Battery click sound, now showing far too low of voltage.

[fechter]

I must be missing something in the picture. Looks pretty normal to me.

To measure the voltage on the end cells, measure from the main battery - wire to the bottom tap wire, and from the main battery + wire to the top tap wire.

With the battery all hooked up, if you measure (carefully) across the two outside legs on a FET, it will tell you if it's getting a gate signal. Below about 2v, it is off.

 

[ihategeeks]

I must be missing something in the picture. Looks pretty normal to me.

To measure the voltage on the end cells, measure from the main battery - wire to the bottom tap wire, and from the main battery + wire to the top tap wire.

With the battery all hooked up, if you measure (carefully) across the two outside legs on a FET, it will tell you if it's getting a gate signal. Below about 2v, it is off.

This is normal? http://i.imgur.com/XjfPPKR.png

 

[ihategeeks]

http://i.imgur.com/o4Fgpda.png

voltages across outside legs of mosfets.

 

[ihategeeks]

To measure the voltage on the end cells, measure from the main battery - wire to the bottom tap wire, and from the main battery + wire to the top tap wire.

Not sure I am doing this right.
I put the black lead into the black discharge, and measured the "bottom tap wire?" which would have been the bottom in my picture.
http://i.imgur.com/7GCW4rO.jpg
Then I put the black lead on the "top tap wire?" and put the red lead into the red discharge wire?

The bottom tap wire gives me 14.9
and the top tap wire gives me 9.9
everything in between is still 3.32

I feel as if that is drastically wrong, or I am doing it wrong.

 

[dnmun]

not sure what you are doing. what is a tap wire?

in your picture you have a yellow line in between the two adjacent pins of the sense wire plug and that gives you the voltage across the cells. to measure #1 you put the black probe on the B- spot of the BMS and the red probe on the first pin, then #2 is between the bottom pin and the second one, and so on.

you should have a higher gate voltage than 4.75V on the discharge mosfets and the .45V does not make sense. i have one 16S v2.5 signalab on a 16S ping pack here and it has 13.3V on the discharge mofets and 9.7V on the charging mosfet.

 

[ihategeeks]

not sure what you are doing. what is a tap wire?

in your picture you have a yellow line in between the two adjacent pins of the sense wire plug and that gives you the voltage across the cells. to measure #1 you put the black probe on the B- spot of the BMS and the red probe on the first pin, then #2 is between the bottom pin and the second one, and so on.

you should have a higher gate voltage than 4.75V on the discharge mosfets and the .45V does not make sense. i have one 16S v2.5 signalab on a 16S ping pack here and it has 13.3V on the discharge mofets and 9.7V on the charging mosfet.

Do you take the gate voltage with the sense wires plugged in and the battery on the charger? Or just the sense wires plugged in and the battery off the charger? I did it with the sense plugged in and the battery off the charger

 

[dnmun]

if you have the sense wire cable unplugged the discharge mosfets will be turned off. you do not have to put the battery on the charger to measure the gate voltages. i just do not know where you are measuring.

 

[fechter]

Tap wire = sense wire. Wires that go to the connections between cells.

It looks like the discharge FETs are getting gate feed and should be on and the charge FET is off. I might have a Signalab board like that around here. I'll try to find it and take a look.

The suspicious looking FETs in the picture may have just been a sloppy assembly job, but be intact. Generally when they blow, they short out. If you orient the board so the FET tabs are pointing up, the left leg will be the gate and the right leg will be the source. The middle leg is the drain, which is connected to the heat sink tab. If you measure resistance (battery disconnected from board) across the source and drain in both directions, it should look like a diode, giving a high resistance in one direction and low in the other direction. If your meter has a diode check mode, this works the best. The discharge FETs are all in parallel, so if any one is shorted, you'll get the same measurement on all of them. If it's shorted, it will read zero ohms in both directions.

 

[ihategeeks]

Tap wire = sense wire. Wires that go to the connections between cells.

It looks like the discharge FETs are getting gate feed and should be on and the charge FET is off. I might have a Signalab board like that around here. I'll try to find it and take a look.
The suspicious looking FETs in the picture may have just been a sloppy assembly job, but be intact. Generally when they blow, they short out. If you orient the board so the FET tabs are pointing up, the left leg will be the gate and the right leg will be the source. The middle leg is the drain, which is connected to the heat sink tab. If you measure resistance (battery disconnected from board) across the source and drain in both directions, it should look like a diode, giving a high resistance in one direction and low in the other direction. If your meter has a diode check mode, this works the best. The discharge FETs are all in parallel, so if any one is shorted, you'll get the same measurement on all of them. If it's shorted, it will read zero ohms in both directions.

http://i.imgur.com/o4Fgpda.png

So in this pic, the top leg is the gate and the bottom is the source on each mosfet?

 

[ihategeeks]

to measure #1 you put the black probe on the B- spot of the BMS and the red probe on the first pin, then #2 is between the bottom pin and the second one, and so on.
.

I put the black multimeter lead on B- , then the red lead on the first spot of the sense wire. (the very bottom contact in my picture, below the first yellow line)
The volt meter reads 39.9

I put the black multimater lead on the last spot of the sense wire, and the red lead on the red discharge wire. (the very top contact in my picture, above the last yellow line)
The volt meter reads 9.9

 

[fechter]

So in this pic, the top leg is the gate and the bottom is the source on each mosfet?

Right.

Something seems wrong with the measurements on those end cells. You should see something more like you had on the other cells.

You might have a bad wire connection somewhere between the actual cells and the BMS board. If there is any way to put the meter directly on the cells you could test.

 

[ihategeeks]

So in this pic, the top leg is the gate and the bottom is the source on each mosfet?

Something seems wrong with the measurements on those end cells. You should see something more like you had on the other cells.

Actually, it is exactly correct. The answer is pretty obvious too. 39.9 is the total voltage of 12 cells @ 3.32.
I was measuring the top of that series cable instead of the bottom, therefore I was getting every cell in that cable instead of just one.
Doing it right Cell #1 is 3.32

However, with the other sense cable, It should be reading 13.3, but it is only reading 9.9.
Testing each cell ... Cell #16 is not showing anything at all.

#1 3.32
#2 3.32
#3 3.32
#4 3.32
#5 3.32
#6 3.32
#7 3.32
#8 3.32
#9 3.32
#10 3.32
#11 3.32
#12 3.32
#13 3.32
#14 3.32
#15 3.32
#16 0.00

Ninja Edit* when I measure from B- to Cell 16, I get 53.2 Volts, so all cells are accounted for at the expected balanced voltage.

For some reason I am unable to take a measurement from cell #16 (black multimeter pin) to the red discharge cable (red multimeter pin). If I take the votlage for all 4 pins on that sense wire, I still get 9.9. Individual I can test all 4 pins to get 3 cell voltages. I do not understand this. Would this indicate that the sense wire is damaged, or that the red discharge cable is damaged?

 

[ihategeeks]

A summary of where we currently are.

I can get 3.32 from every individual cell except for #16.
I assume cell #16 has 3.32 because the voltage from B- to cell #16 is 53.1+
I am as of this point, unable to get any voltage reading by going from the sense cable at cell 16 to the red discharge wire.

A voltage reading across the discharge wires is 20.7

Voltage readings from all the mosfets are still the same as this picture http://i.imgur.com/o4Fgpda.png

The battery will not charge.

The battery will not power a cycle analyst.

 

[fechter]

It seems like the red discharge wire is disconnected at the battery or somewhere between. Those are usually pretty hefty wires and don't break so easily. The red wires all connect together and go straight through the BMS (or at least they're supposed to). All the switching happens on the negative side.

At this point, I'd suggest looking closely at all the red wire connections, especially right where it attaches to cell 16. It's possible they made a cold solder joint or something cracked and broke.

The cell voltages all seem OK and the BMS seems to be doing what it's supposed to.

 

[dnmun]

the voltage difference between the top sense wire from the top of #16 and the red power wire will be 0V. they are connected to the same spot.

 

[fechter]

The BMS has only 16 sense wire connections. If there were 17, then I would agree. At least one sensing connection is done with the main wire.

If you disconnect the BMS, you should be able to power a load directly from the main red and black wires coming form the pack (not going through the BMS). A voltmeter might 'lie' if there is a high resistance connection. Something like an old fashioned 120v incandecent light bulb should glow if connected to the battery wires.

 

[ihategeeks]

The suspicious looking FETs in the picture may have just been a sloppy assembly job, but be intact. Generally when they blow, they short out. If you orient the board so the FET tabs are pointing up, the left leg will be the gate and the right leg will be the source. The middle leg is the drain, which is connected to the heat sink tab. If you measure resistance (battery disconnected from board) across the source and drain in both directions, it should look like a diode, giving a high resistance in one direction and low in the other direction. If your meter has a diode check mode, this works the best. The discharge FETs are all in parallel, so if any one is shorted, you'll get the same measurement on all of them. If it's shorted, it will read zero ohms in both directions.

I also tested my 2 diode packs to make sure the stickers indicating polarity were correct, they are.

 

[dnmun]

what does 1,1 mean?

 

[ihategeeks]

what does 1,1 mean?

With the voltmeter in diode mode, 1 is the default reading on the display.

 

[ihategeeks]

The BMS has only 16 sense wire connections. If there were 17, then I would agree. At least one sensing connection is done with the main wire.

If you disconnect the BMS, you should be able to power a load directly from the main red and black wires coming form the pack (not going through the BMS). A voltmeter might 'lie' if there is a high resistance connection. Something like an old fashioned 120v incandecent light bulb should glow if connected to the battery wires.

The black discharge wire is connected to the BMS (P-), regardless if the BMS is connected to the battery through the sense cables, the current has to travel through the Mosfets.
If they are causing a problem, wouldn't that give me wrong numbers?

If I take the black at (B-) and the red at red discharge, that goes around the BMS entirely, and I get a volt meter reading of 53.2. As I should.
This indicates to me that the battery would work as it should if it did not have a BMS attached.

Also, when I measured the Mosfet gate voltage, if you add up all the values, you get around 20 volts.
When I take a volt reading at the discharge wires, I get around 20 volts.
Coincidence or proof of a problem at the mosfets?

 

[dnmun]

what is the actual number? is it just 1 ohm or is that the offset in your voltmeter? what does the diode tester read when you touch the probes to each other? sounds like the mosfets are shorted to me.

 

[ihategeeks]

what is the actual number? is it just 1 ohm or is that the offset in your voltmeter? what does the diode tester read when you touch the probes to each other? sounds like the mosfets are shorted to me.

I don't know what 1 means, it is just want the meter says in diode mode by default.
probes touched together the meter emits a steady beep and shows 0.

http://imgur.com/QAlCVkg

 

[dnmun]

while discharging the battery through the BMS put your finger on the mosfets and find the one that gets hot. cut the source leg and the gate leg off of that one and then go back and use the diode tester on each of the discharge mosfets again.

 

[ihategeeks]

while discharging the battery through the BMS put your finger on the mosfets and find the one that gets hot. cut the source leg and the gate leg off of that one and then go back and use the diode tester on each of the discharge mosfets again.

The fun stops here until I get word back from Li Ping then, I can't start cutting things off the board just in case he wants me to send it back to him.

 

[edventure]

what is the actual number? is it just 1 ohm or is that the offset in your voltmeter? what does the diode tester read when you touch the probes to each other? sounds like the mosfets are shorted to me.

I don't know what 1 means, it is just want the meter says in diode mode by default.
probes touched together the meter emits a steady beep and shows 0.

http://imgur.com/QAlCVkg

I checked the manual for that meter here http://www.southwiretools.com/tools/file.get.do?file_id=243 and a 1 indicates an over range situation which basically means high or infinite resistance. This is what I would imagine you should see between the Source Drain pins if the FETs are not turned on. The one that sticks out is that top one, but I never trust testing components like FETs when they are left in the circuit because this can give you spurious results. I think Dnmun's idea of checking to see if one the FETs is getting warm may be a quick and easy way to try. Have you tried running the pack with the BMS bypassed to prove to your self it is the BMS and not a connection issue? This may be something else to temporarily try in an attempt to confirm it is a BMS related issue.