Trouble balancing headway pack

W

Wurly

100 W
Joined
Dec 29, 2008
Messages
156
Location
Somerset, UK
One of my packs cutout on me last week, so before i charged it, i measured the cell volts to see which one might have been the culprit .
1 2.714v
2 2.885v
3 3.222v
4 2.688v
5 3.199v
6 3.209v
7 3.221v
8 3.221v
9 3.211v
10 2.929v
11 3.264v
12 3.267v

Although my pack takes a while to balance ok (charger shuts off and switches on many times) i was a bit disappointed to see about 3 cells quite a bit lower than the others. I recharged the pack, saw a few leds switching on and off, went back later and they are all off....odd?
So, i thought i'd ride the bike a couple of times, discharge the pack and re-charge.
This time i watched the pack while charging and sure enough , there is balancing going on but there is a long time between the charger off and on times and no leds illuminated....as if BMS(signalab) was shutting off charger.
I remeasured the cell volts after charging
1 3.369v
2 3.383v
3 3.396v
4 3.377v
5 3.459v
6 3.902v...wow!
7 3.413v
8 3.450v
9 3.511v
10 3.393v
11 3.492v
12 3.415v
Again, charger switches off and doesn't switch back on. I presume the BMS is shutting down because that channel reaches of HVC?
Why would cell 6 be so much higher than the other cell volts?
 
Either your BMS is dieing, or your pack is near the end of it's life and the cells no longer have matched capacity.

Try charging each cell individually with an RC charger. Get them within 0.01 volts of each other manually, then take it for a ride down close to LVC. if the cells are still reasonably balanced, then they're still matched and the problem is the BMS. if they aren't balanced at the end of the ride, the pack is at the end of it's life and it's time to replace it. .. or just start manually balancing it every ride.
 
Although the cells were 2 years old when i bought them, they had not been used. I measured the IR with a load on each cell and they were all around 4mohms. I never did a capacity check though! The pack has only 10-15 cycles on it and i had it nicely balanced a short while ago.
I could try another known good BMS, or change cell6 (i have some a couple spares).
 
why do you think the BMS is not functioning when it clearly is functioning?

it sounds like the battery was assembled by someone who knows nothing about battery construction and just never balanced the cells to begin with. there is no evidence you have done anything to balance them either.

there is no mention of the BMS that you are using. no pictures.

you can use a power resistor or flashlight bulb to drain down the high cell to force the BMS to continue charging.

the pack is so much out of balance it will have to remain on the charger for weeks or a month if you expect the BMS to actually balance it but i suspect you just remove it from the charger when the light turns green anyway so i don't see how it ever will balance if you continue doing that.
 
A lot of assumptions there.
I charged each individual cell with an imax b6 charger when i built the pack. I know that charger doesn't have a great reputation on here, but it got every cell voltage pretty close ( i have the cells voltages written down somewhere). After doing that i cycled the pack with minimal discharge and put it back on the charger and let the BMS balance it. Several times. All leds were illuminated after leaving it on the charger balancing. At that point i was confident the BMS was doing its job well.
I have no reason to think the BMS isn't working. I didn't say that. i said i have a spare that i could try.
If the pack has become unbalanced through use, i would like to know why? i ain't an expert...just learning as i go.
And i do not unplug the charger when the green light goes out, i wait until all leds are on except in these two cases i mentioned. (I needed to use the bike)
BMS is signalab, pictures to follow
 
It's been a few days so i measured cell volts again because i was going to charge each cell individually,. No need, all cells between 3.23 and 3.25V so i put the pack back on charge.
So far all cells are around 3.4V with 2.5Ah into it part way though charging. Must have gone way off balance when i discharged it right down. Cells are probably mismatched but looks like it'll balance back ok (my theory fwiw).
 
Yep, you just drove it further out of balance than usual. BMS seems to work to me, those high cells were just what it does when you get the pack that far out of whack. I've seen 4v many times on an old ping pack when it was getting old. The cells I did not sell to Dnmun would fill crazy fast.

Next time it gets that far out of balance, just do a manual balance with the RC charger. Run your regular charger till the first cells hit 3.65, then bring any really low cells up to 3.5v with the RC charger one by one. Then put it back on the regular charger to bms balance. PITA, but likely quicker than waiting on the bms.
 
Well this is the one that climbs to 3.9V, all the other cell volts were increasing nicely. So i bled this one down and plugged in the charger drained the pack a little, charged again and finally watched this one catch up. I also started to see other cells reaching 3.9v while this was happening.



And eventually all leds on



So how long should i leave the pack connected to the charger once all leds are on?

I am also wondering if this BMS is the best one to use for these headway cells? does the signalab BMS bleed down cells volts or does it just switch them off? should i use another design BMS? :?:
 
'One more thing. When you build this new headway pack, a few extra wires spliced to your balance wire harness will allow monitoring the cells manually with a cellog 8. Three 8s jst plugs will do it. Being able to watch the cells while they charge can be very reassuring. And much easier than probing cells one by one with a DVM.'

Just picked this statement by dogman,on johnnyz thread...... yep, damn good idea. I need a cellog
 
if all the cells are at 3.9v then your charger voltage is set far too high. the v2.5 signalab turns on the shunt transistor at 3.60V so if you have a 12S such as this then the charger voltage should be 43.2-43.4V. i use 43.8V for discharge testing.

you can test the capacity of the cells by fully charging it to 3.65V/cell and then discharging to 2V/cell.

you should discharge through an accurate wattmeter and record the total amount of Ah discharged when the lowest cell hits the LVC first.

if you do not use a separate auxiliary power source for the wattmeter it will lose the Ah values when the BMS turns off the battery so you need to monitor the cell voltages closely at the end if you don't use a aux power. but you should monitor the cells all the time as you discharge so you know which are strongest and which are weakest, and then take out the weak ones and replace them with stronger ones.

i discharge into a radiant floor heater and oil bath type heaters.

that may help you understand more about how a battery works.
 
Once all lights are on it's done needing the charger. It just needs to bleed off the high cells. I agree, if you have all cells at 3.9v, then you do need to turn the charger down. 44v at the very most.
 
This is interesting. Similar to my situation with the signalab BMS. In my case the voltage on my new garbage cells began increasing very fast up and over 4 volts while the others didnt move.
The only logical explanation in your case must be that the particular cell that reaches 3.9 volts so quickly is beginning to lose its capacity. Lower capacity cells will charge up much quicker than cells that have larger capacity. The real question is...the function of the signalab BMS itself. It appears and please someone correct me here if im wrong...that when a cell begins to lose capacity in relation to the others in the pack the signalab bms becomes wholly inefficient. Im not an expert on this by any means but they way it balances a pack is ok as long as the pack has very similar capacities. It would be better if the BMS design were to "shut off" a cell that reaches its maximum voltage cutoff rather than attempt to "bleed the voltage".

Dnmum says that your charger must be too high. What is the charger output in volts and amps?...seems to be to be of lessor concern since the difference of 3-4 volts going into all the batteries would still result in the lower capacity cell hitting the upper threshold quicker than the others. That is why i am going to make sure all of my headways (when i get them) are extremely close before charging but not more than 3.4 volts as i want to see how the signalab bms handles it from there. I will have cell loggers on the cells so i can monitor them. Im assuming that at 3.4 volts none of the 24 headways should "jump" to over 3.65 volts whilst the others try to play catch up, otherwise the logical assumption is that the one that does rise is lower somehow in capacity.


John
 
i use a big 10A single cell charger and for the headways i mount 6 at a time in a strip of plumbers tape. the holes in the tape are just big enuff for the headway screw so i can charge 6 headway cans at once up to 3.65V.

but i don't do stuff like you guys because i test capacity of each cell before i include it in the battery.

so once i have the headway cans all charged up, i assemble them into a 12S 36V battery under a v2.5 signalab and discharge the 12S pack into a radiant heater through my wattmeter. then i can measure the capacity of each can accurately so i know how close it is to spec.

i do the same with all of these ping pouches too. i cut the batteries apart so that i have individual pouches that i can charge up together in parallel through the tabs with some alligator clips on jumper wires to 3.65V and then i lay them out on my work table with the pouches connected in series and discharge them through the watt meter into a radiant heater.

that way i can measure the capacity of each pouch accurately and i record the measurements on the pouch. just finished doing this on about 150-200 pouches so far. this is how i am assembling all my ping pouches into a big 24S pack for my ZENN car.

i am up to 75Ah now and gonna add another 12Ah that i am working on today. that will get me to 87Ah wide.

but since i know all these cans and pouches are close to the same point of discharge, i assemble the battery out of these discharged cells and then after i put the battery together, i charge it up for the first time to see how close the battery is when they all climb from their discharged state. watching to see the order in which each channel reaches full charge voltage is important information for me to understand the health of the assembled pack.

i would not do it any other way myself since i have to know what i am putting into the pack because once i solder the tabs together and put the battery together, i do not wanna have to cut it open again.
 

Attachments

  • DSC01769.JPG
    DSC01769.JPG
    58.2 KB · Views: 12,537
dnmun said:
i use a big 10A single cell charger and for the headways i mount 6 at a time in a strip of plumbers tape. the holes in the tape are just big enuff for the headway screw so i can charge 6 headway cans at once up to 3.65V.

but i don't do stuff like you guys because i test capacity of each cell before i include it in the battery.

so once i have the headway cans all charged up, i assemble them into a 12S 36V battery under a v2.5 signalab and discharge the 12S pack into a radiant heater through my wattmeter. then i can measure the capacity of each can accurately so i know how close it is to spec.

i do the same with all of these ping pouches too. i cut the batteries apart so that i have individual pouches that i can charge up together in parallel through the tabs with some alligator clips on jumper wires to 3.65V and then i lay them out on my work table with the pouches connected in series and discharge them through the watt meter into a radiant heater.

that way i can measure the capacity of each pouch accurately and i record the measurements on the pouch. just finished doing this on about 150-200 pouches so far. this is how i am assembling all my ping pouches into a big 24S pack for my ZENN car.

i am up to 75Ah now and gonna add another 12Ah that i am working on today. that will get me to 87Ah wide.

but since i know all these cans and pouches are close to the same point of discharge, i assemble the battery out of these discharged cells and then after i put the battery together, i charge it up for the first time to see how close the battery is when they all climb from their discharged state. watching to see the order in which each channel reaches full charge voltage is important information for me to understand the health of the assembled pack.

i would not do it any other way myself since i have to know what i am putting into the pack because once i solder the tabs together and put the battery together, i do not wanna have to cut it open again.
Click to expand...

So in essence, what i am proposing once i get the headways is not unlike what you do.

Charging them all up to 3.4 volts (so that there is something left to charge) and then hooking up to the signalab bms and with the addition of cell loggers, watch and see the voltages rise...they should all be relatively the same, then i know the capacity of each is the same. If one begins to rise much quicker than the others or if the bms has trouble controlling the HVC of one then i know i may have a problem with that individual cell.

John
 
3.4v is pretty close to top of charge, but still in the zone where it will be hard to say what state of charge is. If you want to find out how your bms handles a cell 10% out of balance, fully charge them all, then discharge one 10%. Then you know what is happening and why.

Even the first charge to 100% is going to result in a few cells that have not developed their full capacity imo, unless you have charged them fully and done a few low dod cycles.

Just charge them to 3.5v-3.65v. Then do some low dod cycles. Like a ride around the block. You will get to see the bms work plenty, as the slower to perk up cells get to charging fully. On my ping packs, it took 3-4 light cycles to start balancing up quicker.
 
dogman said:
3.4v is pretty close to top of charge, but still in the zone where it will be hard to say what state of charge is. If you want to find out how your bms handles a cell 10% out of balance, fully charge them all, then discharge one 10%. Then you know what is happening and why.

Even the first charge to 100% is going to result in a few cells that have not developed their full capacity imo, unless you have charged them fully and done a few low dod cycles.

Just charge them to 3.5v-3.65v. Then do some low dod cycles. Like a ride around the block. You will get to see the bms work plenty, as the slower to perk up cells get to charging fully. On my ping packs, it took 3-4 light cycles to start balancing up quicker.
Click to expand...

Ok you have me convinced. When i get them ill hook up 4 sets of 6 cells initially in series and charge to 3.6 volts. Once done ill put them together in a pack and then do a short test around the block, re-test and charge...all the while monitoring with cell loggers so i can see whats going on. Im going to fully document this on here when i get them.

John
 
Still if all are at 3.45v. This pretty full. 3.59v or 3.65v is tops for a charge daily battery. Yesterday i charged my A123 20ah pouches to 3.75 to 3.82v with a cheapo 36v 3a. 43.8v bluck charger to see how they would run up to the top without bms and from 3.59 balance charge to shutoff. Then the bluck charger and all cells to 3.75v to 3.82v. I know this is no test. But different types of cells have more vairable in I.R. and don't balance as well as anothers.
Meaning all batteries are different even from the same box of cells. Plus I don't use a bms because I'm hauling around 20ah cells.
 
You must log in or register to reply here.
Back
Top