Best place to put thermistor in China-controller

[underdog]

I have tried to find the a thread that answers this question, but cant feel I have got the answer I am looking for. Please correct me if such a thread allready exist (as it probably does, somewhere...)

Background:
Am am modifying the shunt on a 9FET 23A china controller, for a bitt of extra boost. I figured I can use the CA3 to limit actual current via a external shunt. And using a thermistor in the controller connected to the CA3 I can monitor the extra heat, to stop it from burning up.

I think the best place for the thermistor would be mounted on the FET`s "cooling-block". Am I right? ...or is there a another place/component more likely to blow from the tuneup?

 

[Skedgy Sky]

Put it directly on the FET body. That's the best place.
The FET becomes hot before the cooling block, and the rest of the controller does.

One time I went for a ride just down the block with my new controller I was testing. Temp sensor on FET body.
I go down the street and controller blew. I look down at my temp display and it was reading above 90C. Case was cool to touch.

 

[mvly]

Sounds like a defective controller. FETs don't blow at 90C. I have seen it happen all the time. Controller is cool and one or more FETs just die for unknown reason. For me it was faulty controller design so one FET was being stressed more than others.

 

[Skedgy Sky]

Sounds like a defective controller. FETs don't blow at 90C. I have seen it happen all the time. Controller is cool and one or more FETs just die for unknown reason. For me it was faulty controller design so one FET was being stressed more than others.

90C was gotten when I looked down maybe 5 seconds after coming to a stop. I am sure it was higher at the time it blew.

 

[underdog]

I am thinking best solution would be to have a indoor-rig that let me stress the controller while it is open. That way I could meassure alle the FET`s temp under load. And put the thermistor on the hottest one... But then again. Maybe a lot of work for a small increase in "security"...

I am sitting there moding it right now. Only tok away about 30% of the resistance in the shunt, so should not be super-dangerous anyway...

 

[underdog]

Sounds like a defective controller. FETs don't blow at 90C. I have seen it happen all the time. Controller is cool and one or more FETs just die for unknown reason. For me it was faulty controller design so one FET was being stressed more than others.

How much heat can a FET be run with without malfunction under normal conditions? I see that the standard value in the CA setup utility is 130C. But this might be intended for a referance for hubmotors, not controller? Also considering that the thermistor is mounted on the outside, and some glue inbetween, some "heat-loss" should maybe be calculated in the max-value?

 

[Skedgy Sky]

Sounds like a defective controller. FETs don't blow at 90C. I have seen it happen all the time. Controller is cool and one or more FETs just die for unknown reason. For me it was faulty controller design so one FET was being stressed more than others.

How much heat can a FET be run with without malfunction under normal conditions? I see that the standard value in the CA setup utility is 130C. But this might be intended for a referance for hubmotors, not controller? Also considering that the thermistor is mounted on the outside, and some glue inbetween, some "heat-loss" should maybe be calculated in the max-value?

I would say that safe FET body temp with some room to spare is 90C.

I remember pushing my 18FET through 1 foot ice covered snow straight through a local park in the winter. My bike was plowin' through it on one (rear) wheel with me hanging on... and temps got up pretty high. 90C.. it was probably doing like 30+ amps continuous at very low RPM to cut through the icey layer. ... controller did not blow.
I was surprised how hot it got as I never have seen temps go so high without the controller blowing.
So that's just my limit.

 

[underdog]

Sounds like a defective controller. FETs don't blow at 90C. I have seen it happen all the time. Controller is cool and one or more FETs just die for unknown reason. For me it was faulty controller design so one FET was being stressed more than others.

How much heat can a FET be run with without malfunction under normal conditions? I see that the standard value in the CA setup utility is 130C. But this might be intended for a referance for hubmotors, not controller? Also considering that the thermistor is mounted on the outside, and some glue inbetween, some "heat-loss" should maybe be calculated in the max-value?

I would say that safe FET body temp with some room to spare is 90C.

I remember pushing my 18FET through 1 foot ice covered snow straight through a local park in the winter. My bike was plowin' through it on one (rear) wheel with me hanging on... and temps got up pretty high. 90C.. it was probably doing like 30+ amps continuous at very low RPM to cut through the icey layer. ... controller did not blow.
I was surprised how hot it got as I never have seen temps go so high without the controller blowing.
So that's just my limit.

Haha. I got a seriously funny image of that in my head :lol: Thanks for input. I think I will test with a current roll-back starting av 70C, and max 90C... And then do some hillclimbing...