Question about typical BMS charging FET Vds specification

[Sweet Tatorman]

I have a good understanding of how common LiFePO4 BMS work. This is not a question about that. It is also not a question about what you would consider good practice if you were designing a BMS. My question is what have you actually seen done with BMS that you have examined. Background: in a single port BMS both the charging and discharging FETs are required to carry the same current. It appears that the voltage that the charging and discharging FETs must stand off differs. In the case of the discharge FETs it is straight forward. The FETs in off state must withstand the full battery voltage is there is any external load path on it's terminals. So, for example, at nominal 48V battery with a BMS designed for a 48V battery (e.g., not four 12V batteries seriesed together) would need a minimum Vds rating of 60V as the battery voltage can be no higher than this. In the case of the charging FETs their standoff requirement appears different. In the case of a 48V battery without low temperature protection the stand off voltage only needs to be the difference between the highest allowed charging voltage and the battery voltage when the BMS determines to cut off charging on either total voltage or individual cell overvoltage limit. This typically would only be a few volts. In the case of a 48V battery with low temperature protection, one can conjecture that the actual battery voltage is down at an allowable minimum, of say, 40V in which case the charging FETs would need to stand off the difference between maximum charging voltage and 40V which would be 20V. I have several questions based on your actual observations of commonly used commercially produced BMS. Since for a given Ron rating, a FET with a lower Vds rating is generally less expensive, have you observed lower rated Vds FETs used in the charging FETs vs the discharging FETs? Have you observed an equal number of FETs used in the charging and discharging groups? This would be expected if the number of FETs is being determined by the individual FET package current limitation which is typically 30A (I think determined by die attach bond wire limitation) for the commonly used TO-252 package if a very low Ron part. TIA for all learned responses.

 

[LakeHouse]

Just a single case, but this video appears to show symmetrical and identical FETs in the JK BMS for charge and discharge. It may just be a matter of convenience and manufacturing ease to make them symmetric: stock and install just one type of FET on a given BMS rather than multiples.

As for the case of a 12V battery in a 48V setup, I follow your logic and I can't come up with an obvious scenario where a single set of FETs would have to interrupt 60V. I initially thought about a case where three BMS have cut off and one remains: on the face of it, you would think that BMS would see ~60V, but it actually doesn't. Rather it sees zero volts because the circuit is open elsewhere. And similarly, the first BMS to cut off still only sees a maximum of 15V (maybe a little more, but certainly not 60V) because the other batteries must still be operating and keeping the circuit closed.
I'm very much open to being proven wrong on this though. It does seem like each BMS should have to be capable of interrupting 60V. Maybe a case where one or more of the other batteries in series fail to a short?

 

[Sweet Tatorman]

Actually, the more I think about it the more I think I had the four 12V series scenario wrong. I think the discharge FET would see the full 48V. I will edit out that particular question.

 

[Hedges]

If a 12V battery BMS with FET on negative terminal disconnects, the battery terminal remains a 0V (relative to Earth, assuming cell Bat- is earth grounded but it might be floating. Just use Earth as a stable point of reference).
Other terminal of BMS FET gets pulled up to +12V or so by resistance of load.

If 4x batteries in series, each with its own FET on its 12V Bat-, then the BMS sees +48V or so.

Fuse or disconnect switch/breaker on positive wire, if fuse blows or switch opens, resistance of load pulls it down to 0V while input from battery remains at +48V.

I think series connecting batteries, if a reverse-polarity diode was placed across each 12V module+BMS, then none would ever experience 48V, clipped to the 12V of the pack itself. If load continued to draw current at 36V, diode has to be heatsinked to handle its power dissipation and 1V or 2V drop. Just like PV panel bypass diodes. Most 48V inverters would shut off at 36V due to low battery voltage.

Preference is to have 48V BMS. But especially for passive balancer, ought to be possible to work with multiple BMS. Like those Kohler 8s 100 Ah CATL modules we're seeing. Each 8s pack has BMS (without FET or contactor to disconnect), and master control unit talks to multiple series connected modules, has contactors for entire HV battery.

 

[Sweet Tatorman]

I think series connecting batteries, if a reverse-polarity diode was placed across each 12V module+BMS, then none would ever experience 48V, clipped to the 12V of the pack itself. If load continued to draw current at 36V, diode has to be heatsinked to handle its power dissipation and 1V or 2V drop. Just like PV panel bypass diodes. Most 48V inverters would shut off at 36V due to low battery voltage.

I occur with this thought experiment.

 

[Hedges]

I occur with this thought experiment.

Did I just manifest you into existence?