Repair of BMS ? ISD 10S 40A

jerome_speedy

100 mW
Joined
May 25, 2015
Messages
44
Location
Lyon, France
Hi all,

I have an old problem that I want to fix with the help of people here (if they can/will). My battery is 10S 8P Li-Ion.
I've noticed for some time that It wouldn't take the charge anymore (displays 0A of charge current if I try). I have postponed the repair until I could afford to spend some time to research the cause of breakdown and fix it. I don't want to change the BMS if I can avoid it.

So far my tests:

B+/B- = B+/Cell- = 40,2 V
B+/LOAD = 34.4 V
B+/PACK- = 33.9 V

Cells0: 4,16 V
Cells1: 4,18V
Cells2: 4,17V
Cells3: 4,05V
Cells4: 3,98V
Cells5: 3,93V
Cells6: 3,90V
Cells7: 3,94V
Cells8: 3,93V
Cells9: 3,93V

I'm well aware that the cells are way out of balance but this is not the cause of my problem (I think). I used an active balancer at one point in time to get there (the balance was even worse before).

While "charging" at 42V (the charge socket being tied to a 42V voltage source), the (presumably) charge mosfet exhibits:
V G-Drain = 1,5 V
V Drain-Source = 1,9V
I think it's normal from that data that the BMS prevents the charge since the Gate-Drain voltage is too low (the threshold for activating the FET is around 3-4V if I recall correctly). So I suppose some part must be playing up and blocking the FET somehow. I checked the other "repair BMS" threads on this forum and concluded that it couldn't be that a sense wire is bad, judging from the cells voltage. But this is as far as I can go.

I checked the discharge FETs and they are willing to send "voltage" to the output, since:
V G-Drain = 5V and V Drain-Source = 5V (bizarre ?). I also noticed that when a suitable voltage (ie 42V or so) is applied to the charge socket, the discharge FETs shuts down (V G-Drain= 0V, V G-Source= 0V), so THAT BMS doesn't like charge and discharge at the same time (I noticed more recent chinese BMS accept that, usually).

If anyone has some more tests I could perform, I'm all hears... Thanks !

PS: I have a feeling the BMS could stop the charge because of too much imbalance of cells. Is that common in old BMS ? The next thing I want to try is charge the lowest cell to 4,2V, connect my active balancer, wait for a few hours and when the voltages are settled, try to charge the battery another time. What do you think of it ?
 

Attachments

  • P_20210415_115250.jpg
    P_20210415_115250.jpg
    349.9 KB · Views: 280
jerome_speedy said:
My battery is 10S 8P Li-Ion.
<snip>
B+/B- = B+/Cell- = 40,2 V
B+/LOAD = 34.4 V
B+/PACK- = 33.9 V

Cells0: 4,16 V
Cells1: 4,18V
Cells2: 4,17V
Cells3: 4,05V
Cells4: 3,98V
Cells5: 3,93V
Cells6: 3,90V
Cells7: 3,94V
Cells8: 3,93V
Cells9: 3,93V
Cells10: 3,93V

You list 11 cells, but if it's a 10s pack, where are you measuring the 11th cell voltage?

Also, which cell do you call "Cells0"--is it the most positive, or the most negative?

There are 11 test points TB0 thru TB10 on the edge of the board by the balance shunts; is that where you are testing? I suspect (but can't tell visually) that TB0 is the negative side of the most negative cell (Same as main battery negative).




While "charging" at 42V (the charge socket being tied to a 42V voltage source), the (presumably) charge mosfet exhibits:
V GND-Drain = 1,5 V
V Drain-Source = 1,9V
Do you mean "Gate" rather than "GND"? Because GND would be ground, or battery negative. Gate is the pin that is driven to a certain voltage to turn on the FET, which then connects Drain to Source so current can flow.

Usually when looking at the "front" (labelled side) of a FET, with the heatsink tab pointing up, the pins are Gate, Drain, Source, You'd have to verify the datasheet for your specific FETs to be certain of that, but it's pretty standard.

When testing to verify Gate voltage, you'd put your meter's black lead on Source (or system/board ground, or battery cell block negative), and red on Gate. For your charge FET, I think TP5 is the gate drive test pin. Depending on the FET part number, it could be a 5v gate drive, meaning it's "off" at 0V, and fully on at 5v. In between you'd need to see that part's spec sheet to find out what happens. Other FETs can be much higher gate drive voltages, some up to 12v or 15v or more.

If there is not enough gate voltage, then the FET doesn't turn on fully, and you may get partial current flow, or no current flow, depending on that specific part number FET's designed behavior.

I think TP6 is probably the discharge FET gate drive test pin.

These pages may be helpful:
https://www.allaboutcircuits.com/textbook/semiconductors/chpt-2/insulated-gate-field-effect-transistors-mosfet/
https://electronics.stackexchange.com/questions/157065/explain-in-laymans-terms-vgs-and-vgsth-of-mosfets
Others
https://www.google.com/search?q=gate+source+voltage


I see that the charge FET (upper right in your picture) has been resoldered. Was it replaced? If so, was it replaced with the same part number? If it was not, the gate drive circuitry on the BMS may not be able to turn it on.

If it was replaced, then why was it replaced? Depending on the reason for failure, damage to the gate drive circuitry could have occured during the original failure.




I checked the discharge FETs and they are willing to send "voltage" to the output, since:
V GND-Drain = 5V and V Drain-Source = 5V (bizarre ?).
You would want to check Gate-Source voltage if you're checking to see if the gate drive is working.

Drain-source is just how much voltage drop there is across the FET, meaning how much voltage is lost across it (power wasted as heat). 5v is quite a lot especially if there is no current flow (no load) yet. Generally it means the FET is not fully on, or is damaged, since it should have only milliohms of resistance at most, and to generate 5v across say, 10milliohms, takes quite a lot of amps. ;) (V/ohms=A)


I also noticed that when a suitable voltage (ie 42V or so) is applied to the charge socket, the discharge FETs shuts down (V GND-Drain= 0V, V GND-Source= 0V), so THAT BMS doesn't like charge and discharge at the same time (I noticed more recent chinese BMS accept that, usually).

That's probably a safety feature, so you can't accidentally ride away with the charger still plugged in (at least, not using motor power). ;)
 
You list 11 cells, but if it's a 10s pack, where are you measuring the 11th cell voltage?

It was a typo, it was listed correctly on my paper :( Fixed on 1st post.
You're right Cell0 is the row of cells closest to Ground. I definitely use the test points of the balance shunts to measure the voltages.

Do you mean "Gate" rather than "GND"? Because GND would be ground, or battery negative. Gate is the pin that is driven to a certain voltage to turn on the FET, which then connects Drain to Source so current can flow.

Of course, I meant Gate instead of GND :mrgreen: I also know (in theory) the pinout for common mosfets.

Taking your remark as for measuring correctly, I get on the charging FET: V_DS=1,1V and V_GS=OV

True, I have changed the original FET. I was unsure if it was the FET or something else so I changed the FET anyway. I don't remember the part Number, but I recall they had compatible ratings (RDSon, VGS(th), max voltage). If I can trust my tester (TC1 transistor tester) the old FET was good anyway.

As for the discharge FETs, I have V_GS=0V and V_DS=5V, so I figure the FET is off and there's not a ridiculous current crossing it :)

BMS can't charge and discharge at the same time
It might be a safety feature, but that means you can't use a solar panel to charge the battery while pedaling. Someone like you would certainly miss something like this, I'm sure !

Thanks for your help !

I have the following new info to submit, all the voltage are when the battery has its charge connector at 42V. All voltages are measured respective to the charging FET source pin

TP5=0V
TP7 = TP8 = TP9 = 1,8V
TP6 = 1,6V
TP11 = TP12 = 42V
 
Back
Top