Half bridge SMPS - transistors blow up under inductive load

I have built a half bridge SMPS that outputs 110V at around 5A amps to power a CNC motor. It's based on SG3525 with a GDT. The power supply works fine under resistive loads (some light bulbs), there's no abnormal heating up of the components even after a long time under load, but when connecting the motor the transistors blow up after short time. There is soft start implemented.
Am I missing special considerations when driving these kind of loads?
 

Silvio

Member
I have built a half bridge SMPS that outputs 110V at around 5A amps to power a CNC motor. It's based on SG3525 with a GDT. The power supply works fine under resistive loads (some light bulbs), there's no abnormal heating up of the components even after a long time under load, but when connecting the motor the transistors blow up after short time. There is soft start implemented.
Am I missing special considerations when driving these kind of loads?
I believe that when you are using those kind of motors it is surely the fact that as you are saying the transistors are getting hot it is clearly seen that the peak currents are exceeding a certain limit. A simple cure is to use a more efficient heatsink and also added fan cooling if heat is the problem. Considering you said even after a short time It can also be the case that some spikes are getting back to the smps and disturbing the waveform on the gates. In this case added filtering is to be considered at the motor's input.
You have to take measurements and see what's happening on the current at the switching transistors and also on the DC output. Preferably use your scope to see the current peaks. Just remember when smps is used for continuous loads the current rating of the transistors must be rated at 4 times better.
 
Top