Monitoring cell voltage

What is the best way to monitor individual cell voltage for a 20s pack?
I have a small project going on, where I am upgrading a small EV from lead acid batteries to Nissan Leaf batteries. It has a 13kW motor and I am having troubles deciding to go with a BMS or not, because of the cost. I guess the BMS would have to handle some amps. A cheaper way could be to hook up a BMS for the charging part only, but then I would want to monitor the cells when driving to avoid too deep discharge levels.

Any ideas are welcome!

Solarpower said:

...A cheaper way could be to hook up a BMS for the charging part only...

Click to expand...

That's what I've taken to doing which I learned from my recent lawn power tool pack experience. 20A and above BMS kinda pricey and when they die out on a deserted road it really sucks. Many 5-7A BMS options on eBay.

I'm not an EE but it seems we should be able to construct a simple LED/buzzer warning circuit plugged into the charge port while riding to alert cell level ODDV (over discharge detection voltage)?

Something I wanna play around with but for now a small BMS can keep a pretty good handle on charging conditions and if charge terminates early I know to check cells and manually bring 'em back into balance, if needed.

You could use any of the various multicell monitors for RC LiPo packs, mounted in the dash and with cell numbers marked next to each one for reference. I'd recommend a type like Celllog that can log the data for later review if you have a problem.

To prevent draining the pack when not in use, add a set of relays between the monitor s and the cells, so t hat only when your key is in one of the non-off positions does it monitor them for you. Just power the relays from the 12v system at a point past the keyswitch relays.

You can do the same thing with the charging BMS cell wires, so it also does not drain the pack when not in use, but you can make the relays for that powered by the charger instead.

While monitoring all the cells all the time sounds great, you really don't have to do that.

Using the RC stuff, such as cellog 8, you can find out which cells consistently have the least capacity. Then you need only monitor them in daily use. You stop when your lower capacity cell is near empty, and the rest are fuller.

Yes, of course, you do have to periodically make sure one of your better cells has not become the worst, but you can alert to that happening from changes noted when monitoring the capacity of the entire pack. I mean, you take your routine drive to the store, and routinely come home at x voltage. When you start getting home with less voltage on the whole pack, something changed.

Here’s what I’ve taken to using for 16S RC Lipo ebike pack -


http://www.ebay.com/itm/201434924156?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

It’s too early to declare a win but 4qty for $10 and they appear to play well together in series. I’m currently trying to determine how much parasitic drain while sitting? Also, still unsure if they draw power evenly from all cells? Or, is it the CellLog 8 problem of a couple cells powering the circuit and over long periods of time will drag the pack out of balance?

Got ‘em plugged into 4qty 4S Turnigy hardcase (16S, 60-67V) in the appropriate cell order and charge 2-5A through dedicated charge port using stock JST balance wiring/connectors. JST extensions soldered to the pads.

For discharge, I plug a cheap lighted volt meter into the battery charge port and if/when that goes OFF, I then know about a potential cell voltage problem. It shouldn’t be too hard to rig slightly more elaborate warning lights/buzzer that simply plugs into the battery charge port?

Basically, it’s cell voltage warning only during discharge and reliance on controller LVC but reasonably well protected charge and none of that stupid-ass bleed balance resistor heat and shorted switching transistor nonsense.

These boards have been terminating charge around 4.18V/cell. But keep in mind, I’ve only logged and compared a couple different channels so far. Other channels might be higher or even lower but I do like 4.18V…

If/when cell do ever go out of balance, I can go in with a USB power bank along with a cell phone BMS or other Lipoly 1S USB charger and merely bring up a low cell. Knowing when there’s a cell problem is most of the battle.

As the lawn power tool companies have apparently realized, discharge is the least of a battery pack fire hazard concern in these applications. Much like a bike, lawn tools are used outdoors so smoke isn’t so much an issue while finding a reasonably non-flammable spot (middle of a paved road if nothing else) to dump it and get away isn’t too difficult. Plus, fire during discharge is much more likely to occur during lower SOC so there’s also less energy to dissipate and it's an attended situation too.

Charging however is a much more sinister/dangerous thing. Often happens indoors, unattended and cells are more likely to have their maximum or above full capacity available to dissipate into a building and/or enclosure. Living/garage space, smoke, etc.

Sorry to drone-on but this is a particularly pertinent topic to my recent experimentation with a different way of managing home built battery packs. Maybe something that deserves a little more investigation?

One can make and program an Arduino UNO microprocessor and a 48 channel multiplexer to continuously read [every few milsec] cell voltages 0 - 5 v of some 24 cells and program the desired output or actuation you want when such and such criteria are met. The time varying voltage read out of each cell can be sent to a PC if the data set is large. This monitoring setup cost is less than $100 and setup is very versatile.

I did not have this project fully wired before I got some bad cells and was able to zero in on the bad cells [ with a voltmeter, many labeled leads and some quick connection for RO's] without such sophisticated battery monitoring as the microprocessor/multiplexer permits.

hi. Can you share some details about multiplexer? thank you

Thank you for all the input! Much appreciated

Does anyone have experience with a 72V BMS? I found these two on eBay:
http://www.ebay.com/itm/72V-84V-20S-30A-20x-3-6V-3-7V-4-2V-Lithium-ion-Li-ion-Li-Po-Battery-PCB-BMS-/321745436164
http://www.ebay.com/itm/Li-ion-BMS-PCM-for-72V-20S-50A-Li-PO-Battery-Pack-With-Balance-LiCoO2-LiMnO4-/121911748057

I just sent the sellers a question asking how many charging amps they handle...


@DingusMcGee, do you have any more information on your solution? I wish I had any Arduino experience, but I dont

@Ykick, I guess I need a bigger BMS than the one you use? How many amps does that handle while charging? I am planning to use the original charger for bulk charging the batteries, and I think it is 30A. Is this a very bad idea? I found the charge curve, and for me it seems that this would be OK except for the last part, T3?