Yixiang DIY Battery Case- Voltage between cell terminals and chassis

[hgg]

Hi,

I am building a DIY Yixiang battery box with 16 MB31 cells and the JK 16S 200A BMS.
The issue is that I can measure voltage from the chassis to each cell. The voltage
is not stable, its like a phantom voltage and when I measure it with low-Z, it shows
a much smaller value. I've turned it on and it works, but I would not like to use the
battery case like that.

It seems like there is a conductive path from the JK BMS to the chassis.
Did anybody else had a similar issue?

Thanks!

 

[Mav986]

You will need a fire extinguisher

 

[hgg]

Can you be more specific?

You can measure with a multimeter voltage from the chassis to the +terminal for example
and the voltage will slowly go down until around 0.5V Same from chassis to individual cells.
I have already checked the two bars with the balance leads and they are no shorts. It looks
that its from the JK BMS.

 

[Mav986]

Looks like the main cables are way undersized + loose pins in BMS connectors, balance PCB screws shorting to case etc... May want to check every single connection and wire on this thing and make sure the PCB mounting screws don't short out to the case.

 

[hgg]

I will have a look, thanks.

 

[hgg]

I have none of the issues mentioned in the video, thanks anyway.

In my case there must be some kind of conductive path between the JK BMS and the chassis.
I think I need to remove the BMS and check again.

I was wondering if anybody else had a similar issue.

 

[mikeenea]

I have none of the issues mentioned in the video, thanks anyway.

In my case there must be some kind of conductive path between the JK BMS and the chassis.
I think I need to remove the BMS and check again.

I was wondering if anybody else had a similar issue.

@hgg I have built 2 of the V1 boxes so far and have not experienced this feature I have a few more boxes to build and will do a lot more testing on those when I do the builds -- thanks for bring the issue up!

 

[hgg]

@mikeenea I have uploaded a small clip showing the issue and how I measure the phantom voltages.

I have disconnected everything from the case and I am using only a single 8s cell string.
Have you seen anything like that? Have you checked your cases in this way?
This might be dangerous because you can accidentally short for example the front
positive terminal with its screws that are mounted on the chassis. Although this
voltage seems like a capacitive voltage or from a high impedance path.



When I measure the JK BMS on the bench from B- to a screw terminal, I get from 11 to 60 MΩ.
Is that normal?

 

[mikeenea]

@hgg - I have not done that kind of testing but will be assembling another V1 box in about a week and will run through your test scenario.

In the meantime, take a look at this video from Off Grid Garage -- I think it may be relevant even though it is a different box it is a JK Inverter BMS. I just went to my battery bank and tried the test he ran just after 11:30 and did not measure any voltage or continuity on the battery boxes.

Mike E.

 

[hgg]

That was the video I was looking for because I remembered that I had seen it but I couldn't find it. Thanks.
Andy had a dead short from B- to the BMS face plate. The coil in my BMS has the heatshrink, but it was
touching the metal hole. I pushed it aside leaving a tiny space from the metal hole, but I can still measure
from 10MΩ between the B- and the screw hole. Resistance goes up slowly to 100MΩ during the measurement.
With the working BMS he replaced, he still had some phantom voltages from chassis to cell positive.

Maybe this is normal and these voltages are capacitive.
I would appreciate it if you can measure with a multimeter the resistance from B- to the BMS screw hole.
Assuming that you have one lying around.

(or anybody else by the way) Thanks!

 

[mikeenea]

OK -- I took my spare BMS (JK-PB2A16S20P) and set my multimeter to Ohms and measured between the B- screws and a chassis mounting screw.

The meter started counting up the milli ohms and then went to OL which is the meters normally open state. I tried connecting and measuring multiple more times and still OL. It is almost like it pre-charged the BMS from the meter - I am clueless now. There are no shorts between B- and chassis screws at this time.

Mike E.

 

[hgg]

That's the same BMS I have.
Strange indeed. milli ohms? Do you mean mega ohms?

 

[mikeenea]

I reread the multimeter manual, and it may have been m farads as I tested again it started reading u farads can counting down. Here is a short video

 

[hgg]

You are measuring capacitance, μF.
What about resistance?
Can you switch your multimeter manually to Resistance mode?

 

[mikeenea]

@hgg Yes - I can set it manually and measure resistance and/or continuity, and I did that it read "OL" - open, and then no beep for continuity when I changed to that mode

Mike E.

 

[hgg]

Ok, continuity will only show a dead short, not a very high impedance path.
I do not know the model of your multimeter, but it needs to be able to measure resistance of above 10MOmhs.

 

[hgg]

Anyway, next week when you build your next battery box, maybe you can take some time
to measure the cell voltages from the chassis with the JK BMS screwed on the case and let
us know. Thanks!

 

[hgg]

By the way, the resistance of the two Yixiang v1.0 balance bars is horrible...
From 5mΩ up to 117mΩ ! This means that some cells will always take
much longer to balance.


Can this be rectified from the JK BMS cable resistance calibration menu?
Does anybody know what calibration values do you use in that menu?

Thanks.

 

[Mav986]

From the front connector pin to the end of the balance wire? Then it is likely the solder of the lead (s) on the PCB is bad. To me this is unacceptable and I would request a replacement unit.

 

[hgg]

Yes and I have resoldered all the connections with lead solder.
Do you have a similar balance bar that you have measured?

If the JK BMS resistance calibration can take into account these values, it is not a problem,
but I do not know how to use that menu. Do I enter the measured values?

 

[Mav986]

No I have a Basen Green, Flawless so far. Top balanced to < 0.005V,

 

[hgg]

The Basen Green also uses balance bar and might have similar issues.
The cells will be balanced. The problem is that they will take a longer time,
if that imbalance is not taken into account.

It definitely has a better DC breaker though than the Yixiang toy switch...

 

[740GLE]

Thread 'Well this happened! What is happening? Help!'

Jan 25, 2025

This was the second time starting this thing. And this happened

• Nef003
• Replies: 182
• Forum: Up in smoke... learn from my mistake!

• 
• 

Another users issue with terminals shorting.

 

[hgg]

Thankfully I do not have anything like that... A reminder that you do not play with LiFePo4...

In my case I can measure voltage from chassis to cells, but when I keep the probe connected,
the voltage starts dropping until it goes to less than a volt. So there must be something capacitive in play.
I hope Mikeenea tests his new battery box the same way and tell us if he has similar measurements.

I connected a 10Ω resistor between the chassis and cell 16 while measuring current with a
clamp meter and there was no current at all.

I've also just replaced all busbars with flexible ones and already I saw a great reduction in the resistance
values.

 

[Mav986]

Looks good