefinitely, etc. So what was connected? Or what were you using it for?

1. Any idea how this big variance in voltage within a group might have happened and how to I prevent it from happening again?
2. Is there anyway to 'save' the cells that didn't charge up?

By group do you mean connected in parallel? As one big cell? Because if so, it simply isn't possible to have any voltage variation between cells, they are all bolted together. So it might be interesting to see a picture... You may have them wired wrong or something basic. Or one full GROUP of parallel cells were below 2V? In which case it may be one bad cell pulled the rest down. Or a cell log has. Please elaborate. What chair? Are you using the roboteq? Etc.

Did you disassemble the pack? Or how are you charging individually? Any cell below 2 volts, is toast. And any that read less than say 11ah if capacity tested to 2.5V at 1/2C are also ruined.

[cdb0ewm]

Thanks - I'll send pictures as soon as I can

And yes the balancing connector was connected to the pack the whole time

[Burgerman]

Well there's PART of your problem. Parasitic drain while its sat not charging through mosfets. You sent it dead. Also good way to kill chargers. As you should always connect power supply first, and disconnect it last.

[cdb0ewm]

Got someone to take pictures for me

If you look at the top row where there are 6 negative cells. Going from right to left, the voltages are .8v, 1.4v, 2v, 3.4v, 3.4v, 3.4v

The charger is a Roboteq and I didnot disassemble the pack before trying to charge a single cell

[Burgerman]

I cant follow that. Theres balance wires all over the place. Looks like some wont get balanced?

Worse. You have protective large sized heat shrink crushed under bolt heads, and between bus bars on many many places. Those will almost all be bad connections. I would be surprised if a 1/4 of those cells or more are not actually disconnected from the pck. You will have endless issues and balance problems. Theres no way that pack was ever charged or balanced correctly. You need to get rid of all of that trapped heat shrink before you begin. And clean all contacts, and go over every connection about 4 or 5 times o0n reassembly as its easy to find ones you missed. Or you will have much worse problems. And balance wires seem very long, too thin, and fragile. Every cell needs to be removed, tested for capacity, and reassembled with much more care.

If you look at the top row where there are 6 negative cells. Going from right to left, the voltages are .8v, 1.4v, 2v, 3.4v, 3.4v, 3.4v

IF they were electrically connected they will all read exactly the same. Anything else isn't possible. You have all that plastic heat shrink in the way on the bus bars, so most of the cells are disconnected! They only LOOK connected... These .8v, 1.4v, 2v are therefore not... You cannot build a pack this way. And you cannot build a chair this way. Much better understanding and accuracy of build is essential.

This sort of thing causes huge problems with one loose bolt! Its why you cannot 'get one built' and need to take charge and DIY.

That must be disassembled, and any cell less than 3.3v tested properly. Cleaned up properly, reassembled with care. And non of that heat shrink unless you are extremely careful to get non of it between bus bars, on top of cells, under screw heads etc. And loctited as you go.

[flagman1776]

I immediately began wondering about faulty connections... which I believe is Burgerman's first concern. After you tested each parrallel cell group, I'd start at each end to check series. If you find a break in the series from opposite ends... that'll be your bad connection.

[Burgerman]

.8v, 1.4v, 2v, 3.4v, 3.4v, 3.4v

THOSE for a start, are supposedly connected in parallel! Or rather they are not... Because that isn't actually possible. The first three at least are DISCONNECTED from the rest electrically. Looking at the pack, I expect the rest is about as good. You cannot work like that. These things must be assembled right and with care! I suspect there are many such bad connections and cells disconnected, or disconnected from balance etc simply by looking at the pictures.

The whole pack must now be disassembled, each cell tested individually fully charged, tested for capacity while completely disconnected from all else, reassembled properly and any failed or substantially different Ah reading cells replaced. And all of that heat shrink removed totally before reassembly.

I suspect you are going to need to find someone that understands electrics and electronics well, and that has a much higher standard of build capability/quality to help you finish this chair. Or it will never be finished or reliable, and will just keep costing you money. And may end up either a fire hazard or a danger in use to you at best. (I am thinking about the roboteq wiring issues you were having here too). You have no choice but to do these things correctly. This is a complex job, everything has to be 101% right to stand any chance of success.

For eg, You *must* see cell balance as accurate and close as this both during and at full charge to know all those connections and cells are correct. http://www.wheelchairdriver.com/BM3-con ... arging.gif

[steves1977uk]

That pack looks like it was built by someone who doesn't understand how to build a lithium pack, half of those cells are probably knackered

Steve

[steves1977uk]

Also now counted 15 balance wire connections? For a 13S pack there should be 14 balance wires.

Steve

[expresso]

i may be the least knowlegable about this than the rest here - BUT - it will be a waste of time to try to fix that with out taking it apart completely and do it over correctly - like BM says - i know you hate to hear that - and that was my fear when i did mines - the idea of doing it over was enough for me to make sure i did it correct the first time -

its a new year - take it easy - slow and do it over - that would be the fastest way - but also its really the only way - too many issues - take your time - you have the whole winter - if you start now - you be done by July

[Burgerman]

Or done by tomorrow.

[flagman1776]

I had to save & enlarge the photos to examine closely.
I'm seeing (8) bottom stacks of 6P as well as the 2 top horrizontal rows of 6P. There appears to be a U shaped 1 P group of 6 between the top rows & the bottom stacks... (picture 1 to Left) but to the right there appear to be (2) P groups of 4, cable connected to 2 more P cells on the end of the top rows. It's very confusing. I don't know how the cabling connections will affect those P groupings.
The red shrink must be pos, the black neg, the blue + to - jumpers.

Again, I'm agreeing with Burgerman that the nice looking heat shrink may be interfereing with the proper tensioning of the bolts, thus poor connections. I would cut away enough heat shrink that there was NONE between the metal to metal parts.
Other thing to consider is if the stepping of connector bars is interferiing with best contact. EVERY cell in a Parrel group should read the same. If you had a bad cell, it would drain them all, equally. If they are not, then there's connection problems.

I would try to remove the excess shrink tubing with a razor blade. It is insulating... None between metal parts! A lot of connections to check but who ever does it just has to start & be systematic.
The pack does not need to be physically disassembled necessarily... there's time for that if defective cells are identified... With the connectors removed, individual cells can be charged.

[Burgerman]

Heres what you need to do. In this order.

1. Remove every bolt, every bus bar, and all the wires.
2. Voltage test every cell individually. Discard any that are 2V or less as a matter of course.
3. Charge up all of the rest fully, holding each of them at 3.600V for at least 1 hour.
4. Discharge EACH at 10A down to 2.500V and write the Ah capacity removed on each cell as determined by the charger.
5. Recharge all (as 3 above.)
6. Discard any cell that is more than 3 to 5% less Ah capacity than the best cell.
7. Add new cells as required. And fully charge each one individually. (as 3 above.)
8. Dont refit any bars, wires or anything for a month.
9. Recharge every cell individually, after a month to test for a faulty cell with high self discharge. This will be very quick, a few mA only. If any are much more discharged than the rest, discard and replace. Write down the mAh figure on each.
10. REFIT all the bus bars without that heat shrink in the way... Clean the end of every cell, and the bars with fine wet/dry or that green scotch-bright cloth. Use loctite on every bolt. Be sure each is tight! And check all several times over as its easy to miss one and it causes havoc.
11. CHECK loom, use shorter cables, no crimp alone, solder properly. And then re-fit.
12. Charge fully as a pack, and post graph so we can see its correct.

Nothing else will work.

[cdb0ewm]

thank you
I'll get going

[cdb0ewm]

Got started on recharging individual cells. I found 3 groups of cells.
Those with a voltage of less the 2volts - Any attempt to charge them is futile - into the trash
Those with a voltage of 3.25volts and above - These charge fine. Go up to 3.6volts and 100% capacity
Those with a voltage around 2.6volts - Some of these take a looong time, like 4 hours to fully charge, while others charge slowly to 9% to 20% then the Hyperion finishes. If I leave it connected the capacity rapidly to 100%. If I recharge it the capacity starts of again at 9% - Are these ready for the trash too?

BTW - fixed the Hyperion charger - it was a loose wire inside

Thanks

[Burgerman]

Any that when fully charged, sat at 3.600V for a few hours, and then capacity measured on discharge down to 2.5v at say 10A that measure substantially less than the rest are junk.

[cdb0ewm]

If I am chargiung a single cell, whart Ah shoiulkd I use

[Burgerman]

Ideally anything from 2A to about 8A will be fine. Terminate at a manually set 30mA or 4 hours CV whichever is sooner.

[expresso]

curious here - about this Pack your redoing now -

how large is this pack - is it 24V or ??

where are you fitting this pack - in a wheelchair ?

looks huge to me - i cant see how it fits in a wheelchair -

if using in a chair - what kind of chair is it ?

[cdb0ewm]

Clarification please

The Hyperion charging screen has 3 values that I want to make sure I understand
- Output Voltage - as I understand the guidance above after about 1 hour of post charging this should be 3.6 volts for a good cell. What if it is less?
- Output Capacity - is the the total capacity of the cell i.e. 12000mAh for a fully charged cell or the capacity added during the specific charging
- Charge Level % - after about 1 hour post charging I believe for a good cell this should be 100 or close to it. Is this correct?

Thanks

[cdb0ewm]

Sorry didn't see expresso questioins.

It is a BM3 capacity pack. It's going in a modified Quickie P222 frame - so far it all fits

[Burgerman]

- Output Voltage - as I understand the guidance above after about 1 hour of post charging this should be 3.6 volts for a good cell. What if it is less?

It will ALWAYS be less. And will fall to around 3.4x volts over time. Forget about this. Just be sure you leave it on charge for an hour AFTER it says done and starts beeping... Because its still charging. Esp if these are cells that have been left to discharge.

- Output Capacity - is the the total capacity of the cell i.e. 12000mAh for a fully charged cell or the capacity added during the specific charging

Its a measure of capacity REMOVED by discharging, with the charger, when it reaches say 2.5 to 2.8V under load. The actual figure will be less than 12 because of the way it works. You are looking for cells that are lower than average use same discharge Amps and same termination volts for all.

- Charge Level % - after about 1 hour post charging I believe for a good cell this should be 100 or close to it. Is this correct?

No, take zero notice of that. If its still on the charger it will read 100. If not it could read anything...

[cdb0ewm]

Thank you

[expresso]

Sorry didn't see expresso questioins.

It is a BM3 capacity pack. It's going in a modified Quickie P222 frame - so far it all fits

i dont even know what the BM3 pack size is -

i have the same chair also P222se - how did you get that thing to fit in that chair ? i find that chair to be very tight with the brace bar over the top - one battery fits one way - the other battery fits different way

thats pretty good if you can fit a large pack like that in there - is that 24V ?

whats modified about the Quickie P222 ?

i though about it a few times - to make it lithium also - i have a small ADD ON with that chair - i decided to leave it alone - i plan to end up using the P222 for indoor use once i get another new chair - in my case its a pain to remove the top off the base to check on things or change things etc, for example like what your doing -

its easy if you can do it yourself or have someone to do it for you when you need it done and for free like a friend - - no such thing here so i have to always rely on help and have to pay if you want them to help you -

[Burgerman]

Its a 72Ah, 45V pack, 13S. So uses 78 cells.
And stores 3.2kwk of power.

kWh is the way to compare different voltage packs.
Yours is 105Ah x 26.4V so is 2.77kWh.
So the BM3 type pack has around 18% bigger in total storage capacity. But it also gains a little efficiency too because of higher volts.

At least it would have done if done it properly!
And will again once rebuilt with care/new cells.

[expresso]

whow thats great - i wonder whats been done to that chair to make it fit - i have that chair P222 - you have to drop it in the case - no sliding out - have to remove the top half of the chair - then you have room to work thats for sure since the base is free and nothing in your way - just a cross bar over battery box - which can be removed - if it can be fitted back - thats another story -

that pack just looks BIG - i wonder how its fitted in that chair - that thing must be FAST then and FUN To drive it - that chair is light and tippy - wheelies are so easy with the P222 - i would love to see PICS of that project - chair etc,

what controller - joystick is used for this higher volt pack ?

78 12ah Cells ?

[Burgerman]

http://www.wheelchairdriver.com/BM-MK3- ... rchair.htm

The same as this!

[expresso]

thats great - love the chairs - speed - but in a P222se chair - that thing is going to Rock and Roll -

so the joystick - controller is the same as yours BM ? a custom setup - -

i wonder if someone can get a hold of the 48 V joystick and controller from Bounder - - that can be used on any chair then ? as long as you make the battery 48V ? or 45V etc -

if thats the case - that would solve the custom programming of this Robotech setup your using right ?