Poll: my BMS testing regime is CURRENTLY:
You do not have permission to vote in this poll.
I test my BMS's and you are crazy -- they all work.
33.33%
2 33.33%
I test my BMS's and get the same result -- most are duds.
16.67%
1 16.67%
I dont test my BMS's. (living life on the edge!)
16.67%
1 16.67%
The BMS I use is better than this, so there.
16.67%
1 16.67%
I refuse to answer on the grounds it may incriminate me.
16.67%
1 16.67%
Total 6 vote(s) 100%
* You voted for this item. [Show Results]

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
[SOLVED] all my cheap BMS's failed testing (DOA?)
#1
Hello All.

I sat on this for a good week ... trying to think how to word this post, what pictures to take, and ... reflecting on whether I had just 'missed many clear warnings from fellow members' ... finally, today I did an SLS search on various search terms, EG failed BMS from china, cheap BMS, testing BMS, BMS failures, ... and some more.  pretty much nothing came up.

So it is with a heavy heart that I deliver this post to you all.  Not so much that I got 'ripped off' a few measly $$, but more that I had no idea that tech had gotten this trashy ... hadn't been WARNING other members ... but also ... that no-one here had warned me ... about these cheap BMS's.   So here goes ... *deep breath*


Background: recently I have decided to get a whole heap of cells into production.  So I ordered a variety of the 'standard' cheap BMS's we have all seen ...  before assuming that my $$/precious batteries would be safe, I thought it prudent to test them ... so I built a jig, and here's what happened:



sorry for bad images, new camera is literally on its way to me now ...

so the setup (above) is pretty basic.  the two purple cells wired in 2s go through a linear regulator (LM317 based) to simulate the first cell in a 6s string.  that way I can dial up/down the 'cell' voltage, and see what happens to the output of the tested BMS.

as you can see, cell #1 is well under 2.8V and the output is still ... Vbatt (!!).

my testing has revealed that it makes no difference to vast majority of BMS's even if the balance lead/harness is completely disconnected (!!!) which to me would seem to be the worst type of battery failure, as then there is just no info to 'MS' the 'B'.

I did pause and think before posting this.  it did strike me as unlikely, but this is not a dream ... it's true!

NB votes CAN BE CHANGED, so put your current situation now, and if your situation changes, then you can come back and update your vote.  the poll result can be useful for us all, please participate!  cheers all  Smile  well, I goofed, evidently not all polls can have votes changed.  perhaps its cause its a public poll, or no end date ... dont know ... my bad.
--    ----    ----    ----    ----    ----    ----    ----    ----    ----    ----    ----    --
I am an iSDT fanbois.  also a proud owner of two SC608AC's .. totally recommended.
-----------------------------------------------------------------------------------
Reply
#2
I don't see how using the LM317 board is helping here. It's not bidirectional, is it? Or are you only testing the "discharge" side of things?

Also, double check your capacitors on the LM317. In your second image the one on the right side looks like it might be a little puffed. This could also through off readings if it is indeed failing.

Could you provide a circuit diagram of your current wiring?

How did you determine your BMS's all failed? Was it only based on the overall voltage not being what you expected?

And if indeed they are all dead, man that sux! I've been down the road, but with the STM32 "Blue Pill" boards I got. All 5 were bad mcu's. The actual IC was a factory reject.
completelycharged likes this post
Proceed with caution. Knowledge is Power! Literally! Cool 
Knowledge is Power; Absolute Knowledge is Absolutely Shocking!
Certified 18650 Cell Reclamation Technician

Please come join in general chit-chat and randomness at Discord Chat (channels: general, 3d-printing, linux&coding, 18650, humor, ...)
(this chat is not directly affiliated with SecondLifeStorage; VALID email req'd)
Reply
#3
Did you put a load on the bms? I think that is the only way you can get it to trigger(stop passing current). The bms even when it triggers, you will still be able to read voltage (due to the electronic relays it uses will leak voltage) but you won''t pass current. 

From what I see in your experiment, the bms seems to be working ok, if the voltmeter read "0" then you might have problems. 

When I test my bms, I always test with a load either charging or discharging, and I monitor the cell voltage when it triggers to stop charging/discharging. I also use the cheap 4 dollar bms, I rarely found a bad one. But some of the more expensive ones I tried in the 10 dollar range actually was trying to overcharge the battery. You have to test everyone and maybe have a voltage alarm just in case.
completelycharged likes this post
Reply
#4
on the ball as always, thanks for the speedy response Korishan  : )

(05-06-2019, 12:45 AM)Korishan Wrote: I don't see how using the LM317 board is helping here. It's not bidirectional, is it? Or are you only testing the "discharge" side of things?

yes, well spotted, I just threw this together to test (only) the discharge functionality of BMS.

had better plans for a v2 of this test jig, but have stopped testing as results so very bad, so far!

Quote:Also, double check your capacitors on the LM317. In your second image the one on the right side looks like it might be a little puffed. This could also through off readings if it is indeed failing.

true dat, BUT this is why the BG-8S (and 7 segment voltmeter) is there also.  I know the photos arent the best, but the BG shows clearly that cell #1 is low ok... nice try though ...

Quote:Could you provide a circuit diagram of your current wiring?

I could, but my diagram making skills are.. pretty lame.  I used colored wires neatly laid out hoping it was easy to follow.  FYI 'Cell' 1 at the bottom, #6 at the top, -ve's all on the right ... output goes to multimeter ... all cell interconnects go through 1A blade type fuses, then parallel outputed to a balance tester (the BG-8S) and the clips to the test BMS.

Quote:How did you determine your BMS's all failed? Was it only based on the overall voltage not being what you expected?

based (entirely) on the fact that the BMS specs said low cutoff at 2.8V or whatever ... and I cant get it to cutoff.  even when all clips removed (other than B+ and B-).  I was expecting ... 0V on output if any cell got low.  This is sane, right? (very sincere question, as Im completely stumped!).

Quote:And if indeed they are all dead, man that sux!  I've been down the road, but with the STM32 "Blue Pill" boards I got. All 5 were bad mcu's. The actual IC was a factory reject.

yeah, it sucks for me, but only in the sense I now have to re-source BMS's and wait (again) for supplies to arrive.

I am however, greatly concerned that many others wont have tested their BMS's, and ... : /  Remember, wasnt just the ones I got, also reproduced same reults when I borrowed different BMS's from others for testing.  I could only find one so far that turned output off ok.  ONE.


that'll teach me for not buying tested/quality stuff from a fellow member.  this is about as positive as I can be about this whole situation.

(05-06-2019, 12:54 AM)jonyjoe505 Wrote: Did you put a load on the bms? I think that is the only way you can get it to trigger(stop passing current). The bms even when it triggers, you will still be able to read voltage (due to the electronic relays it uses will leak voltage) but you won''t pass current. 

From what I see in your experiment, the bms seems to be working ok, if the voltmeter read "0" then you might have problems. 

When I test my bms, I always test with a load either charging or discharging,

ah, yes, I had thought of this, but not tested it, as ... well, ... that just seemed insane to me.

I was sure that MOSFET's didnt have such leakage ... but that said ... I should have tried ... will do that right now ...

EGAD, you were right jonyjoe505 ....!!!!!!!! it does cutoff ok with the 24V fan I just tried. no good on the release though ... but thats another story. problem solved I guess. THANKYOU!

bah humbug ...
completelycharged likes this post
--    ----    ----    ----    ----    ----    ----    ----    ----    ----    ----    ----    --
I am an iSDT fanbois.  also a proud owner of two SC608AC's .. totally recommended.
-----------------------------------------------------------------------------------
Reply
#5
(05-06-2019, 01:01 AM)DCkiwi Wrote: EGAD, you were right jonyjoe505 ....!!!!!!!! it does cutoff ok with the 24V fan I just tried. no good on the release though ... but thats another story. problem solved I guess. THANKYOU!

bah humbug ...

Big Grin Glad you got it figured!
DCkiwi likes this post
Proceed with caution. Knowledge is Power! Literally! Cool 
Knowledge is Power; Absolute Knowledge is Absolutely Shocking!
Certified 18650 Cell Reclamation Technician

Please come join in general chit-chat and randomness at Discord Chat (channels: general, 3d-printing, linux&coding, 18650, humor, ...)
(this chat is not directly affiliated with SecondLifeStorage; VALID email req'd)
Reply
#6
In fact, i have had made BMS tests where i replaced one Li cell against 2 or 3 Nimh Cells. 2 for undervoltage Tests, 3 for overvoltage Tests. And small real load, as described.
I found a defect in a B6AC, but all BMS worked.
1 kWp in Test
4 kWh battery target - plus Mobile Home battery
Ultra low cost
Electronics ? No clue. Am machinery engineer.
Reply
#7
(06-04-2019, 11:57 AM)Cherry67 Wrote: In fact, i have had made BMS tests where i replaced one Li cell against 2 or 3 Nimh Cells. 2 for undervoltage Tests, 3 for overvoltage Tests. And small real load, as described.
I found a defect in a B6AC, but all BMS worked.
Replying to myself, as somebody asked for better explanation.

Let me I talk about a 4s BMS having 4 cells connected in series.

I wanted to make BMS-Test under real conditions, means:

- an ongoing current drops one cell below 2.7 V (or whatever the Limit of the BMS is)
- an ongoing charge raises one cell over 4.2 V (or whatever the limit of the BMS is)

- i want  to drop or raise the cell Voltage with an imminent imbalance of the cells and having charge/load on the whole Chain

But, Dropping/raising a LiIon cell below/ over the Limits is factual dangerous, so how i do this safely ?

First, lower Limits, dropping the cell below 2 2.7V:

I replace one cell on your choice of the 4 s chain against a 2 cell Chain of NiMh. This two cells have fully loaded 3 Volts, and on discharge Voltage can safely belowered well down to 2 Volts.
(all LI cells should be and stay within normal limits)

And the behaviour of the BMS observed.

Second, upper limits, raising voltage over 4.2 V: (should be obvious now)


I replace one cell on your choice of the 4 s chain against a 3 cell Chain of NiMh. This three cells have fully loaded 4.5 Volts, and on charge Voltage can safely raised from lower Voltages upt to that 4.5 Volts.

And the behaviour of the BMS observed.

If somebody is really painfully accurate, he repeats the process with the NiMhs at all positions of the 4s (or any other) chain of LiIon one after the other.

(It should be clear that the NiMh chains are preoladed to the desired Voltage for testing procedure before being placed in the LI Chain).

All clear now ? Big Grin
DCkiwi likes this post
1 kWp in Test
4 kWh battery target - plus Mobile Home battery
Ultra low cost
Electronics ? No clue. Am machinery engineer.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)