Smart BMS reporting over-voltage - how best to fix?

[bengee]

I have the 52V 14S7P from em3ev and according to the smart-BMS, I’ve over-charged the battery.

In an attempt to discharge, I’ve left the battery switched-on unconnected to any devices and the voltage has decreased slightly. I’m also seeing a voltage diff of 188mV between highest voltage and lowest voltage of all the cells.

What options do I have to fix this? Can I simply connect to controller and run the motor until the charge decreases to 90%?

Or have I done irreversible damage to the battery?

Thanks
Ben

 

[john61ct]

I have the 52V 14S7P from em3ev and according to the smart-BMS, I’ve over-charged the battery.

In an attempt to discharge, I’ve left the battery switched-on unconnected to any devices and the voltage has decreased slightly. I’m also seeing a voltage diff of 188mV between highest voltage and lowest voltage of all the cells.

What options do I have to fix this? Can I simply connect to controller and run the motor until the charge decreases to 90%?

Or have I done irreversible damage to the battery?

Thanks
Ben

The BMS is not doing its job if it allowed the charging to continue as that top cell got so far past the 4.2V max spec

unless you allowed the setpoint to go that high.

Yes any load to bring voltage down for now, bad for the pack to sit fully charged for long.

Likely you did not "damage" that weak cell, but didn't keep allowing that abuse, the weakest link dictates the performance of the pack as a whole.

You need to get the pack balanced, within say 20-50mV.

Getting the BMS to do the job may be a challenge, lower the balance-start setpoint so it gets going earlier

lower the charger's termination voltage until that weak cell stays well below 4.20V

either doing it manually or using a 14S active balancer will be quicker

thus less stressful, for good longevity. Reducing the current rate to say 0.2A will also help a bit if you can.

The goal should be to stop when that weakest cell hits say 4.15Vpc, and all the other cells are usually within say 50mV

Then rebalance once the delta goes higher

 

[bengee]

thanks John - that information is extremely helpful.

So I’ve just ordered a 14S Active Balancer, should be here on Thursday. In the mean time I’ll learn how to use it.

I’ve not changed any battery/charge settings myself, so perhaps the BMS is “faulty”.

What’s confusing me now is, the overcharged cell’s voltage is still increasing whilst not being charged. I’ve just connected to the Bluetooth BMS and the highest voltage is now up to 4295mV.

How can that be? I guess charge can move from cell to cell?

Thanks again
Ben

 

[fyfo]

Are you able to view and adjust your BMS settings via Bluetooth?

You definitely have a 52v charger?

 

[john61ct]

Sorry I was not clear, but you should not be letting things stay the way you are

every hour you are killing cycles off the back end

get that voltage down!!

Of course the BMS may be inaccurate in reading voltages

get a decent DMM and ammeter to verify devices with cheap components, never just trust chargers either.

I’ve not changed any battery/charge settings myself, so perhaps the BMS is “faulty”.

What’s confusing me now is, the overcharged cell’s voltage is still increasing whilst not being charged

That should **not** be happening, other way around, the BMS should be "resistance burning" off that weakest/highest cell.

so yes most likely faulty

unless the "start-balance" setpoint is too high, like accommodating those so-called High Voltage cell type supposedly designed to go to 4.35Vpc or so

Another rule is, never accept the default settings on an adjustable device

make sure you understand what each setting means, and what you want them to be

before putting the system into production.

 

[bengee]

Are you able to view and adjust your BMS settings via Bluetooth?

You definitely have a 52v charger?

None of the phone apps that can interface with my BMS allow me to change configurations. That’s not to say there isn’t one. It could be that I’ve just not found the correct app.

I’m not sure of the model of Smart BMS. I plan to take apart the case later today.

The charger I have came with the battery and is 58.8V with a High(100%)/Low(90%) switch, manufactured by STCharger. The most recent charge was using the High setting (accidentally).

 

[bengee]

Sorry I was not clear, but you should not be letting things stay the way you are

every hour you are killing cycles off the back end

get that voltage down!!

I’m currently without a controller so can’t use the motor to discharge. I’m waiting for a baserunner and a BAC855.

I’ve ordered a DMM and a wattmeter, which I will use from now on when charging/discharging.

Whilst I wait for those, I plan to wire together 5x12v100w halogens for discharging.

I’ve also ordered a DC variable PSU to help with balancing once I’ve got the voltage down.

At what point should I consider replacing the bad cells? The cells are Samsung 35E.

 

[E-HP]

I’m currently without a controller so can’t use the motor to discharge.

Do you have a light bulb or some other load?

 

[bengee]

Do you have a light bulb or some other load?

I have lightbulbs for 220-240v applications. Will the voltage not be too high for the 52V battery?

I also have a couple of 12v halogens, but I’ll need at least 4 (possibly 5) to meet the required voltage.

So I’ve ordered a pack of 100W 12V halogens that should get here by the weekend.

Sorry I’m very much a novice when it comes to electrics

 

[E-HP]

Do you have a light bulb or some other load?

I have lightbulbs for 220-240v applications. Will the voltage not be too high for the 52V battery?

Those should work.

 

[bengee]

I have just soldered this contraption. When I turned battery on there was a pop and faint burning smell coming from battery.



There are 5 100w 12v halogens in parallel, so should be at 60v?

When connected the battery, the battery displays “over current protection”. The battery still powers up normally when not connected.

Where did I go wrong here? Are the halogen wires not capable of the current rating of the battery?

 

[fechter]

Oh, no, the lamps need to be wired in series, not parallel. Rewire in series and try again. I use an old hair dryer but you can't use the switch or it will arc and self-destruct. Leave the hair dryer on and turn it off by unplugging it.

I would guess the voltage is not actually increasing. There is no mechanism for that. You really need to use a separate voltmeter and compare the voltages.

 

[bengee]

Oh, no, the lamps need to be wired in series, not parallel. Rewire in series and try again.

Oops! Thank you for putting me straight fechter. I've since built a new version, this time in series, and managed to get the battery down to 60% (according to the BMSapp).

In an attempt to get the BMS to balance the cells (balancing kicks in at 50%), I plugged in the charger on the Low setting (charges up to 90%). The BMSapp almost immediately adjusted its capacity reading to 0% (?). Once it got up to approx 28%, cell 14 was over-voltage (>4.2v) and the battery stopped charging... so I cannot use BMS to help with the balancing. Could the BMS be playing up?

Here's a screenshot of the readings in the app:


And to confuse me further (doesn't take much). The BMS (at some point during charging) reported cell 14 to be undervoltage, but now it's overvoltage.

Could the BMS be at fault here?

 

[bengee]

Time to try the Active Balancer. Since I'm new to all of this, I've "drawn" a diagram to indicate where I plan to solder each of the 15 wires and married it up to a photo of my battery.

Can anyone confirm if this is correct or not?




Other developments since my last post:

1. I have multimeter (just need to work out how to individually measure each of the parallel cell "blocks" - the voltage across the wires going to the BMS (e.g. B14 to B13) read between 4.05 and 4.08 for all 14 cells, but I don't know if this is what I need to be checking).

2. A grin satiator arrived today, which I will use as my charger from now on.

 

[fechter]

1. I have multimeter (just need to work out how to individually measure each of the parallel cell "blocks" - the voltage across the wires going to the BMS (e.g. B14 to B13) read between 4.05 and 4.08 for all 14 cells, but I don't know if this is what I need to be checking).

Yes, that is the right way to measure. On mine, I use a pair of solid strand wires to insert into the BMS plug.



This can also be used to manually charge or discharge a single cell group.

Your balancer diagram looks basically right. The typical numbering convention is to start at the negative end of the pack and work up to the positive end (opposite of what you are showing). Cell 14 should be at the positive end.

If your voltmeter measurements were all between 4.05 and 4.08, it would seem the BMS is not accurately reporting the voltage.
Since it's reporting strange numbers on the last cell, you may be suffering from "end cell effect". This happens when charge or discharge current is passing through the sensing wires going between the BMS and the cells. Make sure you have a separate wire going from the positive end of the pack to your charge/discharge connectors and keep the one going to the BMS only going to the BMS. Same thing on the negative end. The little sensing wires will have too much voltage drop at charge or discharge currents.

When not charging or discharging, this voltage drop won't happen, so the numbers should be accurate. If you still see high or low numbers for cell 14 when resting, I would suspect the BMS. Once you get the cells balanced, they tend to stay there unless you have bad cell or bad BMS.

 

[bengee]

Yes, that is the right way to measure. On mine, I use a pair of solid strand wires to insert into the BMS plug.

This can also be used to manually charge or discharge a single cell group.

Thanks for sharing this trick. What do you use to discharge a single cell group at ~4v? I suspect that in the next day or two I'll want to manually draw from the misbehaving cell group and see how this affects the BMS readings.

Your balancer diagram looks basically right. The typical numbering convention is to start at the negative end of the pack and work up to the positive end (opposite of what you are showing). Cell 14 should be at the positive end.

Thanks for confirming wiring and correcting the cell numbering. I had to flip the connector numbers and overlooked doing the same for the battery cells.

If your voltmeter measurements were all between 4.05 and 4.08, it would seem the BMS is not accurately reporting the voltage.

Right now the multimeter is showing all cells are currently at 3.84V except cell 13 (3.78V). The BMS is reporting cell 14 to be 3.95V.

Since it's reporting strange numbers on the last cell, you may be suffering from "end cell effect". This happens when charge or discharge current is passing through the sensing wires going between the BMS and the cells. Make sure you have a separate wire going from the positive end of the pack to your charge/discharge connectors and keep the one going to the BMS only going to the BMS. Same thing on the negative end. The little sensing wires will have too much voltage drop at charge or discharge currents.

I think the wiring on the battery is as you describe. Two thin gauge wires running to the BMS-header from each terminal and nowhere else. There's 10AWG connecting the pack's negative terminal to the BMS, then from the BMS to the battery's discharge connector. And lastly the positive terminal wire goes directly to the charge/discharge connectors.

When not charging or discharging, this voltage drop won't happen, so the numbers should be accurate. If you still see high or low numbers for cell 14 when resting, I would suspect the BMS. Once you get the cells balanced, they tend to stay there unless you have bad cell or bad BMS.

I plan to active balance comparing voltmeter and app readings, leave 24hrs then compare readings. Providing they are stable, I'll then charge to 58V using the satiator.

If the BMS is misreading cells, it will stop charging as soon as it reads a cell at 4.2V. Can I disconnect the BMS header connector (the 1 with 15 wires) and still charge the battery with the satiator via the battery's main charge port?

 

[bengee]

The Active Balancer didn’t do any balancing, as all the cells were within 50mV of each other. I believe it is working, but to be sure I need to push the cells out of balance. Not sure I have the equipment to drain a single cell group. Perhaps I could charge a single cell group instead, as I do have a dc power supply to hand.

When I put the battery on charge just now, the BMS was reporting 3.4V for cell 14. However my voltmeter shows me 3.83V (same as all other cells).

Is this enough evidence to categorically say the BMS is at fault, or are there other avenues I should explore first?

 

[E-HP]

The Active Balancer didn’t do any balancing, as all the cells were within 50mV of each other. I believe it is working, but to be sure I need to push the cells out of balance. Not sure I have the equipment to drain a single cell group. Perhaps I could charge a single cell group instead, as I do have a dc power supply to hand.

When I put the battery on charge just now, the BMS was reporting 3.4V for cell 14. However my voltmeter shows me 3.83V (same as all other cells).

Is this enough evidence to categorically say the BMS is at fault, or are there other avenues I should explore first?

Are the readings inconsistent for each cell, between the BMS and voltmeter, or just the one?

 

[fechter]

I would trust the multimeter more than the BMS.

If you unplug the BMS, it won't allow any charge current. You'd have to bypass the BMS (put the charger straight to the cells). But if you do this, you need to watch the cell voltages like a hawk during charging.

To discharge a single cell group, you could use those halogen bulbs in parallel again, or just a single one. I use a big resistor. The bulbs are actually nicer since you can see some glow.

It seems like your BMS is not measuring some of the cells correctly. Your wiring looks good. You may need to replace the BMS. I don't think those have software to calibrate the individual cell readings but that would be a nice feature.

 

[bengee]

Are the readings inconsistent for each cell, between the BMS and voltmeter, or just the one?

The readings are inconsistent for cells 13 (+1%) and 14 (+5%).

They have identical readings for the remaining 12 cells.

 

[bengee]

I would trust the multimeter more than the BMS.

If you unplug the BMS, it won't allow any charge current. You'd have to bypass the BMS (put the charger straight to the cells). But if you do this, you need to watch the cell voltages like a hawk during charging.

To discharge a single cell group, you could use those halogen bulbs in parallel again, or just a single one. I use a big resistor. The bulbs are actually nicer since you can see some glow.

It seems like your BMS is not measuring some of the cells correctly. Your wiring looks good. You may need to replace the BMS. I don't think those have software to calibrate the individual cell readings but that would be a nice feature.

Thank you for this! I think I better start looking for a new BMS, preferably one that's configurable.

 

[bengee]

I've searched for a replacement BMS, and I'm yet to find one small enough (<50mm width) to fit in the dedicated slot. The current faulty BMS is 47mm in width.

The battery has a max current rating of ~47A, so I've been looking for a 60A device. Or would a 45A rated BMS be suitable?

I've looked at JBD and Heltec devices. Neither seem to have any compatible in size.

I've scoured Aliexpress and Ebay so far, as I'm not aware of any specialist retailers.

Where else should I look?

 

[fechter]

The current rating should be higher than what your controller limits at. You don't want the BMS to trip on hard acceleration. Stated ratings tend to be exaggerated sometimes too, so a 60A unit might be good. If your controller only takes 30A, then a 45A rated one should be OK.

The market for these things changes constantly. I can't make a recommendation. The ones I've used in the past were "dumb" and needed a little modification for my application. eBay, Amazon, Alibaba, etc. are places to look.

 

[velias]

try Speedict.com

 

[john61ct]

try Speedict.com

Is that your product?

Heard good things