How does a BMS board ACTUALLY work?

[Peter Sternersson]

My first electric bike build is coming to an end, I got my BMS Battery Alloy Shell charger yesterday along with a "battery management system"

http://www.bmsbattery.com/bmspcm/323-17s26s-24a-max-discharge-current-bms.html

I connected everything up according to the scheme found where I bought my batteries (they have the exact same BMS, but a better wiring diagram).

http://www.ev-power.eu/docs/pdf/GWL-SBM-Product-Specifications.pdf

Hooked the charger up and it started charging for a minute or two. My cell voltage was around 3,3V when I started (53V total), and the charger quickly reached 57,6V and shut off.

After a minute or so it turned on for 5sec again...

When I took the pack of the charger it read 53,X Volts still. Perhaps this is normal, I understand that the voltage will drop some as soon as you take it of charge, but 5V??

How exactly does one of these BMS boards work?
I was under the impression that the MOSFETS on the board "portioned" out the Amps to the lower cells or something?

Or does the charging stop when one cell reaches full voltage, and then the effect resistors burn that cell down in voltage and charge everything up again?

The later whould mean that you cycle that one cell much more then the others as that cell will also take charge much faster than the others as well?

Also, how does the low voltage cut off work? If one bad cell reaches the cut off way before everyone else, the rest of the capacity in the battery can not be used??

 

[dnmun]

the best way to understand it is to measure the cell voltages while the pack is charging. the charger turns off because the BMS has detected High Voltage Cutoff on one of the cells and so it shuts down the charger to allow the high cell to drain off the excess charge through the shunt resistor.

that is the primary function of the BMS, to allow the cells to all fill with charge at their own rate as they age and the internal resistance changes so they don't charge up at the same rate any more. this is balancing and is done through the shunt transistors diverting charging current around the cell as the cell fills up and can no longer accept more charge.

by watching each of the cell voltages you will be able to identify the cells which reach the HVC first and which ones reach the LVC first. this characterizes the limits of the battery capacity.

if the battery is well balanced all the cells will charge up together and very little time is spent in balancing. when the pack is new, all the cells are far out of balance and it takes a long time for it to balance naturally under the charger and BMS.

the low voltage cutoff is where the BMS detects that one of the cells has dropped below 2.1V (or 2V for some BMSs). when any cell reaches the LVC then the BMS turns off the output mosfets so the battery is now 'turned off' and you have to recharge it to cause it to work again.

the other function the BMS provides is to protect the battery from pushing more current than the battery is capable of. that is the 50A limit you read about. when the current reaches 50A for a few milliseconds the BMS will shut off the output mosfets and you have to stop and disconnect then reconnect the BMS to reset it.

 

[dogman dan]

Your battery may need a few lightly discharged cycles to perk up all the cells. At that point, you might see it balance sooner and perhaps have a higher voltage when taken off the charger.

Meanwhile, the charger cycles on and off periodicaly trying to get all the cells topped up. When off, the bms is lowering the highest cells voltage. It can take quite some time, particularly with a fresh cell that is still a bit sluggish.

Charge it, ride around the block, and put it back on the charger. Repeat this about 5 times, leaving the battery on the charger overnight for a few days. Soon you should see a much quicker balancing, and much less need to leave on the charger overnight. Some times a new lifepo4 cell just needs a few cycles to develop it's full capacity.

 

[Peter Sternersson]

Thank you for your answear guys, I just love the internet when I can sit in Sweden and get direct support from the US

I contacted the battery saleman, and he said that I should start by charging each and every cell seperatly to 4,0V, then the balancing whould be much faster.

Also, he explained that when one cell reaches it´s max voltage, the charge current to that cell goes to the resistors instead.
But in my mind this whould mean that the charger should keep on going, and not stop as it did yesterday.

However, when I get home today I will charge each cell to 4V and go for a ride, I clocked it at 45km/h yesterday at only 53V

 

[dnmun]

not 4V. charge each cell to only 3.65V. if you put the pack on the charger and measure each one while it is charging, then ask advice, it will be better advice and you won't be so far in the dark. keep it on the charger for now even though it appears that it has stopped charging. keep track of how long between intervals that the charger comes back on.

 

[Peter Sternersson]

The charge level of 4V I got from the battery supplier, so I hope that´s correct...?
I have emailed them again just to be sure.

My plan was to use my faulty Turnigy Accucel 8 and charge in LiIo-mode to 4,1V and monitor and shut it of at 4,0V.
But it whould be more easy to set it to LiFe so it stops at 3,6V, that way I dont need to monitor it all the time...

I´ll post again later this evening... (its 3 in the afternoon here)

 

[dnmun]

if it is lifepo4 then the cells are fully charged when the cell potential reaches 3.65V. we have all accidentally pushed cells to 4V or over but that usually provokes panic. 3.65V is charged.

 

[Peter Sternersson]

I have connected the BMS Alloy charger nog and fully charged the battery, it went much higher then last time, I think it´s something like 58V when I take it off the charger.

I´m waiting for some connectors from a friend of mine so I can measure each cell, then I will see if the balance board has made its job.

Has anyone had problems with this type of balance board?

 

[dnmun]

you can just leave it on the charger, 24/7 for awhile.

 

[dogman dan]

Assuming this is lifepo4, 3.7v per cell. As it said on the spec sheet when I looked at that. Some charging and bms systems will go as high as 3.75v, but anything above 3.65 is not usable power. Definitely not 4v.

As soon as you start riding, it will quickly drop to 3.5v or less so expect to see about 55-56v as soon as you start, and about 48v is discharged.

Sounds like it's perking up fine, just charge it overnight for awhile. If more convenient, leave it on the charger every night. Anytime you unbalance it, like a fully 100% discharge, it's good to leave it charging a long time to completely balance again.

 

[999zip999]

No need for 4.0v as 3.5v is charged and 3.65v is tops.. There not much useful ah up there. Leave it on the charger.

 

[Peter Sternersson]

The first serious ride taken on the bike this morning, an average of 35km/h, and a 10km trip using 4Ah, going to work.
I love the feeling of Superman strong legs when you hit the accelerator

However, I must change the previous statement on the charging, when it´s fully charged and I actually disconnect it from the charger, it reads 53,xV. When the charger is connected (and not charging) it reads 58,6V or something.
Is it normal that a LiFePO4 battery drops that much when taken off the charger?

Also, the balancing by the BMS board seems to kind of work.
3,3xV on all the cells except one that read 3,8V.
The BMS should cut that cell off at no more than 3,65V according to specs but...

I´ll throw in a picture I took down stairs in the lobby as well.

Specs are

KTM Avento 27 light bicycle
Ebike-kit.com wheel and controller, 48V 20A (peaks at 22A)
15Ah LiFePO4 cells from ev-power.eu in 16S.
BMS Battery BMS board, 16S 30A

The only real problems I have had has been with the batteries and their charging, and also I had to weld some reinforcements to the frame since it was a light weight aluminium frame.
Now I just need to ride it enough to cover the cost

It´s not 100% finished, I have a lid for the batteri box where a main switch will be mounted. Also I got some "Speakon" connectors yesterday that I will use as a charge/discharge port.
Some kind of wrapping of the cables have to be made as well.
But other then that I´m very pleased how it came out...

 

[dogman dan]

53v is a bit low on the normal range. 3.3v. My 48v lifepo4 battery only holds about 54v now,about 3.4v after some abuse. 3.5v should be the final charge, after excess surface charge is dissipated. Your charger is going to 58v, which is normal. A slight overcharge helps it balance.

It's possible that while your charger is set to 3.63v, your bms seems to be set to only 3.3v, and mayb be drainging down the cells to that voltage. The bms is supposed to be set to 3.6v according to the spec sheet.

One way to test it would be to disconnect the bms, and try charging without it and see if the cells hold 3.6v then.

 

[fivari]

Reading this post I have an additional question.

When I charge my battery pack, all cells are at 3.6V except the first series. That seems to remain lower: 3.35, 3.4 max. I've tried to SLOWLY charge this series separately up to 3.6V. However, the number of Ah I put in that series are very high and the BMS stays warm when I try to do that. My assumption is that the BMS is drawing energy out of this first series to bring down the voltage. First I tought this was for feeding the functioning of the BMS, but reading this post, I guess that the BMS of this first series is calibrated to a voltage that is too low.

==> Is this a plausible explanation of what is happening
==> How to correct this? I suppose I have to replace/modify some resistors on the BMS

 

[dnmun]

put the battery on the charger and measure the cell voltages while it is charging and post them up.

when you say the BMS is hot, where is it hot? is it on the shunt resistors?

 

[Peter Sternersson]

This doesn´t answear the above question regarding BMS, but I found it very interesting.
I got this by email from ev-power.eu where I bought my LiFePO4 cells.

"
LiFePO4 cells have a bit complicated "maximum and minimum" voltage levels.

Initial full charge is 4V (teoretical maximum is about 4,15V).
Optimal balanced charge (the voltage you will charge it to
periodicaly) is 3,65V - 3,70V
Minimum Voltage is about 2,5V, this would indicate ethier damaged or nearly damaged cell requiring special care (should be slowly charged to 4V again, as single cell).

These voltages are valid for unloaded cell, depending on situation, cell under load can go as low as 2V without being damaged.

Basically LiFePO4 behave like : Charge to 3,65V and balance, once charger is plugged off the voltage will drop to 3,2V. When connected to load it usually drops to 3V. This is caused by chemical reactions of the battery.
"

This whould indicate that, atleast my batteries, are ok as they behave...

 

[999zip999]

The battery. What is the battery you talk like it is a lifepo4 3.65v and give voltage numbers for lipo 4.0v ? The battery is a what from where ?

 

[Peter Sternersson]

http://www.ev-power.eu/LiFePO4-small-cells/Lithium-Cylindrical-Cell-LiFePO4-3-2V-15Ah-48Wh.html

Just got home from work. The BMS low voltage cut off engaged. Measured the cells just now, and two cells where at 2,6V. So the BMS worked as it should at least.

The question now is why two cells has such low voltage. The others had 3,24-3,26V, so thats nice and even.

My charge cycles this far are as follows.

From the box -
- charged to full - (probably not balanced because I thought it was finished when the charger turned green)
- 3,7Ah drawn-
- fully charged- (left it on the charger for several hours)
- 4,0Ah drawn-
- fully charged- (cells measured 3,3xV all but one that was 3,8V)
- 7,0Ah drawn- This is where the BMS cut off.

Can I expect the two lower cells to wake up, or should I replace them? (I have spares)

 

[dak664]

3.3 volts says little about the state of charge of LiFePO4. If the low voltage disconnect occurs after a 7 amp-hour discharge then that cell is 3 amp-hours low (in a 10AH pack). Your BMS equalizing bypass is probably 100-200 milliamps, that means 15-30 hours of charge time if the charger does not cycle.

If the charger cycles that time will go up. The charger typically tapers the rate as the pack approaches the target voltage, however when a couple of cells are low the charger could put out an amp or two which will overwhelm the bypass shunts and cause immediate shutdown. Adding 10 ohms or so in series will reduce the charging current to a level the shunts can handle continuously (but it will still take 10-30 hours of charge time).

To do fast equalizing, use a single cell charger on the low cells.

 

[dnmun]

if you put a 10 ohm resistor in a circuit flowing 4A then the heat produced by the resistor would be ((4)^2)x10=160watts.

 

[dak664]

True, but the charger would have to increase its output by 40 volts to push 4 amps through the added 10 ohms. It depends on the charger, most have a maximum voltage around 4 times the number of cells for natural taper charging. 100 milliamps through 10 ohms raises the charging voltage by 1 volt which would compensate for three cells that are still 0.3 volts below maximum and hopefully reduce the charger current to something that the BMS can continuously handle

An ammeter in series is useful for monitoring, also bear in mind that all but one shunt may be in bypass mode so that the (say) 50 volts at 200 milliamps may be generating 10 watts of heat internal to the battery pack, with only a tenth of the energy being used to charge the single low cell. All the others are being overvoltaged for tens of hours which can't be good for them. Safer to balance for several days at 100 milliamps - it only needs to be done once a year or so.

 

[dogman dan]

Not even my abused ping cells drop to that low a voltage when unplugged from the charger. Crap defective cells might. Or maybe ruined cells that were charged to 4v.

I'm no expert, but this guy is misinformed. But it might be normal for the crap he's selling. I think you got sold something crap, perhaps the cells themselves. Good cells won't self discharge more than to 3.5v. Really good cells will hold 3.65 for awhile, perhaps a few hours.

 

[The Mighty Volt]

Not even my abused ping cells drop to that low a voltage when unplugged from the charger. Crap defective cells might. Or maybe ruined cells that were charged to 4v.

I'm no expert, but this guy is misinformed. But it might be normal for the crap he's selling. I think you got sold something crap, perhaps the cells themselves. Good cells won't self discharge more than to 3.5v. Really good cells will hold 3.65 for awhile, perhaps a few hours.

Wow...are you sure of this??? I've charged {CC-CV} cells to 3.6v and they drop down to 3.3X v and then stay there for weeks.

When tested, they give good capacity and discharge rates.

 

[scriewy]

for now it just sounds like his 2 cells are very out of balance.

till he doesnt singly charge them and then charges as a hole pack a few times to see how it acts or test capacity with smart charger, cant say if the cells are bad or faulty bms or out of balance cells.

 

[Peter Sternersson]

I´ve done some more riding now, and got 8Ah out of the pack before the BMS shut off, due to the same two cells as before was low.

So I charged full again, I know that the two low cells have 8Ah in them, now I have charged them seperatly with an additional 7Ah each.

It´s getting late and cold here now, but maybe I´ll go out for a spin to see how much I can get out of it now.

It´s a bit sad that dogman so easily jump to the conclusion that everyhing I bought is crappy shit not worth the name...
If you don´t have any more constructive things than that to write...please don´t...