BMS not working properly

[Pablo_1985]

hi guys, i have a 16 cells 38120 cells pack, (48V 10ah),the BMS is a 200/80A and because i have spent an entire month (but always conected to the charger) not using the ebike i wanted to check voltage in every single cell. and this is what i get

3,859
3,575
3,577
3,333
3,566
3,340
3,335
3,576
3,335
3,576
3,349
3,566
3,566
3,573
3,582
3,575

WTF 0,526V is the maximun difference! tell me what do you think about this...i have already sent an email to Michelle (michelle@xhnykj.cn), and i´m waiting for a response. i have use very little the ebike, because i have broken controllers, motors, furthermore personal problems... so i have use them just a little! and never deep cicles, in fact i think i never use more than a 20% of the total capacity.

Regards

 

[snowranger]

This doesn't seem weird at all. Once you burn off the surface charge, they'll probably all be around 3.3 volts under load.

 

[jondoh]

I agree, this doesn't look too bad.

The couple that are at 3.3x volts... maybe hit them with a single cell charger or power supply and WATCH THEM as the voltage rises. Bring them up to 3.5v (no need to go much higher) just to get them in the range. Try to get a sense of the milli-amp-hours going into the cell. If it takes very little like say 0.2 amp hours or less, the pack is balanced enough.

The problem that i'm finding with lifepo4 cells is that the charging "knee" is around 3.5 volt or so. Any additional charge can dramatically increase the voltage on the cell-- a few hundred milliamps can do it. While 0.5v seem like a lot-- it's not at the top end. If we're talking about cell ranges from 2.8v to 3.3v, then something may be wrong.

 

[dnmun]

you are right, the BMS is not balancing the cells properly. how did the #1 cell get up to 3.9V and there is one down around 3.33V?

do you have a picture of your BMS you can post up? i am assuming these are headway cells even though you did not state so.

how many balancing cycles too?

 

[snowranger]

I think he says that the battery has been sitting around for a while. When I first got my headway pack, it was way out of balance. We single-charged various cells until full capacity was achieved. The bms balances very slowly, and you can accelerate the process with some single cell chargers. The bulk charger has to be set at the right voltage to activate the shunting too (appx 58.9 volts).

 

[sabb]

So most bms just limit the upper voltage on the cell in this case it looks like it is set at 3.9 volts and has not allowed any cells to go higher.

when I add these up I get 56.283 and I would guess the charger is not putting out enough voltage to get most of the cells to cutoff

the charger on my 16s cell man pack charges to about 58 volts before going constant voltage.

Did you get the BMS and charger from same source they dont seem well matched

Stephen

 

[jondoh]

you are right, the BMS is not balancing the cells properly.

Ok... yes... maybe the BMS is not working. Having charged a headway pack without a balancing BMS, I got similar numbers. I still don't think there's anything wrong with this pack. Stay above 80% DOD and you won't have a problem.

Lifepo4 is not lipo. Different chemistry, different animals. Lipo is definitely high maintenance in comparison to Lifepo4. If this were a Lipo pack getting these numbers, yes the BMS is not working and this pack is NOT balanced. As this is a Lifepo4 pack, I think it's fine.

 

[Pablo_1985]

Wow! i didn´t expect so many answers!

I thought also about the great diference that makes some miliamps when the cells are high in charge, but i´m a newbie here and i prefered to check with you guys.

The cells are headway, and i bought the charger from the same person. BUt yestarday after post this, i frocked up....i´m a jackass! the bms is conected a every single cell by two conectors, one with 5 pin and the other with 12 pins. so, because i wanted to see if any component look damaged or something i put out this connectors, but fool of me when i connected them i missone pin, so i put one of the connectors moved one pin... something smoke ...i put out as fast as i could the connector, but too late... because there is a heatsink i didn´t see wich component i burned, so when i put out the heatsink there is a tiny zener wich could look bad... i check voltage in each terminal and it is 0V, i didn´t check resistance because i was too upset to think with clarity, today when i came back home i will check, and i will upload some pictures, if the BMS was good, is not longer good because of me.

I will talk again with michelle telling about this, let´s see if they can help me. (she has already respond saying basically the same as you guys.)

Regards.

 

[dnmun]

make sure you disconnect the sense wire plug if you leave it like that.

if that was your initial cycle then the pack may not have had a chance to balance yet but that was too high on #1 and have all the others so low.

why not post up a picture of your BMS and what smoked? was it one end of the plug inserted part way or the entire plug? not sure what coulda happened.

 

[Pablo_1985]

make sure you disconnect the sense wire plug if you leave it like that.

if that was your initial cycle then the pack may not have had a chance to balance yet but that was too high on #1 and have all the others so low.

why not post up a picture of your BMS and what smoked? was it one end of the plug inserted part way or the entire plug? not sure what coulda happened.

Now i´m at work this afternoon i will upload some pictures.

 

[dogman dan]

Cycle the battery one time, and then look again at the voltages. I think it's just a pack that sat around. Some cells hold a charge better than others, plus the effect, if any of the load from the bms on one or two cells.

 

[Pablo_1985]

Ok, guys i have some time to upload some pictures as i said before.

Can you see the tiny zener? i had to use a lens for making the photo...

Also you can see the temrinal strip, move one of the terminal just one position wrong (i didn´t expect to fix, i thought it was a "just one way" terminal...fooll of mine...)

How could i check if the BMS is working or not once i change this zener (if it is the only problem)....

 

[Pablo_1985]

By the way i´m not sure about why is necesary a BMS and neither how it works... it´s time to learn about this if i want to repair it (i want)

I supose that because little variations between cells even when they are in serires and they "feel" the same current, they could charge differently. So you have a net resistor for connect cells between them, you measure the voltage between cells and see the difference, so you balance the charge by connecting cells using resistors to make the connection. Also for not exceeding the current limitations, the maximum and the minimum voltage (of each cell), you have a power step full of mosfet wich will cut the current as it is needed, not draining too much the cells, not demmanding too much current for the cells...am i close?

correct me if i am wrong please

 

[dnmun]

wow, this is great. you have the headway 16S BMS and there is no black potting compound on the gate drive transistors across the top there.

mine has black potting compound from one side to the other, from J2 over to ZD2 but i cannot see where the trace that comes off your diode goes to in your picture.

very good pictures, but take another, face on, even closer if possible of that area from J2 over past the zener location that is empty at ZD2, with the wires in the chafing sleeve out of the way , and then turn it over and if there is no potting compound on the back then take a picture of that very close up too.

this is the only headway BMS i have seen with no black potting compound on it so we need those pictures to analyze it.

take the picture on the back of the region above the BR5 and up to R1A, all covered by potting on mine. especially the outside edge on the left. way cool.

 

[Pablo_1985]

OK, i will do that but jsut with one condition.... help me to repair my BMS... hehehe, i´m kidding, i will do it when i get home.

 

[dnmun]

sure, that's the idea, help you fix it. i don't know much more than you either so you gotta learn a bunch yet too.

why did you think the diode is bad?

when you plugged in the sense wire plug 'one off' and got the spark, which plug was it and how was it lined up? did you have just one corner on or the whole thing?

but it really helps that we can see it because mine is all covered with the black potting compound and i can't determine where the traces run.

but i am surprised it was so unbalanced because these headway BMSs are famous for balancing to within 10mV. they actually add small resistors to the divider bridge at final test to calibrate each channel to less than .5% or so. that's what those bare exposed pads on the back are for.

 

[Pablo_1985]

I didn´t see any spark, i just see smoke in one component...but because of the heatsink i could not see what was smoking... i can´t remember what i did with the conector... i think* that i put the 5 terminals conector one more inside, so lefting one out....but i just think* i´m not sure...

Im electronic engineer i´m just not familiarized with BMS

THanks a lot, this afeternoon, i will post more photos

 

[dnmun]

ok, thanks, really good you are already familiar with electronics. you will know by the end of this.

on mine it is 4 pins in the second connector, and they go to the top end, but it seems like moving one over would not cause a failure, but if the plug was reversed, then the shunt transistor woulda been reverse biased and coulda smoked that way.

this may take some smart people but we got some here.

leaving off the top plug, may be able to see if it works with just the bottom plug, up through channel #12.

 

[Pablo_1985]

Ok, i have done a million photos, they are here:

http://picasaweb.google.com/nuncaestardesillegaspronto/HeadwayBMS20080A?feat=directlink

I think there you can see everything... if not, just ask for it. I´m going to keep investigating.

Regards

 

[Pablo_1985]

Ah! i have a burned track! yo can see it in the photos! let´s try to find something else...maybe it is just this!

 

[Pablo_1985]

Ok, its a little hard because the electronic is very tiny, but i think i could write the circuit...

The main thing is all the terminals are interconected between 20 ohms resistors (the blue big ones) and a PNP transistor (S8550D)...in the base of the transistor i have an strange 3 legs component with 437 written on the top of the component (maybe a comparator? or a power step for powering up the base of the PNP? )... then....oh, i´m tired guys, i have spent all the fu---ng day making a wiring diagram for a power plant...NIGHTMARE, maybe tomorrow will be a better day my eyes hurts!

At the end i thing it is just some comparators and some easy logic, maybe a smidt trigger for not being working with any tiny difference in voltage cells and having some margin (some hysteresis). also i can see that this "logic" is also connected to the power mosfets, so it make sense to cut the current if voltage is too low, cut current if it is too high (that´s why there are shunts, to convert current in volts using a very little resistance)...

I think if i fix the burned track everything will work, because at the end it has work as a fuse protecting the rest of the component, furthermore the rest of the components seems perfect (i was wrong with the zener (in fact i thing it is just a normal diode) today i checked it and it seems allright, i had doubts because of the color, but with a lens i can see is just paint)

Regards

PS: if anyone thinks that i could be wrong in something, please tell me, i am a newbie in BMS, i just said what seems to be logical.
EDITS2not play station II... bad joke sorry XD), i´ve been thinking, and i think i know what happend.one terminal with some voltage the othe one with nothing, so the transistor open and let the current flow, but this is too much difference, so the current is quite high, high enough to burn the track, so if the track is bad, the transistor is ok, in fact i check the transistor and all the rest and everything seems ok!

 

[dnmun]

superb. totally awesome. glad this happened to you so we got such great pictures.

you are right about the diode, does not look blown. could not see the burned trace.

gonna take some time to digest this and maybe someone smarter will offer advice too.

anybody have the a useful code for the chinese surface mount devices?

i figure the *06 is for a 3906 knockoff, and maybe the 431 is a voltage reference, and you are right about the 6 legged sop parts, they are comparators for the HVC and the LVC signal lines which run across the board and the top signal line is the LVC i think because it does not seem connected to the gate drive for the output FETs.

wiring diagram for a power plant? aren't they supposed to do that first? really, this is gonna be a cakewalk for you. thanks for the pictures.

 

[Pablo_1985]

superb. totally awesome. glad this happened to you so we got such great pictures.

Mmmm thanks...? i prefer not to break my electronics, but as we said in Spain "No hay mal que por bien no venga" wich mean something that every bad thing is asociated to another good one.

you are right about the diode, does not look blown. could not see the burned trace.

down to the right, is easy to see...

http://lh5.ggpht.com/_nNioXvumFtY/TOQgdMXwlwI/AAAAAAAADGs/jKdUXHMjH5M/171120101413.jpg

wiring diagram for a power plant? aren't they supposed to do that first? really, this is gonna be a cakewalk for you. thanks for the pictures.

Yeah not the entire plant, you can´t do this in one day, this is just a piece...maybe 500 signals and it took me more than one day, but it is not like electronics, it is just like a huge sudoku hahaha.

Thanks a lot, let´s see in a week or in a moth i hope everything works ok, i don´t have so much tiem now to spent my time repairing... but if it is finally repaired (and i think so) you encourage me to do it!