Hi All,
I have a JBL-PRX12M which had a major blow up. First one I've worked on and seems from discussion with other techs around the country, tech's chose not to touch them. (fingers burnt) Awful to work on!
Blow-up is the common issue if powered from a varying AC supply voltage such as generators or faulty 3 phase distro issues. The main culprit the FAN7380 high side gate driver.
On the unit, both IGBT's were short circuited high side IGBT a big hole where the smoke come out. The totem pole BJT push pull pair transistors and resistors gone to god. Complete mess. The internal reference supply 3 x MPSA43A in parallel with 10R series resistors supplies the +18Vdc to get all this running were all toast including the 20V zener! Why did I start.
I've replaced everything including the TL494 and have now have a good clean square wave drive to the gate of the Low side. clean square wave from pin 5. For testing purpose I have yet to solder in the IGBT's and have no HT supplied to the module. I've applied +22V dc form a external power supply after the supply diode from the Switch transformer to get the switching happening.
The high side I can't seem to get any drive signal for pin 7 & 6 for the high side of the FAN7380.
My question is:-
1. Do I need to have the IGBT's in circuit with HT supplied to have the high side section bootstrap to operate correctly?
2. Do I install the IGBT's and power the HT rail with a low voltage 90 Vdc with current trip to get the FAN7380 to do its thing?
3. Whats the recommend strategy for testing before power-up with high side drivers?
Any information be hopeful.
Many Thanks
Rob
I have a JBL-PRX12M which had a major blow up. First one I've worked on and seems from discussion with other techs around the country, tech's chose not to touch them. (fingers burnt) Awful to work on!
Blow-up is the common issue if powered from a varying AC supply voltage such as generators or faulty 3 phase distro issues. The main culprit the FAN7380 high side gate driver.
On the unit, both IGBT's were short circuited high side IGBT a big hole where the smoke come out. The totem pole BJT push pull pair transistors and resistors gone to god. Complete mess. The internal reference supply 3 x MPSA43A in parallel with 10R series resistors supplies the +18Vdc to get all this running were all toast including the 20V zener! Why did I start.
I've replaced everything including the TL494 and have now have a good clean square wave drive to the gate of the Low side. clean square wave from pin 5. For testing purpose I have yet to solder in the IGBT's and have no HT supplied to the module. I've applied +22V dc form a external power supply after the supply diode from the Switch transformer to get the switching happening.
The high side I can't seem to get any drive signal for pin 7 & 6 for the high side of the FAN7380.
My question is:-
1. Do I need to have the IGBT's in circuit with HT supplied to have the high side section bootstrap to operate correctly?
2. Do I install the IGBT's and power the HT rail with a low voltage 90 Vdc with current trip to get the FAN7380 to do its thing?
3. Whats the recommend strategy for testing before power-up with high side drivers?
Any information be hopeful.
Many Thanks
Rob
Last edited:
What I do with switching supply or amps or led drivers:
1. you could provide all voltages but main, that goes to IGBT's. Here you can have IGBT's in their place. Furthermore, you can then supply you main voltage with something, like external power supply, so you have a lot more control and time to check different test points with scope.
That would be my way of doing it, since all the circuit will work as it would with high voltage, but you have option to supply main voltage with variable power supply, which can be high voltage, so you can start from 0v or almost 0v to your rated HV (if you have such supply) and doesn't hurt if it has current limit, even better if adjustable.
I have one supply that goes up to 130Vdc and if device works at that voltage and I don't see any problematic waveforms, there is good chance that it would work even with 400Vdc.
2. goes about the the same as 1. if you decide to go the way I do it
3. is again done like in 1. You can power up whole circuit, but without main voltage (you will have to check, if there is any lower voltage that is needed, that is stepped down from you main HV. Power supply and led drivers might have HV startup, there you would need supply for driver seperate from HV OR use variable HV and slowly crawl to some HV, that would be enough to power us circuit)
Again, this way (supply all voltages, but main, have all components installed(maybe not some capacitors, depends on design itself(some have HV capacitors after fuse and across fets or IGBT's, which in case, that transistors don't close in time or open at wrong time, short those capacitors, which results always in blows transistors)))
1. you could provide all voltages but main, that goes to IGBT's. Here you can have IGBT's in their place. Furthermore, you can then supply you main voltage with something, like external power supply, so you have a lot more control and time to check different test points with scope.
That would be my way of doing it, since all the circuit will work as it would with high voltage, but you have option to supply main voltage with variable power supply, which can be high voltage, so you can start from 0v or almost 0v to your rated HV (if you have such supply) and doesn't hurt if it has current limit, even better if adjustable.
I have one supply that goes up to 130Vdc and if device works at that voltage and I don't see any problematic waveforms, there is good chance that it would work even with 400Vdc.
2. goes about the the same as 1. if you decide to go the way I do it
3. is again done like in 1. You can power up whole circuit, but without main voltage (you will have to check, if there is any lower voltage that is needed, that is stepped down from you main HV. Power supply and led drivers might have HV startup, there you would need supply for driver seperate from HV OR use variable HV and slowly crawl to some HV, that would be enough to power us circuit)
Again, this way (supply all voltages, but main, have all components installed(maybe not some capacitors, depends on design itself(some have HV capacitors after fuse and across fets or IGBT's, which in case, that transistors don't close in time or open at wrong time, short those capacitors, which results always in blows transistors)))
Last edited:
Hey Luka,
Thanks for you input to my question.
I rather make sure all is good before supplying 340Vdc across the IGBT's. Like you have indicated, if the switching dead time isn't correct. Instant Kapoof killing the IGBT's and possibly high side driver again. I'll take your suggestion about using a dedicated supply to drive the IGBT main supply with some sort of current limit.
Thanks for you input to my question.
I rather make sure all is good before supplying 340Vdc across the IGBT's. Like you have indicated, if the switching dead time isn't correct. Instant Kapoof killing the IGBT's and possibly high side driver again. I'll take your suggestion about using a dedicated supply to drive the IGBT main supply with some sort of current limit.
Most of the time I use two supplies, one is LV for control circuit and HV. It is really great if you have time to check and know that your circuit is functioning correctly, before any real HV is applied. Way better than hoping it will work the first time
- Status
- Not open for further replies.