LM5025A: 12V converter primairy FETS problem

Teun Teeuwen

Part Number: LM5025A

Hi,

I asked a some previous questions about this basic reference design. I seem to still have problems with the main N-ch mosfet. It keeps breaking down and shorting.

I previously used a 150V 30A FET with +/- 16mOhm Rdson. It held up for quite a while and suddenly it broke down and shorted out.This happened after turning up the output current to +/-7A. I used a calculation excel sheet which calculated the stress voltage to +-/170V. Never measured voltages above 100V peak on the input line though. (specification is 36-80V input voltage). Then I upgraded to a 200V 34A 55mOhm mosfet and it broke down immediately without even drawing any output current. I checked the dead times etc. they are not causing massive dissipation losses.

Has anyone ever encountered similar problems?

over 1 year ago

0 Jonathan Wong over 1 year ago

TI__Expert 5615 points

Hello Teun,

I assume that you are referring to the PMP7806 from the last E2E post you asked: https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1135440/lm5025a-pmp7806/4213736?tisearch=e2e-sitesearch&keymatch=%2520user%253A515112#4213736 is this correct? Can you also share with me the calculation excel sheet that you used?

Regards,

Jonathan Wong

0 Teun Teeuwen over 1 year ago in reply to Jonathan Wong

Prodigy 70 points

No it was actually about an older post. On a LM5025A design.

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1082706/lm5025a-fast-switching-gate-signal-during-overlap

and

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1111219/lm5025a-secondairy-side-fets-break-down

These issues seem to be resolved. But on further testing the above happened.

LM5025A Calculator 13Vout.xls

This is the excel file.

0 Jonathan Wong over 1 year ago in reply to Teun Teeuwen

TI__Expert 5615 points

Hello Teun,

Is this problem similar to the first post "LM5025A: Fast switching gate signal during overlap" where the MOSFETs turn on but then stop working after some time? Also, are there any changes since the previous post or are you using the same schematic?

Since your MOSFETs are shorting, I suspect the problem is due to a timing issue where both of the MOSFETs are conducting at the same time. The recommendations that Steve provided for your secondary side MOSFETs would be useful in this problem. Do you have waveforms of the outputs?

Regards,

Jonathan Wong

0 Teun Teeuwen over 1 year ago in reply to Jonathan Wong

Prodigy 70 points

No it is a new problem. The schematics at the front have not changed much, on the secondary side some components are added to reduce the gate driving voltage. But since those are minor changes i didn't think that would affect the front much.

I'm very curious how the fets on the front keep breaking because there is never a higher than spec voltage or current measured at their inputs.

0 Jonathan Wong over 1 year ago in reply to Teun Teeuwen

TI__Expert 5615 points

Hello Teun,

It seems unusual that this problem suddenly happened when you made no changes to the schematic. Since you never have a voltage higher than the spec, I think, as stated above, the MOSFETs are somehow conducting at the same time. Have you tried swapping the LM5025A with a new chip? There is also the possibility that the board is damaged or there is a short somewhere causing the MOSFETs to fail. Is this problem happening on multiple boards or just one?

Without further information, test data and outputs, I cannot provide many other suggestions besides what Steve M provided in the last E2E post.

Regards,

Jonathan Wong

0 Teun Teeuwen over 1 year ago in reply to Jonathan Wong

Prodigy 70 points

It has happened on several boards. The 150V I used before broke down on some and lasted a few hours on others. Even when drawing 5A at 12V. So to me there seems no logic in why they would break or sometimes last.

0 Jonathan Wong over 1 year ago in reply to Teun Teeuwen

TI__Expert 5615 points

Hello Teun,

Do you have any thermal data? If the FETs were working for some time, especially over a few hours, then they may have overheated. Did they feel warm at all?

Regards,

Jonathan Wong

0 Teun Teeuwen over 1 year ago in reply to Jonathan Wong

Prodigy 70 points

I have thermal images. But until they break down there is nothing to see. And it seems like a perfectly fine board.

After they break of course they heat up very fast because of the short in them. So thermal data doesn't really help me in finding the problem.

Power management

Power management forum

LM5025A: 12V converter primairy FETS problem