If an IGBT's collector-emitter max voltage is 600v, do I need 1kv rated capacitors on the output side?

Thread Starter

CptChronic

Joined May 30, 2022
6
Hello guys. I'm fixing this solar pv diverter (pictures attached at the bottom of this post), other than the traces it has a blown IGBT and 2 capacitors.

The datasheet for the IGBT (model FGH40T65SPD) is here: https://www.onsemi.com/pdf/datashee...4QFnoECA0QAQ&usg=AOvVaw3K90Ez6-A2s0c8bYZ8jAPq

I've tested the other non blown capacitors and found the missing ones are 12uf and 56nf, the max output from the IGBT collector-emitter voltage can reach 600v though so do I need to use caps rated at 1kv? On the datasheet it looks like normal operation is around 20v but I'm not 100%. What are your thoughts?


Thanks


20220530_182722.jpg20220530_182715.jpgScreenshot_20220531-001302_Acrobat for Samsung.jpgScreenshot_20220531-001313_Acrobat for Samsung.jpg
 

Attachments

Papabravo

Joined Feb 24, 2006
18,991
It depends entirely on how close to the absolute maximum ratings you want to come. I don't know what choices you think there are between 600V, which I would not use, and 1000V which I would consider using.
 

Thread Starter

CptChronic

Joined May 30, 2022
6
It depends entirely on how close to the absolute maximum ratings you want to come. I don't know what choices you think there are between 600V, which I would not use, and 1000V which I would consider using.
I'm more wondering if I can get away with 250v caps as the 1kv are hard to get hold of and very expensive compared to the 250v ones. Just a little confused as the operating voltage on the graphs is 20v. I've not worked on anything like this before and it's my first time coming across an IGBT so I'm not sure what the output is going to be
 

Papabravo

Joined Feb 24, 2006
18,991
I'm more wondering if I can get away with 250v caps as the 1kv are hard to get hold of and very expensive compared to the 250v ones. Just a little confused as the operating voltage on the graphs is 20v. I've not worked on anything like this before and it's my first time coming across an IGBT so I'm not sure what the output is going to be
I think I can safely advise you NOT to allow 600 volts to appear across a capacitor rated for 250 volts.
 

Papabravo

Joined Feb 24, 2006
18,991
Of course, but there is not going to be 600v going through the emitter though is there. That is the max it can handle, not what it will run at
I haven't the slightest idea. I don't have a clear picture of the circuit you are talking about. A schematic would be worth several thousand words.
 

Papabravo

Joined Feb 24, 2006
18,991
Couldn't agree more, there doesn't seem to be one available though unfortunately
Then your question cannot be answered if we don't know how the capacitor is connected in relation to the IGBT. It is possible that it could be configured in such a way that 600 Volts on the emitter could appear across the capacitor. It doesn't matter what I can imagine, it only matters WTF is on the board. How hard could it be to reverse engineer a schematic from the board?
 

Thread Starter

CptChronic

Joined May 30, 2022
6
Then your question cannot be answered if we don't know how the capacitor is connected in relation to the IGBT. It is possible that it could be configured in such a way that 600 Volts on the emitter could appear across the capacitor. It doesn't matter what I can imagine, it only matters WTF is on the board. How hard could it be to reverse engineer a schematic from the board?
Well I have no clue how to do that so, hard
 

Papabravo

Joined Feb 24, 2006
18,991
Well I have no clue how to do that so, hard
You start with one of the parts, follow the traces, and make a schematic as you go. It is straightforward and tedious, but it is far from hard. Multilayer boards where you cannot see the traces are harder but still far from impossible. How badly do you want the answer?
 

Thread Starter

CptChronic

Joined May 30, 2022
6
You start with one of the parts, follow the traces, and make a schematic as you go. It is straightforward and tedious, but it is far from hard. Multilayer boards where you cannot see the traces are harder but still far from impossible. How badly do you want the answer?
Fair enough, I'll give that a go.

The max voltage for the 10uf smd multilayer capacitors seems to be 100v anyway though so it can't possibly need a 1kv

I need 2 different caps, on my testers they're coming up as 50nf and 11.x uf, so I'm guessing they will be 47nf and 10uf as those other values aren't available and it's within 10%
I can find 1kv 47nf's but the max 10uf's I can get is 100v
 
Top