WEBENCH® Tools/LM5088: Chip blows up when 75 V is supplied

William Dunkley

Part Number: LM5088
Other Parts Discussed in Thread: LM5146-Q1, LM5005, LM5146

Tool/software: WEBENCH® Design Tools

When I supply 53 volts to the board it works just fine. The LM5088 heats up to about 130 F nothing to crazy, but when I supply 75 V with no load the LM5088 blows instantaneously. Can It not really handle 75 V?

over 4 years ago

0 Daniel Li14 over 4 years ago

TI__Mastermind 27470 points

Can you monitor the input waveform on the VIN PINof LM5088, whether it have overshoot?

Thanks

0 William Dunkley over 4 years ago in reply to Daniel Li14

Prodigy 130 points

I blew up a couple more for fun to see if the input voltage overshot and it didn't. It hit a max of around 75.34 volts. It seems like I shouldn't be using a chip at its maximum rated voltage, but it should still work so I'm wondering what I need to do differently for the next one.

0 Daniel Li14 over 4 years ago in reply to William Dunkley

TI__Mastermind 27470 points

do you use full bandwidth to check the input voltage. If the chip damaged when first applied the VIN, most possibility is that chip damaged by input over voltage

Thanks

0 Timothy Hegarty over 4 years ago in reply to Daniel Li14

TI__Guru 74052 points

Hi William,

Rated input voltage for the LM5088 EVM is 55V based on the 60V diode (MBRB2060) and 75V FET (Si7148DP). The diode rating notwithstanding, it's not a good idea to apply 75V DC as there will always be some SW voltage spike / overshoot, thus exceeding the FET rating.

I recommend the LM5146-Q1 100V controller for such high input voltage conditions.

Regards,

Tim

0 William Dunkley over 4 years ago in reply to Timothy Hegarty

Prodigy 130 points

I blew up an LM5005 and LM5088 both on boards that I made(not the eval boards). I sent a board utilizing the LM5146 a few days ago so we'll see if it was just overshoot on the input voltage. Thanks for in the info Tim.

Will

0 Daniel Li14 over 4 years ago in reply to William Dunkley

TI__Mastermind 27470 points

HI William:

thanks, i will close this question, if you have further question, you can reopen it.

0 William Dunkley over 4 years ago

Prodigy 130 points

It turns out my 75V power supply jumped to 108V for 8 nanoseconds when it first starts up. I did end up switching to an LM5146 which also blew up. I then moved 3 10 uF capacitors very close to the input on the LM5146 and the MOSFETs. This also blew up... I then added a 47 uF cap in place of one of the 10 uF caps and this was able to control the initial spike.

I needed much more capacitance near my chip to account for the large voltage spike on power-up.

0 William Dunkley over 4 years ago

Prodigy 130 points

It turns out my 75V power supply jumped to 109V for 4 nanoseconds when it first starts up. I did end up switching to an LM5146 which also blew up. I then moved 3 10 uF capacitors very close to the input on the LM5146 and the MOSFETs. This also blew up... I then added a 47 uF cap in place of one of the 10 uF caps and this was able to control the initial spike.

I needed much more capacitance near my chip to account for the large voltage spike on power-up.

0 Timothy Hegarty over 4 years ago in reply to William Dunkley

TI__Guru 74052 points

Hi William,

Please send the schematic and completed quickstart calculator file for the LM5146-Q1 application.

The input voltage should be kept below 100V, especially if this is also close to the FET rating. I recommend adding an electrolytic input capacitor with high ESR to provide input filter damping. This will also reduce the high-frequency behavior shown in the scope shot above.

Regards,

Tim

0 William Dunkley over 4 years ago in reply to Timothy Hegarty

Prodigy 130 points

Sorry if I didn't clarify it in the last post but I did already add a 47uf electrolytic capacitor and that solved the problem. I was just showing what my problem was without it.

0 Timothy Hegarty over 4 years ago in reply to William Dunkley

TI__Guru 74052 points

Sounds good, William. Thanks for the clarification. Also, check that the SW voltage spike remains below 100V (assuming you're using 100V rated FETs).

Regards,

Tim

Power management

Power management forum

WEBENCH® Tools/LM5088: Chip blows up when 75 V is supplied