Flying ADC 16 channel BMS 96S test

[maiks]

Thank u

[skr]

Thank you, @johu. I just want to confirm—should I remove the components marked in red (as shown in the attached image) and replace them with 10 nF or 22 nF 0603 SMD ceramic capacitors?

see pic on my "implementation", ther resistors are beneath the caps

[maiks]

see pic on my "implementation", ther resistors are beneath the caps

image.png

Thanks for the clarification. I've placed the order for the capacitors—they should arrive within 10 days. I’ll update you once they’re here

[Proton]

So..... I was doing the CAN mappings for the motor and inverter. I saved the mappings and turned off the car and turned it back on.

The screen was not showing the SOC anymore and the voltage was 309V instead of 369V.
2nd BMS was blinking slow and when I looked it is MUXSHORT...

I posted the pictures here:
viewtopic.php?p=83431#p83431

[johu]

Can you repair it, install the caps and latest firmware?

[Proton]

Can you repair it, install the caps and latest firmware?

If you can send me a screenshot with which parts to replace and a link with the parts I have to buy I can try. I tried before on 2 boards with no success.

[johu]

Well the procedure hasn't changed, you have to check the diode inside the FETs with a multimeter in diode mode.
Q1 and Q5 (BSS123) are likely culprits so check them first. Also Q52 and Q39 (ST1005SRG)

The good thing about the power on self test is that no other FETs are destroyed in a chain reaction.

It's also interesting that it started straight into the mux short. It means without even trying switch the mux something is already broken.

Now that it's not longer a "running system" it's a good time to install the latest firmware on all modules and install the caps viewtopic.php?p=83307#p83307

[Proton]

Are this temp senors going to work with the Board? "10K Ohm MF52-103 5% Accuracy High Sensitivity Current Limiter Temperature Sensor "

https://www.amazon.com/dp/B0CH1CC1S6?re ... Doh=1&th=1

[skr]

Yes they will

[Proton]

I installed the 2.4 BMS version.


Do I have to change any settings to start balancing the cells? Mine always show in Idle mode.

[PetersonOctavius]

ubalance should be around 3700 3800

[Proton]

I have another question. I connected the 12v from the BMS-ses directly to the 12v battey and the enable is stil separate and connected to the Ignition.
Now even when the car is off the lights are still on on the BMS. is that normal? should I connect them back to the ignition together with the enable?

[johu]

Yeah that is intended. In this time it can do balancing. After two hours it should power off, although there is still some bug that I will fix next (the sub modules don't power off)

[Proton]

Indeed when I came back in the morning they were off. I turned back on yesterday and again they went off after 2 hours.

[skr]

Johu, IMO it is worth it to add params for sleep time and idle current, both came in very useful in my vx1 fork and I guess others may find them useful as well

[johu]

Johu, IMO it is worth it to add params for sleep time and idle current, both came in very useful in my vx1 fork and I guess others may find them useful as well

Yes worked on that, see latest commit.

Now where do we stand?
- I run v2.1, v2.3 and v2.4 hardware on a 96S system and it has survived multiple power cycles
- Mario runs 96S with v2.2, how does that go?
- Proton runs 96S now with v2.4 modules, works?
- Octavius runs 96S with capacitor patch, no failures?
- skr runs 48S with cap patch, all good?
- maiks waits for the results of the above

Would be good to get some feedback on this!

[mario]

Now where do we stand?
- Mario runs 96S with v2.2, how does that go?

Not much happened since last report. I've received new board to replace #4, but didn't run balancing after installing it, I just did readouts for test. I disconnected boards and after that I was getting car ready for the inspection - decided to go without BMSes.
Got license plates last week and car is back in the garage so I can work on it. I will reinstall boards in the next few days, calibrate board #4 and report back.

@johu should I upload latest firmware to my boards?

[johu]

@johu should I upload latest firmware to my boards?

Yes you should and I very much recommend the "capacitor patch"

[Jacobsmess]

On the above. I bought some capacitors to perform the patch but have ended up with way too many due to an order error. If anyone wants some in the UK let me know.

[maiks]

Yes worked on that, see latest commit.

Now where do we stand?
- I run v2.1, v2.3 and v2.4 hardware on a 96S system and it has survived multiple power cycles
- Mario runs 96S with v2.2, how does that go?
- Proton runs 96S now with v2.4 modules, works?
- Octavius runs 96S with capacitor patch, no failures?
- skr runs 48S with cap patch, all good?
- maiks waits for the results of the above

Would be good to get some feedback on this!

I added capacitors to all six boards, but I haven’t tested them yet. I ran into a setback — the coupler broke, and I had to remove the front battery pack. I’m currently waiting for a better coupler from Brat Industries to arrive. I expect to be back with results in a few weeks.

[skr]

Mine are constantly power cycled, 2.3 with capacitor mod. No death or issues so far, but also seems to do nothing in terms of balancing. A week later the overall numbers get lower from 7mA drain from motor controller dcdc, but the delta, if closing, seems to be only on soc curve, not balancing efforts. I have checked my code multiple times to see if nothing is breaking the balancing commands, but can't seem to find anything wrong there. Latest addition of adaptive balancing cycles depending on delta doesn't seem to have much effect either. Have not yet opened the bike to remove the 100mA ptc in front of each cell tap (3ish ohms during normal conduction). Have not yet set up a test setup to try and experiment with balancing fet switching to bypass the current limiting resistors.

[mario]

Yes you should and I very much recommend the "capacitor patch"

Argh... I wanted to skip that.

[PetersonOctavius]

Yes worked on that, see latest commit.

Now where do we stand?
- I run v2.1, v2.3 and v2.4 hardware on a 96S system and it has survived multiple power cycles
- Mario runs 96S with v2.2, how does that go?
- Proton runs 96S now with v2.4 modules, works?
- Octavius runs 96S with capacitor patch, no failures?
- skr runs 48S with cap patch, all good?
- maiks waits for the results of the above

Would be good to get some feedback on this!

Hi there, my car was on jack stands for a while because of oil pump issues with the gs450h.
it up and running now. for bmses, i had only 4 boards left that was working hw version 2.1 i think. when i powered them on last week, board #1 went bad, now i have 3 working boards.
i havent purchase the 0603 caps to install as i didnt have the time. i received some used bms boards from Proton i still have to hook up and see what their status is.
for now i have over 10 boards "not working" (muxshort) hw version 2.1 2.2 2.3. and the 3 working board running the latest firmware before the 0603 caps firmware.
im going to check the caps prizes on mouser to order some while i diagnose the none working boards.

[Proton]

Yes worked on that, see latest commit.

Now where do we stand?
- I run v2.1, v2.3 and v2.4 hardware on a 96S system and it has survived multiple power cycles
- Mario runs 96S with v2.2, how does that go?
- Proton runs 96S now with v2.4 modules, works?
- Octavius runs 96S with capacitor patch, no failures?
- skr runs 48S with cap patch, all good?
- maiks waits for the results of the above

Would be good to get some feedback on this!

The 2.4 Version works no problems so far. still do not know how to make them balance. never seen them saying "balancing"

[skr]

You need to set param balance to both and ubal below what the cell voltage is, say 3800mV and the balancing is only carries out in state idle, which is switched to once the current from current sensor is below some treshold. Johu afaik made new params for idle state current in latest commits, if your idle current is higher than previous hardcoded treshold. If you have no current sensor you can probably force it to be idle all the time by enabling current sensor and playing with offset and adc value to show 0 or close to that.