Heltec (JK) 200A Smart BMS with 2A Active Balance

[phaberest]

Dude, check the resistance. Watch this video. He starts checking the resistance at about 4:00

Thank you Arthur, I watched both videos and definitely learnt something! I will take care about checking everything and make sure to replace anything that doesn't look adequate

I made the internal copper bars myself with 5mm x 2cm bars and I might need to check they are good enough, in the second video Ray shows the bars might be slightly bent and this is something I missed checking. They might certainly be since I drilled them locking them to the bench

 

[hwse]

If you do not have a very good resistance tester like Ray used in the video, at least check everything for hot spots. If you have a thermo-camera (flir) that is the best resistance tester. If not, just an infrared thermometer or even checking (carefully) with your hand will find connections that are getting warm or hot. Heat = resistance.

 

[Choco]

B2A24S15P (2A/16S/150A)
I've had an issue pop up on my BMS, has anyone seen this...

-Balancing has stopped working
-The highlighted High & Low cells are stuck (i.e it does not dynamically move from cell to cell)

Was thinking of unplug/replug the balance cable harness, any suggestions?
It's been working well for '1Y200D', hoping it hasnt blown something, it never sees over 70A discharge.

 

[hwse]

B2A24S15P (2A/16S/150A)
I've had an issue pop up on my BMS, has anyone seen this...

-Balancing has stopped working
-The highlighted High & Low cells are stuck (i.e it does not dynamically move from cell to cell)

Was thinking of unplug/replug the balance cable harness, any suggestions?
It's been working well for '1Y200D', hoping it hasnt blown something, it never sees over 70A discharge.

Could it be that your balancing threshold is greater than the imbalance in your pack?

 

[labeeman]

B2A24S15P (2A/16S/150A)
I've had an issue pop up on my BMS, has anyone seen this...

-Balancing has stopped working
-The highlighted High & Low cells are stuck (i.e it does not dynamically move from cell to cell)

Was thinking of unplug/replug the balance cable harness, any suggestions?
It's been working well for '1Y200D', hoping it hasnt blown something, it never sees over 70A discharge.

If the cells are not above 3.400 VDC it will not balance either

 

[Choco]

My balance is set to come on at 3.38 and the imbalance while charging is defintely above the threshold.
I adusted the threshold even lower but it still would not trigger.

 

[Choco]

Disconnected B-, then jump started it again, seems to be all good again.

 

[ichernev]

Hello guys! I have recently installed this bms (JK-B2A24S20P(8-24S 200A)) with 16s 280k EVE cells, and started discharging with 5kw XWJNE 48 V pure sine wave inverter (amazon.de) with 20/40A (1/2kW). And the bms started spitting out

abnormal cancellation of coprocessor communication
abnormal coprocessor communication

every few seconds. There were no issues with the load (it's an electric heater, so I doubt it would have a problem), but I stopped it just in case. Is this something to worry about or just business as usual? Somebody in here https://diysolarforum.com/threads/jk-4s-200a-bms.38623/post-519498 mentioned it could be related to THD/not very good inverter. I don't have any advanced (scope) measuring equipment, just a few multimeters.

 

[Sanwizard]

Hello guys! I have recently installed this bms (JK-B2A24S20P(8-24S 200A)) with 16s 280k EVE cells, and started discharging with 5kw XWJNE 48 V pure sine wave inverter (amazon.de) with 20/40A (1/2kW). And the bms started spitting out



every few seconds. There were no issues with the load (it's an electric heater, so I doubt it would have a problem), but I stopped it just in case. Is this something to worry about or just business as usual? Somebody in here https://diysolarforum.com/threads/jk-4s-200a-bms.38623/post-519498 mentioned it could be related to THD/not very good inverter. I don't have any advanced (scope) measuring equipment, just a few multimeters.

I never saw that, but I was not communicating with mine with the inverter. Where is it spitting out the errors to? Is it the inverter or a Raspberry Pi for comms?
It could be a logic error internally though, or maybe a loose connection.

 

[ichernev]

I never saw that, but I was not communicating with mine with the inverter. Where is it spitting out the errors to? Is it the inverter or a Raspberry Pi for comms?
It could be a logic error internally though, or maybe a loose connection.

Well, my inverter is also pretty dumb, so no communication there. I see it in the JK BMS app, it flashes in the main screen (AUXCPU or CPUAUX) and the messages I pasted above are from the event log. Also, it looks like some of the information disappears briefly and then re-appears -- it looks like there is a standalone chip (auxcpu) for some of the measurements, and that goes bad and gets rebooted from time to time.

I torqued all sense leads to 5nm, nothing moves (when I try to wiggle it), the main P- <-> B- line is also pretty solid, not sure what else could be loose. The JST-XH connectors are also pretty snug. It didn't do that while I was charging or when just idling. It does it only when discharging.

 

[Sanwizard]

Well, my inverter is also pretty dumb, so no communication there. I see it in the JK BMS app, it flashes in the main screen (AUXCPU or CPUAUX) and the messages I pasted above are from the event log. Also, it looks like some of the information disappears briefly and then re-appears -- it looks like there is a standalone chip (auxcpu) for some of the measurements, and that goes bad and gets rebooted from time to time.

I torqued all sense leads to 5nm, nothing moves (when I try to wiggle it), the main P- <-> B- line is also pretty solid, not sure what else could be loose. The JST-XH connectors are also pretty snug. It didn't do that while I was charging or when just idling. It does it only when discharging.

Sending items back to China gets expensive. If you bought from a good source, perhaps they would send you a replacement? Its worth a try. I bought a charger that ended up DOA on Alibaba, and the vendor simply send me a new one at zero cost after showing them the issue. Maybe you will get lucky too.

 

[adinindu]

Looking for advice.

I have EVE105 brand new 16S pack with JK smart BMS. Cells are perfectly Top and Bottom balanced a while ago. However, when i'm drawing more than 20A, about 0.2C-0.5C - I see Delta Cell voltage rise up for about 50mv. But it goes down to under 5mV after load disconnected. Is this normal with good quality cells or is BMS cant cope with the voltage difference to actively balance each other under load ?

Expert help would be much appreciated.

 

[ichernev]

Sending items back to China gets expensive. If you bought from a good source, perhaps they would send you a replacement? Its worth a try. I bought a charger that ended up DOA on Alibaba, and the vendor simply send me a new one at zero cost after showing them the issue. Maybe you will get lucky too.

Heh, I wanted to see how the 4 main wires were tied to the board (I was wondering if I should replace them), so I opened it up (broke the seal). I'm not really into returning stuff (I'd better get a new bms and disassemble the old one for fun, some forum members might be interested as well). I'm just trying to find out if that is normal/seen/experienced. As I said it worked -- i.e the load continued to draw power, the bms continued to show stats, just spitting this error from time to time and dropping some stats for a second or so. So I might just use it some more and see if the problem is critical or not (i.e will it fry the BMS eventually or just complaining).

 

[John Frum]

Cells are perfectly Top and Bottom balanced a while ago.

You have to pick one or the other as you can't have both.
If you are using a JK BMS it kinda has to be top balance.

 

[ichernev]

Looking for advice.

I have EVE105 brand new 16S pack with JK smart BMS. Cells are perfectly Top and Bottom balanced a while ago.

You can't have top AND bottom balanced cells. They are either top balanced (you fill them to the top together (in parallel)) or you drain them together in parallel, and then you hook them up in series. If they are both, this suggests they are best quality "matched" cells (which I doubt many people around here use due to price), which are top balanced, and by their virtue of being identical, they'll be also bottom balanced.

However, when i'm drawing more than 20A, about 0.2C-0.5C - I see Delta Cell voltage rise up for about 50mv. But it goes down to under 5mV after load disconnected. Is this normal with good quality cells or is BMS cant cope with the voltage difference to actively balance each other under load ?

The balancing in most BMS-es happens when they are charging and when they're very high in their SOC. That is because when they're not, LiFePo are so close in voltage that it's meaningless to balance them (i.e a cell with higher voltage (slighly) might actually be less SoC than a cell with lower voltage. When you discharge with higher current (well, not very low), then they are under stress, and because they are not EV-grade (as I said above, most cells people use are not) they experience it differently. But in any case the BMS won't bother and shouldn't, and also, if after load is stopped they're close, it means it shouldn't have in either case.

 

[RCinFLA]

Looking for advice.

I have EVE105 brand new 16S pack with JK smart BMS. Cells are perfectly Top and Bottom balanced a while ago. However, when i'm drawing more than 20A, about 0.2C-0.5C - I see Delta Cell voltage rise up for about 50mv. But it goes down to under 5mV after load disconnected. Is this normal with good quality cells or is BMS cant cope with the voltage difference to actively balance each other under load ?

Expert help would be much appreciated.

Make sure you do not set start of balancing below 3.40v. Set delta to 5 mV.

Problem is either a poor bus bar connection or unmatched cells.

Realize each BMS sense voltage wire for cells are shared between two cells with a bus bar in series with each cell voltage measurment (except one cell in series stack). If you have a poor bus bar connection with excessive resistance when there is high current there will be a voltage drop that will add or subtract from BMS cell voltage measurement.

If cells are not well matched their overpotential voltage slump with load current will be different. As cell ages its internal impedance gets greater. Matched cells will have same terminal voltage slump for same cell current load. You have to measure directly on cell terminals base to avoid influence of bus bar connections.

 

[adinindu]

You can't have top AND bottom balanced cells. They are either top balanced (you fill them to the top together (in parallel)) or you drain them together in parallel, and then you hook them up in series. If they are both, this suggests they are best quality "matched" cells (which I doubt many people around here use due to price), which are top balanced, and by their virtue of being identical, they'll be also bottom balanced.

ok, These cells came with EVE Shipment report, Excel datasheet as well as the matching EVE box. Hence i assumed these are matched cells.

The balancing in most BMS-es happens when they are charging and when they're very high in their SOC. That is because when they're not, LiFePo are so close in voltage that it's meaningless to balance them (i.e a cell with higher voltage (slighly) might actually be less SoC than a cell with lower voltage. When you discharge with higher current (well, not very low), then they are under stress, and because they are not EV-grade (as I said above, most cells people use are not) they experience it differently. But in any case the BMS won't bother and shouldn't, and also, if after load is stopped they're close, it means it shouldn't have in either case.

I never liked the default busbar they came up with and waiting on flexible bus bar set. I will do a retest, once they are fixed and post an update. Thank you very much for the clarification.

 

[adinindu]

Make sure you do not set start of balancing below 3.40v. Set delta to 5 mV.

Problem is either a poor bus bar connection or unmatched cells.

Realize each BMS sense voltage wire for cells are shared between two cells with a bus bar in series with each cell voltage measurment (except one cell in series stack). If you have a poor bus bar connection with excessive resistance when there is high current there will be a voltage drop that will add or subtract from BMS cell voltage measurement.

If cells are not well matched their overpotential voltage slump with load current will be different. As cell ages its internal impedance gets greater. Matched cells will have same terminal voltage slump for same cell current load. You have to measure directly on cell terminals base to avoid influence of bus bar connections.

Cells are manufactured in 2022/10 and came with all required test report to verify. Anyway I'm waiting on a new busbar set and hopefully it will resolve this issue. If not, thats a whole new problem - given the cells were perfectly matched when arrived and balanced.

 

[ichernev]

Cells are manufactured in 2022/10 and came with all required test report to verify. Anyway I'm waiting on a new busbar set and hopefully it will resolve this issue. If not, thats a whole new problem - given the cells were perfectly matched when arrived and balanced.

View attachment 136527

My apologies, you got "official" cells. Still I'm not sure they're balanced (factory discharges them so they stay in ~30% SoC, for a lot of time, even if they were balanced 1y ago they are not now). You can read the balance tutorial, but in general, charge them to the top in the 16S config, then you can individually top them off to 3.65V. If they're balanced it would be very quick, if not -- less quick but in either case you'd know for sure they are balanced. BMS balancing is kinda only keeping it balanced once it's balanced. If they're not balanced now it might take a long time for the puny balance current.

 

[adinindu]

My apologies, you got "official" cells. Still I'm not sure they're balanced (factory discharges them so they stay in ~30% SoC, for a lot of time, even if they were balanced 1y ago they are not now). You can read the balance tutorial, but in general, charge them to the top in the 16S config, then you can individually top them off to 3.65V. If they're balanced it would be very quick, if not -- less quick but in either case you'd know for sure they are balanced. BMS balancing is kinda only keeping it balanced once it's balanced. If they're not balanced now it might take a long time for the puny balance current.

Yah, I did try that. With balance enabled, and charging ends - cells get balance within 20mins. Even while charging delta V will never exceed 10-12mv. I charged at 2Amp until COVP kicks in.