transistor switch gets hot quickly and damaged

Thread Starter

ulzi

Joined Dec 16, 2018
9
Guys
please see attached simple curcuit and give me advices
it works normally but igbt can't withstand for 2-4 minutes , it sits on heat sink and also has air fan
from datasheet , as i undestood , diode and igbt are ok imo , or ok?
 

Attachments

drc_567

Joined Dec 29, 2008
714
The data sheet indicates a significant decrease in collector current as the case temperature increases. Have you applied a thermally conducting compound of some type between the IGBT and the heatsink?
 

Thread Starter

ulzi

Joined Dec 16, 2018
9
The data sheet indicates a significant decrease in collector current as the case temperature increases. Have you applied a thermally conducting compound of some type between the IGBT and the heatsink?
No, transistor bolted in heat sink directly. but i think , problem is not because of it , transistor is getting hot very quickly.

as i understood , back emf spike from inductor is heating the transistor and killing it ? but diode f30u60s has quite fast speed that open before bemf spike reach the transistor . am i right ? or diode is slow opened ?
 

AnalogKid

Joined Aug 1, 2013
8,107
Typical saturation voltage is greater than 2 V. With the 3 ohm resistor shown on the schematic, that's a collector current of 4.8 A, and a power dissipation of 9.6 W. How good is the heatsink?

ak
 

ebp

Joined Feb 8, 2018
2,332
What is the function of the circuit supposed to be? It is a "flyback" circuit. If inductor winding phasing in the as-built circuit is not as shown in the simulation, then the behavior is completely different.

If the circuit is built as diagrammed, it should be OK except for the leakage inductance spike at IGBT switch-off. Because IGBTs are comparatively slow, it may or may not be being forced into breakdown due to this spike. Normally in such a circuit a snubber is used to assure the voltage is limited to less than the breakdown voltage of the transistor. In practical circuits, the leakage inductance is usually a much bigger concern than the so-called forward recovery time of the diode in the output circuit. It is rare to be concerned with forward recovery time - usually it is reverse recovery that matters and in this circuit even that isn't very important because it operates in discontinuous current mode.
You cannot simply put a diode across the inductor and not grossly change the behavior of the circuit.

With the expected peak collector current of about 2.4 A, and with a 10% duty cycle, the average current is 120 mA (calculated peak current with zero ohms instead of 3 ohms and transistor saturation voltage taken as 0, is 2.4 A). The transistor should be perfectly OK with no heatsink. However, if the inductor does not have adequate rating or the circuit is not as shown, the inductor may be saturating which will allow the current to rise to the point where the series resistor is the dominant limiter.
 

recklessrog

Joined May 23, 2013
988
There can be some difficulties when using IGBT power devices, as, unlike a normal BJT, they are voltage controlled. You may be experiencing some high frequency oscillations, or latch up. Have you connected a "real" 'scope to your circuit?
Here is an article by International Rectifier that may help you.
 

Attachments

Last edited:

crutschow

Joined Mar 14, 2008
23,353
No, transistor bolted in heat sink directly. but i think , problem is not because of it , transistor is getting hot very quickly.
If the heat sink is not also getting hot, then that is part of your problem.
You need to use a thin layer of thermal grease to mount the transistor on the sink.
 

Thread Starter

ulzi

Joined Dec 16, 2018
9
What is the function of the circuit supposed to be? It is a "flyback" circuit. If inductor winding phasing in the as-built circuit is not as shown in the simulation, then the behavior is completely different.

If the circuit is built as diagrammed, it should be OK except for the leakage inductance spike at IGBT switch-off. Because IGBTs are comparatively slow, it may or may not be being forced into breakdown due to this spike. Normally in such a circuit a snubber is used to assure the voltage is limited to less than the breakdown voltage of the transistor. In practical circuits, the leakage inductance is usually a much bigger concern than the so-called forward recovery time of the diode in the output circuit. It is rare to be concerned with forward recovery time - usually it is reverse recovery that matters and in this circuit even that isn't very important because it operates in discontinuous current mode.
You cannot simply put a diode across the inductor and not grossly change the behavior of the circuit.

With the expected peak collector current of about 2.4 A, and with a 10% duty cycle, the average current is 120 mA (calculated peak current with zero ohms instead of 3 ohms and transistor saturation voltage taken as 0, is 2.4 A). The transistor should be perfectly OK with no heatsink. However, if the inductor does not have adequate rating or the circuit is not as shown, the inductor may be saturating which will allow the current to rise to the point where the series resistor is the dominant limiter.

yes , i also dont think that put a diode across the inductor . inductor is air cored bifiliar winded , so no saturation here.
the citcuit is part of my little research on energy field. if i use snubber , where i put ? diode is also considered snubber , put it parallel with inductor ? btw , english is not my native language, question i have about meaning "across the inductor = parallel with inductor " ??
if RC snubber , where i put ? i think that snubber is working by way of power dissipating , means , losing energy
from your reply , as i understand , i put diode with short forward recovery time at output winding , transistor should be long turn off delay time ?
 
Top