Regen current through BMS

Batteries, Chargers, and Battery Management Systems.
Post Reply
kdog   10 kW

10 kW
Posts: 694
Joined: Feb 02 2014 9:22pm
Location: hobart, tasmania

Regen current through BMS

Post by kdog » Jul 15 2014 6:52pm

Hi
I have a question about regen current through a BMS. I have a 36v headway pack with BMS protection, but the controller HVL is set for 48v. I live at the top of a hill and often regen brake on a full charge as I'm descending. I assumed that the BMS would shut down regen current if cell voltages got too high. Ive discovered two cells that have leaking electrolyte and figure that maybe the bms is not protecting cells from high regen induced high voltages.
Should it? Or do they only protect when wall charging? I thought that regen and wall charge currents followed the same path through the BMS.

User avatar
amberwolf   100 GW

100 GW
Posts: 31723
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Regen current through BMS

Post by amberwolf » Jul 15 2014 9:41pm

The first thing to ask is:

When you are regen braking and the battery is full, does the braking action cut off at all? If not, then for whatever reason the BMS is definitely not cutting off (whether it "should" or not might be another story, depending on design/hookup).



As for whether it will cut off current:

It depends on how the BMS works and is wired. Some BMS's charge thru a separate port (and FETs) than they discharge thru, and in that case there won't be any cutoff of regen charging currnet, because the discharge port the regen comes in thru probably only cuts off under cell or pack LVC conditions, not HVC.

If charge and discharge are thru the same port/FETs, then either HVC or LVC should cut off the FETs, so regen current should also be cut off.

However, if your BMS does not disconnect under cell-level HVC but only pack-level HVC (which some seem to do based on failures people have seen), then until the whole pack reaches HVC it wouldn't cut regen current off, and any already-high cells would then get overcharged.


If your BMS is supposed to do cell-level HVC and cells are indeed being overcharged by the regen, and it also uses the same port for charge and discharge, then it also could theoretically overcharge off the wall charger, too, if that charger was high enough voltage to force current thru the pack past it's normal cutoff, because it means the BMS is not working correctly.

I'd guess the last is the most likely problem--either a design flaw or a failure in the BMS itself.


It could also be that the FETs have failed, which since that's normally a short, would allow current to flow whether or not the BMS is trying to turn them off or not (meaning the rest of the BMS could be working properly, but the FETs can't turn off anymore).

That oculd happen from overheating, or from higher voltage across the FETs than they are rated for, or a number of other reasons.


You can test this a couple of ways.

One is by using an ohmmeter and measuring across the source and drain legs of the charge/discharge FETs, when you know the BMS should have cut off the current. If the reading is low ohms, then it's not working. Then disconnecting the BMS from the battery completely and remeasuring, and if the reading is STILL low ohms, then the FETs are probably shorted and will need to be replaced. remove the FETs and remeasure, and if it's now high ohms hten repalce with new FETs. If it's still Low ohms you have some other problem inside the BMS or PCB, or are measuring in the wrong places.

If the reading of the first measurement is high ohms then it *has* cut off the current. If the reading of hte second measurement is high ohms then the FETs are ok, but just not being told what to do correctly by the BMS when it's powered up. For that you can start measuring the gate of the FETs, relative ot battery negative, to see what they'r ebeing told to do and when.



Another is by monitoring the voltage on either side of the FETs (source ot drain or drain to source) during charging, as well as the cell voltages, and when the BMS is supposed to be shutting off during a cell or pack level HVC event, if the voltage is the same on either side of the FETs then it's not cutting off--if the voltage becomes different after the event than before, they are working. If not, they're either shorted or not bieng controlled ocrrectly by the BMS.

kdog   10 kW

10 kW
Posts: 694
Joined: Feb 02 2014 9:22pm
Location: hobart, tasmania

Re: Regen current through BMS

Post by kdog » Jul 16 2014 12:08am

Thanks AW for that comprehensive reply!!
BMS has different port and FETs for charging and discharging so I suspect that's en if it was working OK it might not have HVL through discharge leads. I'll do some testing with the info you provided. Ebrakes never cut out even when Bms was new As for regen when pack is full, I charge at home then have a 200mt/5kms drop into town. Might have to change some habits/ equipment
Thanks again -K

User avatar
amberwolf   100 GW

100 GW
Posts: 31723
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Regen current through BMS

Post by amberwolf » Jul 18 2014 3:40pm

With separate ports for charge/discharge, you could get or make some "ideal diodes" (there are threads about this) or use regular high-power diode to block regen current from reaching the pack thru discharge port, and another diode to prevent discharge from the charge port, and wire them in parallel. Then you will only be charging thru the charge port, be it from the charger or from regen.

Otherwise you might be safer to disable regen, or else ensure you don't use it till you know you have run down the pack to some particular state of charge.

simpes   10 µW

10 µW
Posts: 6
Joined: Apr 24 2012 9:04am

Re: Regen current through BMS

Post by simpes » Aug 31 2015 4:29am

Hello, came across this thread while doing research on BMS. I'm thinking of getting this BMS:
http://www.bestechpower.com/296v8spcmbm ... 223V1.html
And this charger:
http://quawintech.en.alibaba.com/produc ... 8397337524

I'm building a system 8s4p using Samsung 25R to drive a sensored motor (60A max) for an electric skateboard. My controller is capable of regen (I intend to use regen). I would like to ask:

Should my BMS charge and discharge be configured to be the same port or different port? (my email correspondence with the company ask me which configuration I would like to have, so I would like to know which configuration is better, since the concern with regenerative braking is that with full charge + regen, the batteries might get overcharged.)

My understanding is that I should get the configuration where BMS charge and discharge shoud be of the same port, am I right? In this case will I have to take measures to cut off current to the motor controller while the system is charging?

Thanks a lot!

dnmun   100 GW

100 GW
Posts: 16190
Joined: Jun 09 2008 1:32pm
Location: portland, or and loveland, co

Re: Regen current through BMS

Post by dnmun » Aug 31 2015 7:57am

you should not engage regen if the battery is already fully charged. obviously it is not stopping the overcharging if the regen current was not stopped by the BMS.

some BMS will block the regen current when it causes one cell to exceed the HVC. if your BMS has a separate port for charging and uses a separate set of mosfets to control the charging from the discharge then it cannot block regen current that flows back into the battery through the output mosfets.

look for a BMS like the D167 which has two sets of mosfets with the drains back to back and C- and P- at the same spot instead of the type with a separate set of charging mosfets.

joshuahuang   10 mW

10 mW
Posts: 26
Joined: Dec 28 2018 3:24am

Re: Regen current through BMS

Post by joshuahuang » Jan 30 2020 4:20pm

amberwolf wrote:
Jul 15 2014 9:41pm
The first thing to ask is:

When you are regen braking and the battery is full, does the braking action cut off at all? If not, then for whatever reason the BMS is definitely not cutting off (whether it "should" or not might be another story, depending on design/hookup).



As for whether it will cut off current:

It depends on how the BMS works and is wired. Some BMS's charge thru a separate port (and FETs) than they discharge thru, and in that case there won't be any cutoff of regen charging currnet, because the discharge port the regen comes in thru probably only cuts off under cell or pack LVC conditions, not HVC.

If charge and discharge are thru the same port/FETs, then either HVC or LVC should cut off the FETs, so regen current should also be cut off.

However, if your BMS does not disconnect under cell-level HVC but only pack-level HVC (which some seem to do based on failures people have seen), then until the whole pack reaches HVC it wouldn't cut regen current off, and any already-high cells would then get overcharged.


If your BMS is supposed to do cell-level HVC and cells are indeed being overcharged by the regen, and it also uses the same port for charge and discharge, then it also could theoretically overcharge off the wall charger, too, if that charger was high enough voltage to force current thru the pack past it's normal cutoff, because it means the BMS is not working correctly.

I'd guess the last is the most likely problem--either a design flaw or a failure in the BMS itself.


It could also be that the FETs have failed, which since that's normally a short, would allow current to flow whether or not the BMS is trying to turn them off or not (meaning the rest of the BMS could be working properly, but the FETs can't turn off anymore).

That oculd happen from overheating, or from higher voltage across the FETs than they are rated for, or a number of other reasons.


You can test this a couple of ways.

One is by using an ohmmeter and measuring across the source and drain legs of the charge/discharge FETs, when you know the BMS should have cut off the current. If the reading is low ohms, then it's not working. Then disconnecting the BMS from the battery completely and remeasuring, and if the reading is STILL low ohms, then the FETs are probably shorted and will need to be replaced. remove the FETs and remeasure, and if it's now high ohms hten repalce with new FETs. If it's still Low ohms you have some other problem inside the BMS or PCB, or are measuring in the wrong places.

If the reading of the first measurement is high ohms then it *has* cut off the current. If the reading of hte second measurement is high ohms then the FETs are ok, but just not being told what to do correctly by the BMS when it's powered up. For that you can start measuring the gate of the FETs, relative ot battery negative, to see what they'r ebeing told to do and when.



Another is by monitoring the voltage on either side of the FETs (source ot drain or drain to source) during charging, as well as the cell voltages, and when the BMS is supposed to be shutting off during a cell or pack level HVC event, if the voltage is the same on either side of the FETs then it's not cutting off--if the voltage becomes different after the event than before, they are working. If not, they're either shorted or not bieng controlled ocrrectly by the BMS.
how would you use regen braking in a separate port bms? Do you split the ESC negative wires to both the discharge port of the BMS and the charge port?

User avatar
amberwolf   100 GW

100 GW
Posts: 31723
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Regen current through BMS

Post by amberwolf » Jan 30 2020 5:28pm

really the only reason to do that is if you have high regen currents that you need the bms to protect the cells against, or high voltage from the regen that might overcharge the cells, etc. typically cells will accept 0.5c to 1c charge in bursts just fine. very short spikes of 2c and up may be fine, too. (c = the capacity in ah of the pack, so 0.5c on an 8ah pack is 4amps. 1c is 8amps. 2c is 16amps, etc.)

if you do need this to happen, you can't just connect the two ports, or else you are bypassing the entire bms protection both charge and discharge (because shorted together, the bms will still supply power to controller even if the battery is below lvc thru the charge port, and will still accpet power from the charger (or regen) even if the battery is ful (above hvc) thru the discharge port.

you will need diodes on the ports to prevent inflow to the discharge port, and outflow from the charge port, between the ports and the y-connection point. the catch is the diodes drop voltage, and thus waste power as heat.

so if you don't require this specific form of protection during regen, i wouldn't worry about it and just leave the ports separate.

Post Reply