Yuntong 48V 20 ah BMS Failure??

Joined
Apr 1, 2013
Messages
35
Location
Colorado Springs, CO
I just bought a Jamis Dakar bike with an older Crytallyte hub motor (1st gen originally purchased by previous owner around ~2008). It has a Yuntong LIFEPO4 battery (yuntong-batt.com) that is no longer good. Seller told me that he thinks the cells are still good but that probably the BMS went bad.

I'm new to e-bikes so I have a few questions.

Is there a way to test the BMS to confirm that this is the problem and possible try to save it?...or is this more work than its worth?
Is it possible to swap out the healthy cells from this battery and put it into a "shell" of another battery that previously had bad cells in it?

trying to figure out which route to go...

Kevin

IMGP5904.jpg


IMGP5903.jpg
 
set your camera to VGA and that will allow the picture to be inserted into your thread.

what did the BMS do that makes you think it has failed? is it turned off and won't charge? won't discharge?

can you put it on the charger and measure the cell voltages while the BMS is balancing the pack?
 
dnmun said:
set your camera to VGA and that will allow the picture to be inserted into your thread.

what did the BMS do that makes you think it has failed? is it turned off and won't charge? won't discharge?

can you put it on the charger and measure the cell voltages while the BMS is balancing the pack?

The battery was working fine and all of a sudden it wouldn't work anymore. It is from 2008 so its getting old.
 
coloradoebiker said:
dnmun said:
set your camera to VGA and that will allow the picture to be inserted into your thread.

what did the BMS do that makes you think it has failed? is it turned off and won't charge? won't discharge?

can you put it on the charger and measure the cell voltages while the BMS is balancing the pack?

The battery was working fine and all of a sudden it wouldn't work anymore. It is from 2008 so its getting old.

set your camera to VGA and that will allow the picture to be inserted into your thread.

is it turned off and won't charge? won't discharge?

can you put it on the charger and measure the cell voltages while the BMS is balancing the pack?
 
dnmun said:
coloradoebiker said:
dnmun said:
set your camera to VGA and that will allow the picture to be inserted into your thread.

what did the BMS do that makes you think it has failed? is it turned off and won't charge? won't discharge?

can you put it on the charger and measure the cell voltages while the BMS is balancing the pack?

The battery was working fine and all of a sudden it wouldn't work anymore. It is from 2008 so its getting old.

set your camera to VGA and that will allow the picture to be inserted into your thread.

is it turned off and won't charge? won't discharge?

can you put it on the charger and measure the cell voltages while the BMS is balancing the pack?

FYI, I have updated my first post with high quality pictures of the BMS. Do I have to take the whole battery out of the case before I can test the cells with the voltmeter or can I do it without removing?
 
You can do it without remove if there are not cells in parallel.
Before diagnose the BMS, check voltage of each cell. Just like the way 999zip99 described and post. Make sure cells are ok in advance. If all of them are ok then problem may probably is the BMS. Then you should focus on this BMS.
BTW, what's the reason that your suppllier will indicate this should be BMS problem? This will also help us to know more about your case.
 
you can measure the cell voltage where the sense wire plugs into the BMS. record the voltage of all 13 cells and post up here. once we know more then we can walk you through how you can test the mosfets to see if they are turned off by the controller or if they are bad. and it is possible that the processor is not getting voltage to run because there is a small fuse on some of those circuits that deliver power to the O2 micro processor which can also be repaired.

but if your battery cells are all still ok then it should be repairable. if the BMS is bad you can get another one too.

i see there is goop on the sense wires gluing them to the BMS so try measuring the cell voltage on the top end of those little resistors that say 470.

put your black probe, 20V DC scale, where it says B- on that large solder pad, put the red probe on the first resistor in series on the right side towards the edge and you should see the cell voltage on your meter. that is the voltage across cell #1. do that for all 13. measure each adjacent resistor on the right end. i think that is where it is connected to the cell.

set your voltmeter to the 200V DC scale and measure the voltage on that red wire next to the sense wire plug. put the black probe on the B- spot. that is ground.
 
if one throws away the odd reading from R50, they look good enough, not extremely well balanced, but will wait for the real experts to chime in.

although 4v/cell seems high for lith iron ph cells; sure of meter and sure of chemistry?


d
 
i think the last reading is just the same as the other one before because the spot on the resistor changed in the circuit layout and it would be another 4v higher on the other end i bet. if it is not then that could be the cause of the problem. dead cell or bad connection.

but if that measurement is the same on both ends i think it is possible to hack that BMS down to 12S by connecting the traces together with #12 from #13.

what is the voltage on the red wire too? i think that is where the microprocessor gets juice to run. you should see if you can follow the trace on the pcb for that red wire over to the processor too. it should go to one corner of the chip, and maybe you can find the ground also which i think is in the lower left corner with several pins connected to the ground. i think you can see the trace from the B- spot on the backside over to those through holes next to where it says IC1. so you need the voltage between the red wire and that spot where that wide trace goes to the processor at IC1.
 
Ok, so the only thing I think I understood from this last post is your instruction on the red wire. The thin red wire has a voltage of 47.6. As far as the other items, I'm not understanding these:

1. Reading on last resistor (R50). How do I test to make sure this is not where the problem is? You referred to the circuit layout and the "other end" but I'm not following what you mean by this...?

2. You had mentioned that I should see the trace from the B- spot to through holes by "IC1" and that I should meassure voltage between red wire and this spot. What do you mean by this. What exactly does this mean?

For the time being, I don't have access to the awesome camera I did before but here are multiple pictures of the backside of the board:

IMG_1526.jpg


IMG_1527.jpg


IMG_1528.jpg


IMG_1529.jpg


IMG_1530.jpg


IMG_1531.jpg


IMG_1532.jpg


IMG_1533.jpg


IMG_1534.jpg


IMG_1535.jpg


IMG_1536.jpg


IMG_1537.jpg


IMG_1538.jpg


IMG_1539.jpg
 
that 47.6V on the red wire is what drives the microprocessor. the wide flat ground plane with those 5 through holes in a T crossing is how the processor makes the ground connection i think and it seems to connect in the corner where it say IC1.

your last picture shows the shunt transistors on the underside that the microprocessor turns on to divert the charging current around the cell as the cell fills up with charge. you can see the shunt resistors on the front side and the transistor is connected to it by a through hole in the pcb.

the processor monitors the voltage through those traces on the front side that connect the traces running from the sense wire plug up to the top of the shunt resistor, the ones that say 470, 47 ohms.

the R50 is the top shunt resistor and you should find the final pack voltage of 47.6V, just like the red wire since the current for the processor comes from the top of the top cell, except it is not carried on the sense wire because that would affect the voltage measurement the processor makes of each cell on the sense wire.

so there is the possibility that the BMS does not work because that connection from the red wire to the processor is open. on the BMSs i have seen with the O2 micro processor there is s fuse in the circuit from that red wire to the processor and it has a TVS, transient voltage suppressor diode, in line to blow open the fuse if the battery is connected in reverse to the BMS. so that was why i said you should see if you can follow the trace from that red wire to the processor and measure to see if that 47.6V gets to the processor. i cannot see a TVS and fuse there however.

the mosfets over on the side are in two parts. the two on the bottom are for the charging and the three at the top are for discharging.

a mosfet has three legs. when you look at it from the front side, the legs from left to right are gate, drain, source. your mosfets are upside down so when looking at that picture with the mosfets on the left side of the board, the top leg is the gate, the middle leg is the drain and bottom leg is the source.

you can test those mosfets with your diode tester on the voltmeter. there is a 'body diode' between the source leg and the drain because of the way the semiconductor is manufactured. if the mosfet is damaged then that body diode is disrupted and you can tell that by measuring the body diode with the diode tester of your voltmeter.

you will have to disconnect the red wire from the BMS to evaluate the mosfets. if there is voltage on the mosfets while you test them it makes the meter very unhappy and it can die.

set the dial to the diode on the meter. put the red probe on the source leg, the bottom leg in your picture, and put the black probe on the drain. you should see a reading around 400-550mV for the forward bias of the body diode. reverse the probes and put the black probe on the source leg and the red probe on the drain leg in the middle and you should see open circuit for the reverse bias of the body diode. the meter will read off scale, open circuit, if they are functional. since the mosfets are tied in parallel you only have to test one in each set to evaluate them since they all are tied together through the legs.

but we need to determine if the last cell #13 is actually making voltage too. if each of the cells is 4v and there are 12 of them then that gets you the 47.6V and if there were 13 with 4V each then the voltage would be about 52V but you did not measure that so measure on each side of R50 to see if you can find 52V. if not then i suspect your problem is a shorted 13th channel and that is why we only see 47.6V on the red wire. if that is the case then the BMS would shut off for LVC on channel #13.
 
dnmun, in regards to the red wire, I'm not sure if I did correctly, but I put black probe on B- and red probe on one of the four points of the microprocessor on the underside of the board (by r43). This measurement showed the same 47.6 voltage, so I take this to mean that the 47.6V is in fact making itto the processor.

As far as R50, I could not get any other reading than the 47.6. This resistor is the same as all the others, in that I was only able to get a voltage reading from the one side (I believe it was the right hand side or edge of the board).

I will test the mosfets tonight and report back.
 
i am now thinking that cell #13 is dead shorted. that would cause the BMS to turn off the output mosfets. so you can skip testing the mosfets to see if they work, which is just normal procedure for me.

but i think you need to expose cell#13 and measure it directly. hacking the BMS to 12S is a solution if it is dead. but i have to figure that out yet.
 
Do I have to remove the battery from the case in order to take a voltage reading directly from the cell number 13? Where is the cell # 13 likely to be located in the battery pack?--on top, bottom, or...?
 
if you can measure the voltages without opening the case then do that but if the red wire is 47.6V and the top of #12 is the same voltage it implies there is 0V on #13.

you count the cell number from the bottom of the battery.

you can post up pictures of the battery and case so we can see how it is built. i have never heard of these guys before.
 
Not sure that I did the early step of tracing the red wire to the processor correctly. I went back tried to measure again by flipping over the board and placing red probe on on of the four solder points of the processor and black probe on B- and received no reading. Also, I was told that maybe this battery does not have the LIFEPO4 chemistry. Since this has 13 cells instead of the typical 16 it take a LIFEPO4 battery to make a total of 48 volts, is this true? Can you confirm?

For now, I am unable to remove battery as it is glued into the case pretty good. I can get a voltage reading on the top few cells as per the picture below. Basically, on the end of the battery pack, there is a series of wires attached in rows in what appears to be silver/aluminum foil style leads (see below for the first couple reads I was able to get):
 

Attachments

  • battery pack.jpg
    battery pack.jpg
    19 KB · Views: 2,421
Image 10/11

Big white connector on the left, yank it out it's a 12s balance tap.
Use it to accurately measure the cells without needing access to the cells tabs.
It's stock method on a YT BMS, sometimes they wire in an additions balance tap for diagnostics, deps on the pack builder
 
yep, those tabs are the ends of the electrodes that make up the cell. from looking at your BMS there are 13 channels but you only seem to go up to 48V on channel #12 and then it measure 48V on channel #13 so i assume #13 is dead.

but there may not be a #13, i cannot see your battery.
 
t3sla said:
Image 10/11

Big white connector on the left, yank it out it's a 12s balance tap.
Use it to accurately measure the cells without needing access to the cells tabs.
It's stock method on a YT BMS, sometimes they wire in an additions balance tap for diagnostics, deps on the pack builder

Yank it out? How will this help measure the voltage? Is it the smaller connector as shown below?
 

Attachments

  • white_connector.jpg
    white_connector.jpg
    16.6 KB · Views: 2,406
hmmm, tapatalk app didn't show that as image 10 :(
I should really make a rule not to post on the go, I even made a mistake on the number of channels your bms has :oops:

This is the image I meant, the connector should be easily removable after applying some force (usually they dab a some hot-melt-glue to which easily breaks off)
YTBMSLEAD.jpg


Coincidentally I recently sent this to a friend who wanted similar diagnostics, it shows how to use the balance plug to read voltages with a cell-meter, but you can do the same thing with your multi-meter
https://dl.dropboxusercontent.com/u/24071298/ebikebuildthread/YTBMS.mp4

The only issue is if one of the voltages reads '0V' and you misinterpret that as a dead cell, when it a disconnected lead at the cell-tab.
 
so you have decided it is too much trouble to find out if there is a #13 or to measure it? we could tell you if there is a #13 if you just posted a picture of the pcb where the tabs are soldered in place.
 
Back
Top