openinverter forum

Hi,

I made a simple LED array to test BMS Spaghetti wiring:
- if your connections are correct: all LEDs are on
- if you skipped a cell: you burn an LED out
- if you reversed a connection: the LED is dark

I made them in loops of 4, so you don't have a continuous connection of 380V, but sections of ~15V.
LEDs are THT, so easy to resolder if popped.

You can see in the photo below some LEDs are dark where I removed the bus bars while working on the pack.

Thoughts?

Very nice! Is that on a prototype board or an actual PCB design? I would have a use case for this.

viewtopic.php?p=33537#p33537

I have some PCBs.

I can send you one of the boards to self-populate if you want.

NICE!

janosch wrote: ↑Wed Nov 24, 2021 11:57 am Hi,

I made a simple LED array to test BMS Spaghetti wiring:
- if your connections are correct: all LEDs are on
- if you skipped a cell: you burn an LED out
- if you reversed a connection: the LED is dark

I made them in loops of 4, so you don't have a continuous connection of 380V, but sections of ~15V.
LEDs are THT, so easy to resolder if popped.

You can see in the photo below some LEDs are dark where I removed the bus bars while working on the pack.

Thoughts?

IMG-20211124-WA0003.jpg
test_board_4_led_loop.png

Great idea and implementation!
Is this for GEN1 Leaf BMS or GEN2?

Ruudi S wrote: ↑Sun Nov 28, 2021 12:49 pm Is this for GEN1 Leaf BMS or GEN2?

They are all the same in that respect.
Please don't quote entire posts

Yes, all the same like johu said, as a matter of fact the photos above are from a 24kWh pack ... and the ones below are from an otherwise untouched 40kWh one.

You can see one LED is off, thats where I made a wiring mistake on the board and connected both sides to the same cell. Everything else works perfect, on a full pack everything comes on.

I am just sending two boards to a school in Illinois today, if anyone want the last one, send me a PM.

I want this board!

Are you willing the share the design files??

Ruudi S wrote: ↑Sun Dec 12, 2021 4:13 pm I want this board!

Bratitude wrote: ↑Sun Dec 12, 2021 6:14 pm Are you willing the share the design files??

Hello,

if you want to self print, the gerber files are attached below.

- D40 will not come on, that is a wiring error I made
- D51 is too close to the aluminium case, you need to "correct" the case there to make sure it doesn't touch!

Post here & let me know what your experience is! I hope it helps.

Thanks for the files but I don't know how to use them. Do you have boards available?

janosch wrote: ↑Mon Dec 13, 2021 12:35 pm - D40 will not come on, that is a wiring error I made
- D51 is too close to the aluminium case, you need to "correct" the case there to make sure it doesn't touch!

The board arrived now, thanks very much!
How can I fix D40 ?

@Ruudi: you can upload that zip-file at jlcpcb or some other service

Populated my board now but that immediately brought up a question: how can you check 96 cell voltages with just 78 LEDs? Indeed some inputs do not light any LED

johu wrote: ↑Mon Dec 27, 2021 1:49 pm Populated my board now but that immediately brought up a question: how can you check 96 cell voltages with just 78 LEDs? Indeed some inputs do not light any LED

It is quite possible I made more mistakes, but see sketch below, there is some fenceposting going on with the loops of four. I am making sure I test every wire once, and not every individual cell. https://en.wikipedia.org/wiki/Off-by-on ... post_error

swapping individual wires
Simplified schematic with 8 cells, needs only 6 LEDs to check every sensing wire. You do not need to check between (4,5), because if you did swap the wires, the LED now connected to (3,5) should see 8V and pop.

swapping groups of wires
If you swap a whole group around (1,2,3,4) with (5,6,7,8), it would look as if everything was OK, while it is in fact not, but I assumed that was quite unlikely. The swap would have to land exactly on the boundary of loops on the board to be unnoticeable.

Simplified Sketch: Full Schematic:

Alright, understood very clever. I'll actually be looking for "weak" wires, i.e. that measure normal on the multimeter but might collapse when loaded

All built and sort of tested. Now what was the issue with D40? It lights up but you said there is a wiring error

johu wrote: ↑Tue Dec 28, 2021 10:18 am All built and sort of tested. Now what was the issue with D40? It lights up but you said there is a wiring error

The last pin on LB14 is 48-
The first pin on LB15 is 49+

they are the same.

So if you connect it to a real battery pack, it will just remain dark because both sides of the diode are connected to the same cell. Not sure why Nissan decided to double them up at that one place.

There isn't a "fix" I think, just ignore it.

Looking neat, let me know how you progress. I like the SMD LEDs in the THT spaces.

Do you happen to have one for a Chevy Volt battery?

janosch wrote: ↑Tue Dec 28, 2021 11:45 am The last pin on LB14 is 48-
The first pin on LB15 is 49+

they are the same.

So if you connect it to a real battery pack, it will just remain dark because both sides of the diode are connected to the same cell. Not sure why Nissan decided to double them up at that one place.

Nissan did this because their maintenance switch is located between cells 48 and 49 in the middle of the pack, so they have split the BMS board similarly (you can see some optoisolators on the board for this). Otherwise when the maintenance switch was disconnected current would pass through the BMS - not favorable for longevity.

Isaac96 wrote: ↑Tue Feb 08, 2022 9:15 pm

janosch wrote: ↑Tue Dec 28, 2021 11:45 am ....

Nissan did this because their maintenance switch is located between cells 48 and 49 in the middle of the pack, so they have split the BMS board similarly (you can see some optoisolators on the board for this). Otherwise when the maintenance switch was disconnected current would pass through the BMS - not favorable for longevity.

That makes perfect sense. I saw the division there before.

Unrelated:
The board does not always show you a single loose wire. In the Simplified Sketch above, if you disconnect the wires to cell 2 and 3, you will still get all three LEDs light up there, because they are just in series. Brightness will be identical as well unfortunately.

Learnt that just now by testing a new pack, that had a loose wire, then plugged the test board in to verify and it still lit up all LEDs. So it is good for detecting swapped wires, not necessarily for individual missing wires.

I made a spreadsheet to show the voltages at each connector. I'm not great with spreadsheets so it is pretty rough but if you enter the voltage of your cells it will show the voltage at each pin, assuming all cells are at the same voltage.

Please feel free to check, improve and share.

When testing I screwed the neg of my volt meter to the pack neg and taped a small needle to the pos probe so I could poke it into each pin of the connector in turn and check the voltage from the spreadsheet.

@janosch very nice work! Good idea

Thanks Pablo, I hope it helps.

Just saw that the board even made its way into Johannes recent video about the A2 installation.
Brave man to plug it in with naked hands, I always put the gloves on at that stage!
https://youtu.be/8eKkwPbLwDQ?t=408

That is unacceptable practise by me. As a wise man said: "don't do as I do, do as I say". I say: wearing gloves in that situation is a good idea