Two BC168 chargers on 12S pack?

[major]

I finally got around to trying to charge an EnerDel 12S module using two BC168 chargers. I am aware that that charger is not isolated from the power supply on the most negative cell. So I used a separate power supply for each BC168. And even went to a battery supply for one of them. So the inputs to the two BC168 units were definitely isolated. The module was sitting at 4 V/c and well balanced to within ~30mV/c.

BC168#1 connected to cells #1 thru #6 and works fine by itself. BC168#2 connected to cell #7 thru #12 and works fine by itself. But when I use them together, cells #6 and #7 show 4.2 Volts while the other 10 cells show 4.0 Volts. All cells show 1.0A in discharge mode. The same voltage reading occurs when in the balance mode. Obviously there is crosstalk between the two BC168 even though they are isolated (except for cell#6+ and cell#7-).

And idea why this happens? I really wanted to be able to use the two BC168 together on the 12S modules but can't tolerate the .2V offset.

 

[amberwolf]

I don't see how there could be any crosstalk given that wiring. :?

Obviously *something* is happening, but I don't understand what it could be.

If you leave both connected but only one powered on, what happens?

 

[major]

I don't see how there could be any crosstalk given that wiring. :?

Obviously *something* is happening, but I don't understand what it could be.

If you leave both connected but only one powered on, what happens?

If I disconnect the power supply on one, the other one which stays on goes back to reading normally as in all 6 cells reading 4.0V. Works that way when powering either one with both connected to the cells.

I have a wye harness for the BMS wires so I can hook in CellLog8s while the BC168s are connected. The CellLoggers confirm 4.2V on cells #6 & 7 when both BC168s are powered. The BC168s behave the same with and without the CellLoggers attached.

The battery module is housed in plastic and the chargers are separated on a Formica tabletop so there are no ground loops which I can see. All the wiring is clean and neat with no chance of a problem with that.

 

[friendly1uk]

Why not spend $110 on a bms and 400w charger? It would be a much better result. 24/7 protection and as hard to charge as your phone. Initial assembly is almost identical and you get more functionality for less money.

I know your stockpiling RC gear now, but I really would sweep the toy stuff in to the bin and get professional about it. It is the only fix (I would except) for this rather expensive mistake. I hope it was not people here that misguided you. Or maybe I'm misguiding you..... But I assure you, this RC crap is not sold on EV sites. It is not for us. Your tripping if you think it is. There is no fix for what you have. Only bodges.

 

[major]

Why not spend $110 on a bms and 400w charger? It would be a much better result. 24/7 protection and as hard to charge as your phone. Initial assembly is almost identical and you get more functionality for less money.

Thanks for your friendly advice. But I have numerous applications for these modules as well as several other types of cells on hand. No BMS which I have found fits the bill. I tried to purchase a Bestechpower board but they won't respond. Those are to small for full time protection anyway. I'll need 500A discharge in some cases and have a 6 kW bulk charger. I never intended to use the RC hobby equipment for regular EV duty.

In this case I was attempting to be able to slowly and automatically balance modules outside the vehicle. After a season of kart racing or yard tractor duty without BMS, the modules can drift out of balance sometimes by 300 mV but mostly only 50 or 60 mV on a few cells. It is a PITA to manually balance those.

BTW, I monitor with CellLog8s to watch for HVC and LVC and delta V. I wish there was a simple affordable BMS which could be attached to and stay with the module but I know of no such system.

I still would like to know why the isolated BC168s affect each other.

 

[dnmun]

i find it hard to believe that henry would not respond. he is devoted to the idea of supplying high quality BMSs to the consumer and bypassing the middle man. just the opposite of what you describe.

but i am sure you can figure out why there is this offset when you have to make do with some balancing chargers.

the BC168 is designed and built different from the other balancing chargers that use an internal BMS to balance the cells. it uses a series of single cell power supplies.

 

[friendly1uk]

Hello again
I really can't think why your chargers are messing about. You seem to of done everything right. I'm a tad busy now, but I will try and have a head scratch later. It's certainly not obvious.

If you want 500A, even with a bms, take it straight from the cells. A 6kw charger though? That will hit hvc long before any serious amount of balancing is done. I might be tempted to set up full time balancing above 90% charged. So they can level out after charging. It might trim a few down, meaning reduced capacity. Only reduced as much as they were out of balance though.

Tricky... You sure don't post trivial stuff do you

 

[major]

i find it hard to believe that henry would not respond. he is devoted to the idea of supplying high quality BMSs to the consumer and bypassing the middle man. just the opposite of what you describe.

He replied to my first inquiry but failed to respond twice when asked how to actually purchase the item. I give up easy when a guy doesn't want my money.

the BC168 is designed and built different from the other balancing chargers that use an internal BMS to balance the cells. it uses a series of single cell power supplies.

Yes, I am aware of this and is why I bought them. I have the Thunder 1220 which works but is slow to balance the 35Ah module. My hope was that the BC168 would be faster.

but i am sure you can figure out why there is this offset when you have to make do with some balancing chargers.

It is difficult to separate the EnerDel 12S module in two. I think I'll take a couple of 6S A123 modules I have and hook them up. That way I can measure potential between the two. I really don't want to pop open the BC168. Having 6 separate chargers in the BC168 I was surprised and disappointed to find channel one negative is common to the power supply negative. That sucks and is likely the root of my problem. I thought having isolated power supplies to each BC168 would overcome that.

 

[major]

Tricky... You sure don't post trivial stuff do you

Thanks for your interest I mess around with a lot of different equipment and tested a variety of cells. That big charger is used for high voltage packs, Manzanita PFC30. For the low voltage stuff, my typical charge is 6, 17 or 23A from a Soneil or DeltaQ.

 

[friendly1uk]

So you have two banks of 6, joined in a row. Each bank with a charger that should be an isolated supply. What is the output of the 12s doing? Is there anything connected to it. Anything at all that could be linking the packs pos and neg? (even road dirt)

 

[major]

O.K. Just completed some interesting tests. First I duplicated what I did last night with a different 12S EnerDel module. Same results.
View attachment 4

You can see 4.1V on cell6 BC#1 and cell1 BC#2. Separately all 12 cells = 3.9V.

Next I took two 6S A123 modules I had hanging around.

All cells were at 3.3/3.4V when on both BC168 were on and the two 6S modules not connected together. When I connected the two 6S modules together to essentially make a 12S, the BC168 readings remained the same. You can see the Fluke reading 40.87V for the 12S and the BC168 reading 3.4V/c.

So what gives? For the hell of it I tried a test on an EnerDel 12S module which I had repaired the BMS connector. The standard module comes with a flexible printed circuit cell connector for the BMS and thermistor connections which I have been using. This repaired module was damaged by a student on the eKart race team and I replaced the flex connector with fused #22 wires.

It has a few low cells (1-3) but showed no crosstalk when the two BC168 are powered together.

A photo of the repaired module next to the standard.
Here is a shot of that flexible printed circuit BMS/thermistor connector.

So my conclusion is that somehow the BC168 crosstalk is caused by this flex circuit connector. I do not know how. But I'll probe around on it and see find I can determine why.

 

[dnmun]

can you read the voltage if the charger is not powered up? in that case you should try charging at the two poles with a bulk charger separate from these chargers and see if the two cells then show the same delta voltage.

 

[major]

can you read the voltage if the charger is not powered up? in that case you should try charging at the two poles with a bulk charger separate from these chargers and see if the two cells then show the same delta voltage.

Here is a shot of the two BC168 powered and attached to the EnerDel cells before pressing the mode button (charge, discharge or balance). All cells read the correct voltage.



I normally charge with a bulk charger and monitor with CellLog8s. All cells read normally in that case. Here is an example from a while back.

 

[dnmun]

what about reversing them and using #1 in the place where 2 is now and use #2 where 1 is now. check with bulk charger and then power up and check for location of the delta V.

 

[friendly1uk]

I would pop a cap across the end cells to try and clean up the noise.

 

[major]

what about reversing them and using #1 in the place where 2 is now and use #2 where 1 is now. check with bulk charger and then power up and check for location of the delta V.

Did that this morning.

View attachment 1

No change. Problem stays with cells #6 & 7 in the module, not with the channels in the chargers. BTW, I checked the actual cell voltages with the Fluke and they are 3.9V. So it is apparently the instrumentation readings which are erroneous and not a over potential on the cells. But how can I count on the BC168 to balance when one channel is off by .2V?

And I hooked up my Thunder 1220 to this module and it operates normally at about 5A charge and all cell voltages read normally (3.9V).

 

[major]

I would pop a cap across the end cells to try and clean up the noise.

I tried that with the smallest cap I could find around here. Not marked but about the size of a dime. High voltage like 1000 I think. It made no difference when place across the cells in question. I'll see if I can find some better filter caps. I have bunches of caps but like the size of soup cans or bigger.

 

[friendly1uk]

Just to be sure I'm up to speed..
Using either the bc168's or the 1220 (a 12s charger I presume) as voltmeters, all cells measure the same at rest
Using either the bc168s as chargers, or some other bulk charger, the bc168s report this offset reading.

Having done that test, the capacitors idea is looking thin. Can you use the 1220 as a voltmeter and apply a bulk charger to the pack? If the fault appears it confirms it is the pack, Not a combination of pack and bc168

 

[major]

Just to be sure I'm up to speed..
Using either the bc168's or the 1220 (a 12s charger I presume) as voltmeters, all cells measure the same at rest
Using either the bc168s as chargers, or some other bulk charger, the bc168s report this offset reading.

Having done that test, the capacitors idea is looking thin. Can you use the 1220 as a voltmeter and apply a bulk charger to the pack? If the fault appears it confirms it is the pack, Not a combination of pack and bc168

No, the offset reading does not happen with bulk charging. Only with the two BC168. And always on cells 6 & 7 in the 12S EnerDel module using the OE flex connector.

You ask _Can you use the 1220 as a voltmeter and apply a bulk charger to the pack?_ Isn't that what the 1220 does when it charges? I am hesitant to connect the 1220 and another charger like a DeltaQ for fear of a ground loop frying something. The CellLog8s act as voltmeter when bulk charging.

 

[Burtie]

I

BC168#1 connected to cells #1 thru #6 and works fine by itself. BC168#2 connected to cell #7 thru #12 and works fine by itself. But when I use them together, cells #6 and #7 show 4.2 Volts while the other 10 cells show 4.0 Volts...
And idea why this happens? I really wanted to be able to use the two BC168 together on the 12S modules but can't tolerate the .2V offset.

If the two chargers share a common wire to connect to Cell 6-7 point.
Because the timing of the two chargers is not synchronised , One charger will be measuring the voltage on this wire while the other may be trying to pass current on the wire. This will lead to an innacurate voltage being read for both cell 6 and cell 7.

To prevent this effect, you need to run independent wires from each charger all the way to the cell 6-7 joint. Then voltage drop accross 1 wire (due to charging current) will not affect the voltage reading on the other wire. The chargers should then operate correctly.

 

[friendly1uk]

I

BC168#1 connected to cells #1 thru #6 and works fine by itself. BC168#2 connected to cell #7 thru #12 and works fine by itself. But when I use them together, cells #6 and #7 show 4.2 Volts while the other 10 cells show 4.0 Volts...
And idea why this happens? I really wanted to be able to use the two BC168 together on the 12S modules but can't tolerate the .2V offset.

If the two chargers share a common wire to connect to Cell 6-7 point.
Because the timing of the two chargers is not synchronised , One charger will be measuring the voltage on this wire while the other may be trying to pass current on the wire. This will lead to an innacurate voltage being read for both cell 6 and cell 7.

To prevent this effect, you need to run independent wires from each charger all the way to the cell 6-7 joint. Then voltage drop accross 1 wire (due to charging current) will not affect the voltage reading on the other wire. The chargers should then operate correctly.

I just read elsewhere these b168s charge through the balance leads, which are a foil in this case. You have this nailed I reckon.

Hat off to you Burtie.

 

[major]

So you think I have 0.2V drop at 1 Amp on the foil trace? Good theory. I'll check that out, but will have to wait a day or two until I get back in town.

Too bad there is no way to synch the BC168 chargers.

 

[major]

I

BC168#1 connected to cells #1 thru #6 and works fine by itself. BC168#2 connected to cell #7 thru #12 and works fine by itself. But when I use them together, cells #6 and #7 show 4.2 Volts while the other 10 cells show 4.0 Volts...
And idea why this happens? I really wanted to be able to use the two BC168 together on the 12S modules but can't tolerate the .2V offset.

If the two chargers share a common wire to connect to Cell 6-7 point.
Because the timing of the two chargers is not synchronised , One charger will be measuring the voltage on this wire while the other may be trying to pass current on the wire. This will lead to an innacurate voltage being read for both cell 6 and cell 7.

To prevent this effect, you need to run independent wires from each charger all the way to the cell 6-7 joint. Then voltage drop accross 1 wire (due to charging current) will not affect the voltage reading on the other wire. The chargers should then operate correctly.

Simple test confirmed you are correct. I took the spare flex circuit pictured previously and found the trace to cell#6-#7+. I used a power supply and 5Ω resistor to bias 1 Amp thru the trace. Then probed the total distance of the trace with Fluke multimeter and read .24V. At 2A it was about .5V. Instruments weren't calibrated and probably not entirely accurate, but close enough to confirm in my mind the reason for the discrepancy on the BC168 readings.

Thanks Burtie.

 

[999zip999]

Sounds you need a 12s balance chargher or a bms and bulk charger.

 

[major]

Sounds you need a 12s balance chargher or a bms and bulk charger.

Thanks. I have bulk chargers and I have a Thunder1220 12S balance charger. Nobody makes a BMS to suit my needs. I've been searching for a year.