LMR33630: Large inrush current with inverting buck operation

user5028679

Part Number: LMR33630

Hello,

I designed a board using the LMR33630ARNXT in the inverting buck mode to generate -9V from a 24V supply and I am seeing issues when power is cycled and the output capacitors are not discharged.

On initial power up with everything discharged the inrush current is a quick spike as expected but when power is cycled and the inrush current peaks to 3.5A for ~25ms causing a supply rated for 3A to enter over current protection.

Swapping to a supply rated for 5A the power rail comes up but the large inrush current is still seen when powering VIN without discharging the Output capacitors.

The EN pin was tied to VIN so the device starts up when the VIN rail reaches ~2V (LMR33630A undervoltage lock-out threshold), could this be the issue?

I am not sure why there is a difference when powering the board with the output caps discharged vs charged.

Thanks,

Sean

over 5 years ago

0 Samuel Jaffe over 5 years ago

TI__Mastermind 25590 points

User,

It's interesting that you would see a higher current spike when the output capacitors are already charged. How much have they discharged when you power the LMR33630 on again?

I would expect this pre-bias startup condition to have negative current initially as the low-side FET turns on when VOUT > VIN. Could this be causing the issue?

Please share waveforms so we can confirm the input and output voltages and currents in the condition you described.

Thanks,
-Sam

0 user5028679 over 5 years ago in reply to Samuel Jaffe

Prodigy 120 points

Hi Sam,

Attached is a PDF containing the waveforms. It looks like the VOUT has discharged to ~-2.5V when the power is reapplied.

Thanks,

SeanLMR33630A_Scope_shots.pdf

0 Frank De Stasi over 4 years ago in reply to user5028679

TI__Guru 57555 points

snva856.pdfPart of the issue may be that the Cio cap is seeing a larger voltage when the output is not fully

discharged, when Vin is applied. The switcher will try to react to the pre-biased output by

bringing the output to zero before soft-start can bring the output to the correct regulation point.

You can try to use an input UVLO to hold off the switching until the input voltage is almost up to normal.

I would also suggest that you do not start into a pre-bias load condition.

I have attache an app note that may be helpful.

0 Samuel Jaffe over 4 years ago in reply to Frank De Stasi

TI__Mastermind 25590 points

Sean,

We have not heard back so I will close this thread. Please reply or start a new post if you have further questions (this thread will lock 30 days after the original post date).

-Sam

0 user5028679 over 4 years ago in reply to Frank De Stasi

Prodigy 120 points

Thank you, we will test UVLO and use this in the next spin. And ensure this spin has a load attached to the rail.

Regards,

Sean

Power management

Power management forum

LMR33630: Large inrush current with inverting buck operation