Flying ADC 16 channel BMS 96S test

[johu]

Read here for a "repair manual": viewtopic.php?p=78587#p78587

[johu]

Checking logs, the boards worked for about 110 minutes before shutting down one by one. The same issue happened when I flashed CI #17 and CI #15.

I have fixed this now: https://github.com/jsphuebner/FlyingAdc ... 3518024923

[Proton]

Before I start desoldering and replacing parts:

I rearenged the boards in order of what they are displaying.

measurement with nothing plugge in.
On node 10,11,12,13,14 - Q5=2.3K, Q1,Q44 and Q52 = 5K and all the rest are 10K
On Node 15 few of them are different.

- Node 10 and 11 work fine and displaying cells voltages accurate. OPMODE = Idle (node10) and Run (node11)
- Node 12 opmode = Idle and displaying first 2 voltages inacurate. (Q1 and Q5?)
- Node 13 and 14 no voltages and opmode = error; lasterr = CELL_POLARITY no voltages displayed.
- Node 15 opmode = error; laserr = MUXSHORT

** withouth the battey cables plugged in all 6 red LEDs blink fast.
with the battey cables plugged in node 10,11,12 blink fast and node 13,14,15 blink slow.


- On which node should I replace Q1 and Q5 -?

[johu]

Yes, Node 12 should be repairable by replacing Q1 and Q5. Node 13 and 14 have also blown PFETs which you can detect with diode mode. Also the 10 Ohm resistors tend to go as a result. Node 15 might even have a failed ADC or something.

Lets see how many you can repair and I'll send you some free replacements for the rest. They are newer revision with more protection.

[Proton]

I replaced Q1 and Q5 on node12 but have the same results.

I ordered from mouser.com and it says in the description that isa BSS123 but that does not show on the part. not sure if that can be a problem.

[johu]

Yes small SMD parts only have some code as there's no space for a full label

[Proton]

So, if you want to tell me how to measure or what to measure, I can do it. other that that I am stuck.

[johu]

Did you read this? viewtopic.php?p=78587#p78587

Here is a quote but I think it is better if you read the roughly 8 posts down from the link

u1 if off because of the missing FET, but there also seems a broken PFET messing with the other values. When I fixed your modules previously I went over all FETs with a diode test mode. They should read 0.5V or so from Drain to Source. Any more or less means it is broken and needs replacement.

It is hard to tell from the values alone which FETs are affected because a broken one skews the other channels as well

If you order anyway go for the specified BSS123 (N) and PJM10H01PSA (P)

EDIT: some pictures:
This is a good MOSFET. Red probe on the single pin, black probe on the lower left one (or upper right for the mirrored FETs). Diode test mode

This is a shorted one.

Sometimes you will find some that do not conduct at all, they are also broken. And also measure the resistance of the 10 Ohm resistors directly behind the cell connector

[Proton]

R1,R10 = 2.3k and R2,R33 = 5K; The rest of them are 10K.

Q1 and Q5 that I just changed are showing 0v all others are showing 0.5v
I can swap them one more time, but are the R1 and R2 values correct?

[johu]

R1,R10 = 2.3k and R2,R33 = 5K; The rest of them are 10K.

No, not these. To the left of the FETs

Q1 and Q5 that I just changed are showing 0v all others are showing 0.5v

They a NFET, I was targeting the PFETs, i.e. all others. For N you need to swap probe polarity. Although 0 would mean blown in either direction...

[Proton]

I measuered again Q1 and Q5 with the negative on the middle/single pin and they both show 0.5V.

No, not these. To the left of the FETs


They a NFET, I was targeting the PFETs, i.e. all others. For N you need to swap probe polarity. Although 0 would mean blown in either direction...

Which left? what is their name. R11 --> R20 are all 10k

All Qs are 0.5V except that I had to measure Q1 and Q5 in reverse polarity to show 0.5v

[johu]

Which left? what is their name. R11 --> R20 are all 10k

Left on my pictures. But yes R11, R24, R40 (the numbering is wild). Sure they are 10k? Should be 10 Ohms without k

With all that checked you still get bad readings? Which board are you working on?

Can you check if R9 is 1k or 100 Ohm? If 1k replace with 100 Ohm

[Proton]

R33, R2 = 5K, R1 = 2.3k, R15 - R25 = 10k,
R11 - R20 = 10k Ohm, R39 = 10 Ohm.
Most of the small ones are 10 kOhm and the big ones are 10 Ohm.

my 9 year old phone does not take good pictures.
Changin it with another de-googled phone:
https://www.indiegogo.com/projects/brax ... artphone#/

[PetersonOctavius]

I installed all my 6 bms boards today, but i getting some weird reading here and there.
Is it because opmode is RunBalance?

Im on 0.20B

[johu]

No, another mux failure. viewtopic.php?p=80403#p80403
Or better read this entire thread to understand what is going on

[PetersonOctavius]

sometimes i get good readings, other time its all over the place, then back to good readings.
im going to check the thread

[PetersonOctavius]

on the web interface, i get good readings of individual cells voltages. its the cellmax cellmin cellavg that's wrong here and there.
when i connected the bms boards alone, i got good reading on all of themx

[johu]

Oh, ok. Well it could be worthwhile upgrading to the latest software then as I improved I2C stability.

[PetersonOctavius]

I ll try that

[PetersonOctavius]

updated 4 boards to 0.23B, board number 2 on the chain was flashing slow, all cells were show 0 on spot value tab.
went to parameter, enable was off....aahh turned it on...i heard like popping on the bms board. those weird values appeared on the web

went ahead and updated the last 2 boards. the showed 0 also, i turned enable on on board number 6 in the string, same popping sounds and the same weird values.

i did unplugged all cells from the boards before updating.

[johu]

Well, not happy to hear this.
Today I tested this version on my 32S LFP test pack and v2.0 hardware where it worked flawlessly (on the desk I just test with a 16-cell "simulator" that also indicates mux issues with increased current draw). The ones that blew up were also the ones with wrong min/max? Which components did blow up? Which error did it detect prior to popping?

Also when the board reports an error you better be careful, especially on "MUXSHORT".

[PetersonOctavius]

hi Johu, sorry for late reply. got caught up in life. my wife and i were expecting a baby boy by beginning of may, he was born march 28th due to health issue. but baby and wife are doing well now, so i got time to work on the car.

2 of my modules are reporting MUXSHORT :'(
prior to popping i didnt see any errors.

[PetersonOctavius]

the batteries connected to the 2 BMS boards that had a MUXSHORT went way of balance, even without power to the boards.

[johu]

Sorry, forgot that. Congratulations first

Yes MUXSHORT means random transistors are conducting and may drain cells. I would expect that the board must have also been quite warm.

[johu]

Got another private message about a broken BMS, or 3 in fact.
I'm trying to figure out what is going wrong.

The BMSes are tested with 3.3V per cell here. This test could be part of the problem as it only puts 52V across the FETs. When connected to 16 NMC cells we are looking at 67V. As found here chinese FETs don't always seem to fulfill their data sheet values.

Another possible culprit is the DC/DC converter. It has very little guard distance between HV and LV side. So I wonder if transients could make it over there and mess things up. That is for cascaded arrangements only (which is where all failures happen).

To minimize transients I have now revised the sweep pattern: https://github.com/jsphuebner/FlyingAdc ... 4352787110
Also a version with a different DC/DC converter is coming.

Instead of 0, 1,...15, 0 it now goes 0, 2, 4,...14, 15, 13, 11,..., 1, 0

I hope we can get this stupid bug ironed out as the BMS works super well otherwise