Testing a 10S 36 Volt BMS

I have an ebay 10S 36Volt 35A Li-ion Lipolymer Battery BMS and 10 Li ion cells in series for 36 volts. initally charged to 37 volts, all batteries at about 3.7 volts. Im using a brushless 350 watt motor and controller

all of the cells are now at 3.5 volts except for one that is at 2.9, I will remove it.

I rode the bike a few KM and voltage went down to 30 volts, and something cut out, dont know if it was the BMS or controller. I understand that is low voltage, but I would like to know if the BMS is working. I tested at the battery output at I found it to be 30 volts or so, but if the BMS was working, would there not be 0 volts?

I tried to test the BMS, and each red wire tested with the mult-meter, between black, would read a fluctuating voltage...like 5, 6, 8 etc (dynamic, not a steady state). I thought each wire was supposed to read 3.7, for ex. to each cell. all voltages should be the same? Same issue between red sense wires; fluctuating voltages, not a steady 3. 7 or 3.5.

is there a way to test the BMS? I have included a diagram from the supplier, and a picture of my battery setup. A few of the red sense wires are disconnected now, as I was testing them individually. Same results. My question is, how do I know the BMS is working?

I dont think I have it connected wrong. I have the sense wire connector black wire going to the end series battery negative. then another black wire going from the end negative series battery, to the BMS B (-) pad. In the diagram, they seem to have that negative wire cut out of the picture, but its supposed to be connected to the B (-) pad, on the BMS.

the last red sense wire is going to the series battery end positive, and another red wire going to the load, and the BMS P (-) going to the load negative.

newb123 said:

initally charged to 37 volts, all batteries at about 3.7 volts.

Click to expand...

That can be anywhere from somewhat less than half charged to somewhat less than mostly, depending on the particular cell model/chemistry. 3.7v can be a pretty wide band of state of charge, fairly flat part of the curve, so you could have some cells with 50% charge, some with less, and some with more, so some of them could run out a ways before the others.


4.2v is usually full, and is where the BMS will usually balance at. If you don't charge to 42v for the pack voltage, the BMS is unlikely to balance the cells. (which can take hours to days depending on how unbalanced they are, like the below, which could take days to weeks to fix, leaving the charger on all the time).

I'd recommend doing the cell bank testing after doing a full charge and letting it balance, just to be sure all cells are in the same state of charge (SOC).

all of the cells are now at 3.5 volts except for one that is at 2.9, I will remove it.

Click to expand...

If it's at 2.9v after the BMS cut out, then that's why--under load it's dropping below the BMS LVC, so the BMS shuts off, at least until the load goes away.

I rode the bike a few KM and voltage went down to 30 volts, and something cut out, dont know if it was the BMS or controller. I understand that is low voltage, but I would like to know if the BMS is working. I tested at the battery output at I found it to be 30 volts or so, but if the BMS was working, would there not be 0 volts?

Click to expand...

If the battery is connected to a load, yes. But if it is not, the output can "float" to some voltage anywhere between actual and zero.

The controller probably has an LVC around 30v, to protect the cells from going too low. Higher would actually be better, mroe like 3.2-3.3v/cell, easier on the cells for longer lifespan.

The BMS is probably closer to 28v, but that is just an emergency last-ditch cutoff to save your battery from damage. Regularly running down that far is hard on the cells.

I tried to test the BMS, and each red wire tested with the mult-meter, between black, would read a fluctuating voltage...like 5, 6, 8 etc (dynamic, not a steady state). I thought each wire was supposed to read 3.7, for ex. to each cell. all voltages should be the same? Same issue between red sense wires; fluctuating voltages, not a steady 3. 7 or 3.5.

Click to expand...

Sounds like a meter or lead / connection problem.

As for all voltages being the same--that's only true if all your cells are balanced.

is there a way to test the BMS? I have included a diagram from the supplier, and a picture of my battery setup.

Click to expand...

Depends on what you want to test. Sounds like in this case you want to test the cell-level LVC.


The second best way is to use a load (current drain) similar to the bike's to place on the battery to drain it while watching the voltage of each cell and the main output as well. Then when it does cut off, you can see the voltage of each at that moment, so you know what it is cutting off at.

A toaster oven or other heating element (taht doesnt' have a motor in it) is probably enough, but might require more than one in parallel. You'd have to measure the current the bike draws on average, and then measure the current drawn by the load, to be sure it's about the same. You want something about the same so that it can cause simliar voltage sag to the average bike load (sag will trigger the BMS LVC before the cells are actually empty, so you want to know when that is for your cells).

The best way is to use isolated-output adjustable power supplies to represent each cell, and monitor the output of teh BMS, so you can lower each supply until the BMS output cuts off, and then note the voltage that sense wire was triggered at. Repeat for each one to verify they are all the same and working.

Similarly, you can raise the voltage on each one until the input of the BMS cuts off (this is harder to test; you'd have to be measuring the resistance across the input FET(s)).


The way you'll probably have to do it with equipment you already have is just to partly charge the pack a little bit, then monitor one sense lead and start discharging it, either on the bike or with a heater (or other) load. When the pack cuts out, note down it's voltage. Do the same for each sense lead, and you'll at least know the BMS is working on the lowest voltage group. (you would have to swap groups around to test each group this way, which is too much of a PITA to do for real).

A few of the red sense wires are disconnected now, as I was testing them individually. Same results. My question is, how do I know the BMS is working?

Click to expand...

Were you testing the sense wires, or the cells?

If you measure the sense wire with no cell connected, there's no meaningful reading to be made.

Measuring the cells from negative of one group to that group's positive (the negative of the next group) will get you the voltage of the cell group.

thanks very much for the detailed explanation. I will start by charging each cell to 4.2 volts. I thought the pack was full at 40 V and connected that, soon after I checked each cell and found it to be 3.7.
Dont know if the BMS is doing some balancing. another cell was low after a ride, and the BMS cut out.
Total voltage was at 34 volts.

for the weak cell, will parallel it with another cell to see if I can get more capacity. I only have 11 cells to work with now.

Also, is it best to charge (Im using a CC/CV buck converter, this is a poor mans project for now) in series or parallel?
and, I dont have a 36 volt charger, but have a CC/CV boost converter. Ive seen solar panels with a booster charge Li ion batteries, is it doeable?

newb123 said:

I will start by charging each cell to 4.2 volts.

Click to expand...

INstead, if you want to test the BMS's HVC and balancing, you should just hook it up and charge via the BMS.

Your charger should be 42v.

Dont know if the BMS is doing some balancing.

Click to expand...

Unless you charge to 42v and leave it there long enough, it cannot do balancing.

YOu can measure the cells during this stage to see when they are getting close to balanced.

The charger will shut off and restart, how long between this I don't know. Depends on the BMS's bleed / balancing current on each cell.

HOw long it will take to balance depends on that and the imbalance between cells.

Also, is it best to charge (Im using a CC/CV buck converter, this is a poor mans project for now) in series or parallel?
and, I dont have a 36 volt charger, but have a CC/CV boost converter. Ive seen solar panels with a booster charge Li ion batteries, is it doeable?

Click to expand...

As long as your voltage is 42v, and the current limit is at or below the limit for those cells' spec for charging, then the source doesn't matter (other than being CC/CV).

If you have to go buying stuff for the project, I'd recommend just getting either an actual charger with the ratings needed, but the cheap ones will *be* cheap, usually cheaply made.

Or something like the genuine (no clones) Meanwell HLG or ELG series PSUs. They're almost the same potted and sealed so weatherproof'; can't remember the difference between them. I use the big HLG-600H-54A for 12Amp charging of my big pack on the trike; you could probably use the HLG-150H-42A, or even the HLG-100H-42A, depending on how much charge current you need. If you don't mind really slow charging you can get smaller ones (lower middle number), but the price per watt isn't as good the smaller you go. ELG-100-42 should also work, IIRC (might be ELG-100H-42A. you have to make sure you get the adjustable voltage and current model of whatever you go for).

Probably about $50 to %70 for a Meanwell, depending on which kind.

my boost converter has not arrived yet, so I am just using a 32 volt printer power brick with a CC/CV buck. The HP power brick I have is rated at 750 mA, and charging series strings of 7 cells at over 28 volts. to get the 4.2 each cell.

the remainder will charge individually or a set of 3 in series.

The BMS is not rated at 32 V, so cant use it for charging; sense wires floating, I read, doesnt work, and of course cant charge a 42 volt string with 32 volts, unless I have the boost converter.

The nice chargers or PSUs, you recommend, I will someday get.

I noticed that when I was at .5 amps, the buck converter charged LED (blue) stayed on when I had the voltage at about 29 volts but it changed to red when I brought the amp pot up to about .7 amps. why would it not charge at .5 amps?

It's best to have all cells balance before hooking up bms. Bms is used to keep a balance pack in balance. Not for balancing a wonky pack. Even if at 4.1v before hooking up. Do you have a rc charger ? Most phone charger are to high. Don't trust the voltage printed on the side check.

how do I get the cells to be all balanced at 4.2V?
Do I use a charger that balances the cells?
The BMS I have states that it will balance charge
Can I charge through, the BMS, to get to 4.2 V on each cell, then the BMS will keep the cells balanced when riding?

Check your cell voltage and write down on paper like
1. 3.95v
2. 4.1v
3. 4.0v

10. Xxx volt
Use a single cell charger and bring up the. Low cell to all are even like 4.1v You can use a phone Only if you seat a monitor the whole time without walking away. This so it doesn't over charge. Very dangerous to overcharge past 4.2v The writing on the phone charge my say 4.5v but if you check with meter you will find it is very different and as always chevk polarity also. That is a cheap way. But did I say youmust seat and monitor. Don't walk away. If you need to go unplug. If you get all even at 4.1v plug in and charge still do monitor.
What are you trying to power ?
If those cells can put out 35amps ? Of not they will always go out of balance.

Yes I did that with a CC/CV buck converter, tried to get all cells to 4.2, even though a couple went to 4.25 or 4.27. some I charged in series strings, other individually. all at about 700 mA charging current

Im powering a 350 watt 36 volt 10 amp motor

so, I have all the cells at 4.2 volts +

now, the BMS will keep them balanced?

Get a 12v brake light to discharge high cell. If 36v use a hair dyer on low to drain pack if to high.

Ideally you would put them all together in parallel at say 4.05, then have a charger / PSU that would let you bring them up while in parallel by another volt or even to 4.2, and **hold** that termination voltage until amps accepted drops to .01C.

Otherwise, after you think you have them all pretty close, hook them all up in parallel to **sit** isolated for a night, or couple days if you can.

They will all equalize to each other.

When you then confirm measure each separately after a few hours, any that are significantly lower V than the rest, are your weak links.

All this should be SOP before ever assembling your pack, buying extra cells so you can toss the lemons.

Otherwise, after you think you have them all pretty close, hook them all up in parallel to **sit** isolated for a night, or couple days if you can.

Ok I will try that. I had 9 in series and 1 pair in parallel (they were weak links) and rode the bike. I check each cell with the BMS connected and they had run down to 3.5 but there were 2 ( the ones in the pair I had initially charged to about 4.5V) that were not at 3.5 but at 3.8V.

I dont know if the BMS is working. Does the BMS work while not connected to a charger or load, but instead only to the batteries? I mean, is it working to balance as along as I have the sense wires and B+ and B- connected? perhaps it takes several hours to balance?

so your saying take all 10 batteries, get each individual cell charged to 4.0, then parrallel all the 10 like + - - +
+ - - +
+ - - +
+ - - +

First get them all top-balanced as close to perfectly as possible.

After you re-assemble the pack with the BMS,
First get them all top-balanced as close to perfectly as possible.

In theory every BMS is different, please link to the detailed spec sheet of the one you have, so we can help you understand how it works.

BMSs have many functions, balancing is just one of them, and most of them IMO do it very poorly.

To check, you need to monitor the voltages, ideally at the per-cell level.

Usually balancing only starts when the charger could be finished, but you have to keep going at a high voltage, possibly for days because the balance current is so low.

But, in theory at least every BMS is different, so please link to the detailed spec sheet of the one you have, so we can help you understand how it works.

Also of your charger, at least as important.

I assume you have a DMM ammeter and ammeter of some sort?

After you re-assemble the pack with the BMS, monitor the voltages at the per-cell level.

If the batteries are new and of good quality, the BMS will have less work to do, and ones you've established your regular routines, you won't need to monitor things so closely

This is the BMS I am using, sorry no P/N or other info included


https://www.ebay.com/itm/10S-36Volt-35A-Li-ion-Lipolymer-Battery-BMS-PCB-with-Balance-for-Ebike-Escooter/273371151970?hash=item3fa630be62:g:W9gAAOSwVVlcds7p

I think I made a mistake on the parallel config for charging it should be

+ -
+ -
+ -

etc

with all the + conected together and the - connected together then a lead off both common + and -, correct? to the charger

Yes I have a generic mulitmeter with amp leads and also amp clamp meter as well

Go with at least #2 here
http://www.smartgauge.co.uk/batt_con.html

And frankly you'll have to reverse-engineer if you can't find docs, if you value your time just get a better BMS.

Some you can adjust the setpoints, others can view everything on your phone via a BT connection, or both.

But at least with a good detailed spec sheet. . .

ok I will get the heat sink cover off later and post a pic of that, maybe it will explain more detail

and I will use method 4 of the site you linked to parrallel charge once I get each cell to 4.0V

Usually balancing only starts when the charger could be finished, but you have to keep going at a high voltage, possibly for days because the balance current is so low.

so this means that BMS balancing only occurs after charging has been completed, and at 40+ volts?

During a ride and discharge, how does the BMS keep each cell at the same level?

newb123 said:

how do I get the cells to be all balanced at 4.2V?
Do I use a charger that balances the cells?

Click to expand...

I gave you a method to do that already, via the BMS, but you have to have a 42v charger.

You can also use a 4.2v CC/CV supply to individually charge each cell to 4.2v.

then the BMS will keep the cells balanced when riding?

Click to expand...

As I already said before, the BMS *only* balances during charge, at the end of charge when it's full at 4.2v per cell.

You must leave it on a 42v charger, thru the BMS's charge port, for anywhere from several hours to several days (to weeks) to balance a pack, depending on the factors I already posted before.

newb123 said:

During a ride and discharge, how does the BMS keep each cell at the same level?

Click to expand...

It doesn't.

It only does it at end of charge.

All it does during a ride is to shut off the output if one of the cells drops below the cell-level LVC for that BMS, which is usually right at the lowest voltage the cell spec sheet would allow.

That is hard on the cells.

This is why the controller LVC is usually much higher, so the cells don't get pushed so hard.

The BMS LVC is a last ditch emergency cutoff to prevent damage and fire.

You can also use a 4.2v CC/CV supply to individually charge each cell to 4.2v.

thats what I have been doing, then I connected the set to the BMS. but once the set starts to discharge while riding, the BMS does nothing except protect against over current or discharge?

oh I see you answered that here:

It doesn't.

It only does it at end of charge.

All it does during a ride is to shut off the output if one of the cells drops below the cell-level LVC for that BMS, which is usually right at the lowest voltage the cell spec sheet would allow.

That is hard on the cells.


ok starting to make sense

I dont have anything that can charge up to 42 volts so I ordered the cheap boost converter to take my 20 volt power brick, and boost that to 42 at low amps.

Ok so the BMS will balance the cells once they arrive at 42V during charging with a 42volt CC/CV charger

. what if, after riding, I have some cells that are high, and the rest at 3.7V? That means those high cells are no good? I have no way to balance except by individually charging each cell to 4.2V
under normal conditions, all cells after discharged, should read the same?

also if I may ask, sometimes when I gun the throttle, the BMS will cut off the pack, even at 40V. that doesnt seem normal over current protection. What might be causing this?

newb123 said:

thats what I have been doing, then I connected the set to the BMS.

Click to expand...

Ok, then youre starting out with balanced cells, at exactly 4.2v each? In another post it said you charged at least one cell (pair) to 4.5v....

. what if, after riding, I have some cells that are high, and the rest at 3.7V? That means those high cells are no good?

Click to expand...

It means that some cells have more capacity than others, which means the cells are any or all of the below:
-- used / old
-- poor quality
-- not well-matched

under normal conditions, all cells after discharged, should read the same?

Click to expand...

New, good quality, well-matched cells will stay within hundredths of a volt of each other (or less difference) all the way from full to empty.

The BMS balance function is there because as they age, cells change, and each one ages slightly differently. When they get old or they're not matched to start with, then they can be radically different from each other in characteristics, so they dont' stay balanced.

I have no way to balance except by individually charging each cell to 4.2V

Click to expand...

Right now, sure. But once you have a 42v CC/CV charger, then the BMS will do it for you.

newb123 said:

( the ones in the pair I had initially charged to about 4.5V)

Click to expand...

4.5v is a pretty severe overcharge on a cell, and damages them.

This creates a risk of fire.

If you havent' already, you should look up the specs for your cells, if you know which brand and model they are, and not charge them beyond the limits specified.

Don't charge at a higher current.

Don't charge to a higher voltage.

If you don't know what the model is or cant' find a spec sheet, then:

-- do not charge at more than half an amp (500mA, 0.5A).

-- do not charge above 4.2v

-- do not discharge beyond 3.3v

-- do not discharge at more than two amps (2000mA, 2A).

Those would be safe limits for most typical cells when they're old / used, or not very well matched.

Quite a few cell models are much better than that, but if you aren't sure it's better to be safe than sorry, since exceeding capabilities can overheat or otherwise damage cells, which creates a risk of fire.

so your saying take all 10 batteries, get each individual cell charged to 4.0, then parrallel all the 10 like

Click to expand...

I'm not saying that (one of the other posters in this thread might be), but that is one way you could charge all of them to 4.2v at the same time.

But if you're already charging them to 4.0v individually, I wouldn't bother with the paralelleling, I would just finish the individual charges all the way to 4.2v.

If you parallel them that means you have to disconnect them all completely from each other first, or you short them out ahd possibly make a fire.

Disconnecting them all and reconnecting them in different ways each time you have to charge the pack, and then doing it again to use it, is pretty much asking for a mistake to happen at some point, and damage of one kind or another to occur.

Worst case is accidentally leaving them connected in parallel after a charge and then hooking up the series connections, whcih can lead to a fire if the cells get hot enough fast enough (since they'll be fully charged at taht point.

People using RC LiPo packs sometimes do this, and it has caused fires before.

newb123 said:

also if I may ask, sometimes when I gun the throttle, the BMS will cut off the pack, even at 40V. that doesnt seem normal over current protection. What might be causing this?

Click to expand...

Is the BMS cutting off the pack? That means there is no voltage output at all, and if you have an LCD or other display or lights running from it, those will shut off too.

This means (at least one of) the cells are being so heavily overloaded vs what they're designed for (or are so old and worn out) that they can't maintain their voltage under that load, and are dropping WAY down in voltage to around 2.8v per cell, while the load exists.

If any display/lights stay on, but the motor stops working, then that's just the controller LVC, and it means the same thing, but the voltage drop isn't nearly as bad, because the controller probably cuts out at 30-32v if it's actually a 36v controller.

The BMS may recover output but often they stay off until a charger is connected, for safety.

The controller always recovers (for the ones I've used at least) as soon as the voltage goes back up.

The voltage sag is only while it's under load, so a pack being used too hard or that is too old/weak could sag down to say 30v under full throttle, and may well read 40v once the throttle is off (load is removed).

It could even happen so fast that a typical digital multimeter wouldn't show this; they don't update the screen very often. An analog meter with a needle sweeping across a dial *would* show it if you were watching at that moment.

Just to reinforce

newb123 said:

During a ride and discharge, how does the BMS keep each cell at the same level?

Click to expand...

It doesn't, not how most BMSs work.

When you are confident that your cells are not getting out of balance, you do not need to worry about that.

I work with and know of lithium banks in other use cases that go for years never going out of balance. Others with OCD owners get manually balanced once or twice a year.

With these batts and a high C-rate usage like this, more care is needed, and the ideal setup is fully balanced automatically evey charge cycle.

Very easy to setup, just takes money.

Very few gadgets balance while in use, in fact little point in doing so most of the time.

The norm is, and all the time using the functionality built into a protective BMS, balancing only happens while charging.


> so this means that BMS balancing only occurs after charging has been completed, and at 40+ volts?

If you want to give the BMS time to **finish** balancing, you need to hold the charger above the start-balance voltage setpoint until the process is complete. Can take days if the cells have been allowed to get too far out of balance.

Very few BMS are so adjustable that the start-balance voltage setpoint can go as low as 4.0Vpc.

Hope that helps

The paralleling all of them together to get them precisely balanced is a **one-time** (hopefully) manual process, should be done at commissioning time before the pack is (re-)assembled.

And a substitute for very precise equipment that automates accurate cell-level balancing in batches.

Obviously less practical if welding/solder is involved in the intra-pack connections.

Everything amberwolf wrote I agree completely, and if your interpretation leads you to believe anything I've written is in conflict, go with him, or ask for clarification.

thanks very much for both of yours great advice, I have learned allot

the boost converters that you see on ebay for about 10 USD have allot of features CC CV etc, and with my BMS I have now, seems that set up would get me to 42 V and allow for balance charging?

I got the batteries for free, from a bike shop. Yes, they have been cycled allot but were at 3.2 or so and the rest were trashed. he has got more sets and I plan to series parallel them all up. but Im feeling they have been cycled stressed allot.