MOSFET in the bms getting hot

[ketoneenergies]

when high current is drained, mosfet in the bms is getting very hot(nearly 100° C). Any reasons?

 

[daromer]

To high current for the bms. A mosfet have a resistance.

 

[ketoneenergies]

bms is designed for 50A, but when the current draw increase above 5A, temperature starts to rise. Actual resistance of the pack is 63 mΩ but bms resistance is 59mΩ. Why bms resistance so high?

 

[Korishan]

Cheap BMS with cheap FET and/or fake labeled FET (ie, it's supossed to be 50A FET, but really only a 30A or such)

 

[daromer]

As Korishan said. That sounds like a cheap Ebike BMS

 

[Bubba]

when high current is drained, mosfet in the bms is getting very hot(nearly 100° C). Any reasons?

What BMS are you using?

 

[ketoneenergies]

a Chinese one, but i have 6 other which works fine at 50A. Any idea to fix it?

 

[daromer]

Change BMS. Thats the easiest solution.

 

[Korishan]

Could also be a defective FET. It does happen. If you have other BMS units that work as expected, then you happen to get a dud in the mix. You could replace the FET with another one if you feel like resoldering. But it might be faster and cheaper to just replace the board, esp if the seller will do an exchange for you.

 

[ketoneenergies]

Can you tell me how to check the FET, or share any links? i am using 16s Lifepo4 cells. i have spare16s bms of Li-ion 3.7v, Can i use it's FET?

 

[daromer]

You can use its fet if it is the same. Look at its number. Do not mix fets if there are more than 1 and definitely not if they have different values.

Start with checking if you have visual signs of damage. Like burn marks or burned out legs or something.

 

[ajw22]

bms is designed for 50A, but when the current draw increase above 5A, temperature starts to rise. Actual resistance of the pack is 63 mΩ but bms resistance is 59mΩ. Why bms resistance so high?

Could also be a defective FET.

The high ON-resistance of 59mΩ points to the FETs getting only a 2~4V On-signal ("Gate voltage" or "driving voltage"), when it probably requires 5V~15V to fully turn on. Would need to look up the datasheet for the part to confirm.
The "left" leg of the FET that's soldered to the smaller PCB pad should be the gate terminal, the right leg with the bulkier connection should be the "source" terminal.
Measure the voltage between source<->gate on each FET. If one (or more) gate has near 0V, it's probably shorted and causing the low gate voltage for everyone else.

If you should find the damaged FET(s), you could just lift up both legs (left and right) to disable the FET(s) and use the BMS, albeit at a slightly lower max current. Replacing with a working FET would of course be better.

Another (or additional) possibility is that the "FET driver circuit" that supplies said 5V-15V voltage is damaged. Fixing that will require a lot more work.

 

[ketoneenergies]

thanks you. i have tested the voltage. in normal bms, fet voltage is around 13v for all. The bms having the isssue is showing 4.6v for 6 fet and 0v of 6fet.

 

[ketoneenergies]

The high ON-resistance of 59mΩ points to the FETs getting only a 2~4V On-signal ("Gate voltage" or "driving voltage"), when it probably requires 5V~15V to fully turn on. Would need to look up the datasheet for the part to confirm.
The "left" leg of the FET that's soldered to the smaller PCB pad should be the gate terminal, the right leg with the bulkier connection should be the "source" terminal.
Measure the voltage between source<->gate on each FET. If one (or more) gate has near 0V, it's probably shorted and causing the low gate voltage for everyone else.

If you should find the damaged FET(s), you could just lift up both legs (left and right) to disable the FET(s) and use the BMS, albeit at a slightly lower max current. Replacing with a working FET would of course be better.

Another (or additional) possibility is that the "FET driver circuit" that supplies said 5V-15V voltage is damaged. Fixing that will require a lot more work.

 

[ketoneenergies]

Thanks for your help

 

[ketoneenergies]

The mosfet in the defective bms is crss052n08n and the spare mosfets i am having is kia KNB3308A. i am attaching the datasheets of both. how do i find which is the bad FET and can i replace it with the spare FET i am having.

 

[ajw22]

FETs can only stop electricity flow in one direction (eg. discharge), and cannot stop when flowing the other way (eg. charge). So in a BMS, a pair or FETs arranged head-to-head is used to stop current flow in both directions.
The FETs on the right column are responsible for stopping the charging current, but always lets through discharging. Gate 0V means the BMS is preventing charging... should it be?

The FETs on the left column are responsible for stopping the discharging current, but always lets through charging. Gate voltage of 4.6V for the crss052n08n is kinda-ON, but a bit short of the ~6V it needs to be ON properly.
One (or more) of the FETs on the left might have an internal fault and partly-shorting the gate voltage down to 4.6V. For each FET, try desoldering the gate leg and measure the resistance to the source leg. It should be infinite - if not, you've found the culprit.

 

[ajw22]

The mosfet in the defective bms is crss052n08n and the spare mosfets i am having is kia KNB3308A.

They have similar enough performance data, so I think you can use those. Not ideal, but certainly better than not having.

 

[ajw22]

try desoldering the gate leg and measure the resistance to the source leg.

Also measure the resistance between Gate and Drain (the backside terminal of the FET)

 

[ketoneenergies]

FETs can only stop electricity flow in one direction (eg. discharge), and cannot stop when flowing the other way (eg. charge). So in a BMS, a pair or FETs arranged head-to-head is used to stop current flow in both directions.
The FETs on the right column are responsible for stopping the charging current, but always lets through discharging. Gate 0V means the BMS is preventing charging... should it be?

The FETs on the left column are responsible for stopping the discharging current, but always lets through charging. Gate voltage of 4.6V for the crss052n08n is kinda-ON, but a bit short of the ~6V it needs to be ON properly.
One (or more) of the FETs on the left might have an internal fault and partly-shorting the gate voltage down to 4.6V. For each FET, try desoldering the gate leg and measure the resistance to the source leg. It should be infinite - if not, you've found the culprit.

Thank you ajw22. I have found the culprit thanks to you, i lifted the gate leg of bad FET and i measured the resistance to the the source leg. it was showing some resistance. After that bms resistance dropped to normal. it was difficult to de-solder, so i left it like that. now its working properly. Once again thank you ajw22, you saved lots of time for me. I didn't expect that i will find solution for my problem when i posted this thread, thanks everyone who have commented.