[Elithion]

Did you blow up your BMS and don't know how? It may be because you forgot to disconnect all loads from the battery (including a charger) beforehand. When you open a link between two adjacent batteries, without first disconnecting all the loads, the entire pack voltage appears across the gap, frying your BMS.
This video shows why that is.

[ndplume]

I watched your video, thanks for it.

To clarify : My charger is hard wired in, I think most are. So we'll say that is the normal state.

If you open your breaker, NO BMS-Fry because there should NOT be a BMS board across the breaker.

If you open the breaker 1st, then undo a battery link, NO-BMS fry since the breaker is open, even if a charger is connected.

You only get BMS-fry if you undo a battery link that has a BMS board across it while your charger is connected (as normal).

Correct?

[Elithion]

Originally Posted by ndplume

My charger is hard wired in

Do you mean that you don't have a safety disconnect? OUCH!

Originally Posted by ndplume

You only get BMS-fry if you undo a battery link that has a BMS board across it while your charger is connected (as normal).

Not "only", but, yes, that is correct.

[ricklearned]

Originally Posted by Elithion

Did you blow up your BMS and don't know how? It may be because you forgot to disconnect all loads from the battery (including a charger) beforehand. When you open a link between two adjacent batteries, without first disconnecting all the loads, the entire pack voltage appears across the gap, frying your BMS.
This video shows why that is.

Thanks, I had asked that question on another thread and got no answer. I did read your book and I suspect the answer was in there somewhere but I could not readily find it. I am working on the design of my 2nd phase pack. I currently have 32 90AHr batteries in two boxes under the seats of my VW super beetle. I monitor them with a cellog board but it is limited to 32 cells. I want to add 40 more cells to get a 2p36s pack. At that point I might consider a BMS. I think I understand that the best configuration would be to have the cells paralleled in pairs so that I will only need 36 BMS boards. The available locations mean I will be splitting that pack into at least 3 groups and possibly 4 depending on how many I can get up front.

I was considering a safety disconnect between the front part of my pack and the back part, but if I understand your video if I were to do that I would blow up my BMS. So do I have to find an alternative way to disconnect my pack at one end or the other in order to not fry a BMS? I am on the fence about installing a BMS anyway, and I think a mid pack disconnect has certain advantages so how do I find a workaround, or does the scale of positives vs negatives tip toward no BMS, given my particular priorities?

[Elithion]

Originally Posted by ricklearned

I was considering a safety disconnect between the front part of my pack and the back part, but if I understand your video if I were to do that I would blow up my BMS.

There is a difference between a mid-pack breakout, and working on the battery with loads still connected.
- A mid-pack breakout is completely between cells; it is OK if the BMS on one side of the breakout is isolated from the BMS on the other side of the breakout
- working on the battery with loads still connected is bad regardless of which BMS: you may disconnect a link and, if the BMS is still connected to the link (not to the battery terminal) if will suddenly see the full pack voltage (reversed) where it is only expecting a cell voltage


There can be a breakout mid-pack, depending on the BMS. To the best of my understanding:

• Cannot have a breakout within the battery pack:
  • PCM (protectors)
  • Elektromotus EMUS
  • Thunderstruck
  • PakTrakr
  • Agni
  • Clayton C
  • Lithium balance iBMS
• Can have a breakout but only at specific multiples of cells:
  • Elite Power EPS: 4 cells
  • Flux power: 4 cells
  • I + ME ACTIA: <= 7 cells
  • CellLog S8: <= 8 cells
  • Lithium Balance S-BMS: <= 8 cells
  • Guantuo: <= 10 cells
  • REAP: <= 14 cells
  • Kokam N-tech: <= 14 cells
  • REC: <= 14 cells
  • Ligoo/Anhui/EVLithium: <= 16 cells or <= 12 cells
  • Orion BMS: <= 36 cells
  • GWL Power <= 64 cells
• Can have a breakout at any point between banks:
  • Elithion Lithiumate Lite and Pro
• Can have a breakout at any point between cells:
  • EVLithium MinBMS
  • CleanPower MiniBMS
  • EVPower
  • Pacific EV
  • Genasun

[etischer]

Originally Posted by Elithion

Do you mean that you don't have a safety disconnect? OUCH!

Disconnecting the pack from the charger under load will fry the charger, I too have my charger wired without a pack disconnect. I definitely would have fried a BMS board had I not disconnected a link in the pack before reconnecting my charger.

[Elithion]

Originally Posted by etischer

Disconnecting the pack from the charger under load will fry the charger, I too have my charger wired without a pack disconnect. I definitely would have fried a BMS board had I not disconnected a link in the pack before reconnecting my charger.

As far as I know, the Manzanita Micro charger is the only one that gets fried when disconnected under load.

The correct procedure, in all cases, is:

1. Stop the battery current (turn off the charger and the ignition)
2. Disconnect the safety disconnect
3. Work on the battery
4. Reconnect the safety disconnect

And, if the safety disconnect is mid-pack, use a BMS that has isolated sections, and place the breakout between two isolated sections.

[Tesseract]

Originally Posted by Elithion

...
- working on the battery with loads still connected is bad regardless of which BMS: you may disconnect a link and, if the BMS is still connected to the link (not to the battery terminal) if will suddenly see the full pack voltage (reversed) where it is only expecting a cell voltage..

Thanks for posting this.

It is possible to protect each BMS against this sort of abuse (within reason) by putting a fast 600V rectifier in series with the shunting transistor (if present). For the voltage monitor, using a differential/instrumentation amplifier circuit with a common mode input voltage range exceeding the pack voltage will suffice. Of course, these tricks - especially the latter - tend to increase the cost of the BMS or inevitably degrade accuracy, but in all cases it's cheaper than a battery fire.

[Tesseract]

I found the screen capture of the SPICE simulation I did to explain this to someone else:

[major]

Hi Davide,

First; thanks for taking the time to address questions on this forum and for compiling and composing those "white papers" and lists available on your web site.

Second; maybe a bit off topic; maybe not. Have you done any work, given any thought to, or have recommendation for BMS for hot swap battery systems? Maybe not hot swap in the true sense but the application in particular is our competition EV go-kart. We use Lithium batteries and do a pit stop during the race where the student team removes the battery pack and installs a fresh battery. The kart carries about 4 kWh and they are able to swap it in 7 seconds. The BMS gives us fits and we are not satisfied with our current system, which is pretty much homemade, but have not been able to find a suitable system to buy.

Presently we use three 48V batteries in parallel on the kart, two in the left side package and one in the right side package. The package is connected via an Anderson connector. All charging is done off board.

We need a BMS with cell modules to stay with the battery and quick plug into a "mother" board either on the kart or on the charger. I guess we could have the mother (or control) board on each battery, but we have 10 to 12 of those. The parts must be robust as the races get quite physical if you know what I mean.

This problem has me thinking that some type of RF or bluetooth communication from the cell BMS would be a nice solution. Given any thought to that?

Regards,

major