Should MOSFET get warm?

Thread Starter

Flug540

Joined Mar 24, 2012
60
Finally got a chance to continue the project...
Measured the frequency to be around 478 Hz, which is consistent with what Arduino site says:
"The frequency of the PWM signal on most pins is approximately 490 Hz".
Also finally fixed my reference voltage problem so the control is now functioning properly,
which caused the FET temps to rise to high 50s, on several occasions it reached around 62C, for the most part hovered around 54C-58C.
Another thing that also bothers me is the shape of the signal, attached. How does it look to you, guys?
As the next step I'm thinking to remove the 200 ohm series resistors to see if that will have an effect on the temps.
 

Attachments

ronv

Joined Nov 12, 2008
3,770
It looks like it might be oscillating. How far away from the FET is the Arduino?? You want to keep the path from the mico to the gate and the source back to micro ground as short as you can.
You might also try the single FET and 91 ohms or you can try a low value 10 -20 ohms and see how the micro likes it. See if it gets warm.
Take a look at the gate voltage while your at it.
 

Thread Starter

Flug540

Joined Mar 24, 2012
60
Arduino is probably about 4 to 5 feet away. Currently Arduino sits with the rest of the system's electronics, while the MOSFET module is removed because of the high current and the hazards associated with it.
I should probably move Arduino out of the computer box and into the remote MOSFET box. Currently though, I'm using a shielded wire between Arduino and the MOSFET module.
The inside is the signal and the shield is connected to the ground on both sides. I'll try your suggestions next. Thanks!
 

Thread Starter

Flug540

Joined Mar 24, 2012
60
Time for an update...

Since last time I've tried several different configurations and now the design looks much better.
The major problem turned out to be the distance from the micro to the MOSFET; as soon as I put them next to each other, the temps have dropped.
I've also reduced to using one instead of three parallel MOSFETs and removed the heat sink to see what will happen.
With the heat sink and good ventilation the heat sink remained at ambient temps.
With the sink removed and temp sensor mounted to the FET, it was generating somewhere around 10C above ambient.
Now I replaced PSMN1R1-30PL with IPB009N03L, no heat sink and the temps dropped even more (as expected because of the lower Rds).
Now fully open FET generates around 4.5C above ambient non ventilated.

I also tried replacing the gate series resistor from 150ohm to 33ohm and didn't really notice any significant (if any at all) difference, I guess my frequency is not high enough for that to be a problem.

Next issue I'm dealing with is heat generated in the diode. Yesteday I removed its heat sink and attached a temp sensor.
Turns out it's generating around 15C-20C above ambient.
The one I'm using now is STPS1545, I'm thinking of trying 65PQ015PBF instead, it has a lot lower voltage drop, hopefully it will be much cooler.

Thank you guys for the support!
 

#12

Joined Nov 30, 2010
18,224
Time for an update...

Since last time I've tried several different configurations and now the design looks much better.

Thank you guys for the support!
You were part of the success. You might be surprised by the number of people that ask for advice and then refuse to do what we say. Those threads can go on for over a week and over a hundred posts. :eek:
 
Top