SIGNALAB (Ping) BMS Schematic

Niky1961

1 mW
Joined
Sep 10, 2008
Messages
16
Hey All,

I am going to investigate my SIGNALAB BMS since it looks like it's cutting off prematurely at cold weather.
I couldn't find any complete information on-line, so I decided to make a schematic and identify all the components myself.
I was not trying to identify FETs - there are plenty of different types available and it's not critical.
Here is my result.

Enjoy,
Niky1961
 

Attachments

  • SIGNALAB_BMS_rev.0.tif
    32.2 KB · Views: 3,194
So... did you identify what the DIP chip was? Was it the NAND gates in your schematic?
 
To swbluto:
Somebody on this forum mentioned CD4011, I checked and it functionally and electrically fits. I believe, it can be almost any CMOS 2-input NAND gate.

To fechter:
No, it’s not. It’s using first cell NCP800 chip over current protection circuit to protect the whole battery. The threshold is ~200mV, so in my case with current sense resistor of 40mOhm the maximum current is ~50A. See NCP800 Datasheet: http://www.onsemi.com/pub_link/Collateral/NCP800-D.PDF

Any questions – just ask.

Thanks,
Niky1961
 
fechter said:
The overcurrent protection circuit is tricky.

And rather sucky in my opinion if shorting allowed my battery's BMS to be killed.

And, yes, thank you VERY MUCH for posting this! If only I had seen this about, oh, 2 weeks ago, I might've made an attempt to repair my battery's BMS instead of replacing it. And I think I have a 4011 NAND chip from my childhood 101-in-1 radio shack electronics kit too! Haha.

For the next time I short the BMS! :lol:

(Why do some of these smilies look like aliens all of a sudden? :shock: Can't forget the little green men! :mrgreen: )
 
And to those who don't know, Ping's original batteries uses the Signalab BMS. I think another vendor also used this BMS, but I don't know who.
 
my BMS in the version 2 pack (just got it before the V2.5 started) was signalab but it didn't have any of the big resistors like i have seen in pictures posted on here. it was all surcace mount technology and looked very high tech with LED's for each cell group that indicated something to do with balancing (discharging?). the LED's flashed possibly to indicate sensing or charging and then turn on when the cell is full. during balancing the LED's gradually all switch on, one at a time, and when the charger switches off they gradually all switch off, one at a time. it didn't get too hot (~40*C)
 
Yes, some pictures would be nice. It's also interesting that he's using a different schematic already before announcing an "officially" different battery, but it could just be minor changes/enhancements. Anyways, leds for each parallel cell group sounds nice! Are those Surface Mount LEDs, by the way? He should start packaging operating instructions so we can understand what the blinking means. :lol:
 
Yes, interesting, indeed. This doesn't appear to have any sort of shunt-based balancing, or am I missing something? It just looks like once a cell voltage reaches 4.3V, it cuts off the charge current. Can't see how that would ever let a low cell catch up.

Also, I've never understood why these imported BMS boards do discharge current protection at all. Since all the current has to go through all the cells, all the time, current limiting can be done at the pack level, like in the controller. I guess it is because the packs these are used with need absolute protection from trying to pull too much current.

-- Gary
 
I haven't looked at the schematic in that much detail, but my experience with the charger is that it cuts-off when one of the cells reaches the high voltage threshold and then I assume the cells lower in voltage by some mechanisms(Perhaps its selectively draining the ones that aren't low? Or perhaps the voltage naturally lowers more on the high voltage cells?) and then it resumes charging once "it" goes below a certain threshold. I don't know what "it" is and how it's determined, but it must decide to resume charging based on something. It then goes through this cycle and this is what I presumed was "balancing". I don't, however, think this balancing, if it is balancing, is very effective for large SOC differences as it seemed like it'd never balance the faulty low-capacity parallel cell group I had.
 
It is really not balancing at all. If you watch what happens when cells start to get full, there is a sudden change in the rate of voltage increase, once a LiFePO4 cell gets to the 3.65-3.70V point. At that point, if unchecked, it will rise very quickly up over 4V. The current will be next to nothing. What a proper charger does is to hold the voltage at that 3.65-3.70V point, which will cause the cell to start reducing the amount of current it can accept. This "CV" mode is how you get the last 10-15% capacity into the cell. Cutting off the charge current will let the high cell's voltage drop below the cutoff, but it will very quickly rise up to the cutoff again. With perfectly matched cells, this might work okay, although it would take forever for the cells to get completely full. With some low cells, only the high cell might get close to being full.

What most BMS designs do is to include a shunt circuit in each channel that in effect, performs the CV mode. Basically, it is like a very efficient zener diode. It will hold the cell's voltage at the 3.65-3.70V point, and will bypass any current the cell can't accept anymore, up to a limit. For most imported BMS designs I've seen, this bypass current is usually somewhere between 75-200mA. What this does is let each cell have at least some current to finish charging. If the shunt current is fairly high (our BMS board allows about 500mA of bypass current...), that will shorten the time it takes to balance a pack.

-- Gary
 
I think you're right.

party0038.gif
 
damn it! i'm trying to upload pictures but my home internet connection is broken and thats where the pictures are!
 
the LED's are surface mount. i *think* it has balanced my pack because dogman said it was well out of balance and after a 24 hour charge it seams good -no more BMS cut outs. i'll have to check the cell voltages again sometime but im having too much fun atm.
 
The AP431 looks to be doing the balancing on this BMS. It bypasses some current through the 33ohm resistor whenever the cell voltage goes high enough. Stringing together a bunch of single cell BMS chips to make a multi-cell BMS is clever. Unfortunately the NCP800 look to be a LiCO part with an over-voltage threshold of 4.35volts, less than ideal for LiFePO4. Under-voltage threshold listed on the datasheet is fine at 2.5v.

Lawson
 
OK, I just got a slightly fried one from nicobie, so I'll get some better pics when I get a chance.
I'll be curious to see what it does when I hook it up. Thanks nicobie!
 
just a thought really. it should be possible to add a 10K resistor to the gate of Q2 and Q3 (where R23 is connected). the other end of this resistor could then be wired to a small switch to allow it to be shorted to ground. this way the pack could be turned off and on using a small switch. the pack would be turned off when the resistor is connected to ground.

rick
 
rkosiorek said:
just a thought really. it should be possible to add a 10K resistor to the gate of Q2 and Q3 (where R23 is connected). the other end of this resistor could then be wired to a small switch to allow it to be shorted to ground. this way the pack could be turned off and on using a small switch. the pack would be turned off when the resistor is connected to ground.

rick

Yes, this would be a good idea. It would be like a solid state contactor relay. I don't know why they didn't do something like that. What happens when you switch on against a big bank of capacitors in the controller? No contacts to weld, but the surge might be rough on the FETs. Would you still need a precharging resistor? I suppose if you could make the FETs turn on sort of slow, but still turn off quickly, you could soften the startup.
 
more
 

Attachments

  • DSC00858.JPG
    233.5 KB · Views: 2,380
  • DSC00859.JPG
    DSC00859.JPG
    154.1 KB · Views: 2,658
Interesting. The sample I have appears to be an older model.
Here's the top side: You can see there is an obvious problem here.
Signalab top side.jpg
Here's a closeup of the regulator chip:
regulator chip.jpg
Here's the shunt transistor.
shunt transistors.jpg
These are the optocouplers:optocoulplers.jpg
 
Hi Fetcher,

I see you got the cooked BMS. It is from a Ping V2 split 48 volt pack that is about 4 months old. I wanted to add a plug in the balance wires so I could more easily remove the pack. Well, it appears that I lost track of the wires because when I connected everything up, I let the smoke out of the resistor.

Nick
 
i have split my 48v pack in the same way nicombe. what caused the smoke in your case do you think? i know that without half the balance wires the pack voltage goes down which i found weird. i dont really know how BMS work but if you try and charge or discharge out of this lowered voltage condition then its probably not good.

check my build in the ping battery box thread http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=8659&start=15
 
Back
Top