TPS562212: Voltage transient on the input VIN side

Martin Likavec

Part Number: TPS562212

Tool/software:

Hello Ti,

I build very small board with TPS562212 which is use for stepping down the input voltage 12-17V (4*4.2V lipo cell) to 5V (USB-C output).

I noticed voltage transient during plugging and unplugging my device from battery. The device portable, so it is frequently disconnected and reconnected to the batteries.

On the input I have this TVS - SMF18VTF and 2x 22uF/25V 0805 MLCC. For my input voltage range, I didn't find better TVS with lower clamping voltage. If you know about someone, let me know.

When I did a test without TVS, the TPS562212 exploded. :D

With SMF18VTF it looks okay, but my question is... Voltage peak on the below picture exceed the absolute maximum of VIN. Is this short (4.5us) peak acceptable or not? If not, can you give me advice how to improve it?

Above measurement was done without any load. The board was just powered from battery.

Thanks
Best regards,
Martin

7 months ago

0 Manish Gautam 7 months ago

TI__Expert 8990 points

Hi Martin,

We will check this issue and come back with solution.

Also, can you please share the schematic.

Regards,

Manish

0 Martin Likavec 7 months ago in reply to Manish Gautam

Prodigy 180 points

Hello Manish,

VCC net is directly connected to battery connector using these pins:

Best regards,
Martin

0 Manish Gautam 7 months ago in reply to Martin Likavec

TI__Expert 8990 points

Hi Martin,

This peak is responsible for the damaging the board. Voltage peak should be less than the abs max of the limit.

Please try to debug the root cause of increasing the peak voltage for short time range. please remove the device from the board and then measure, is that the same peak will be coming or not.

Regards,

Manish

0 Martin Likavec 7 months ago in reply to Manish Gautam

Prodigy 180 points

Hi Manish,

As I note, this peak is transient during connecting my board to battery connector. (hot plugging)

Lipo battery is very stiff and powerful supply. So I am asking how to reduce this fast dv/dt voltage spike. The probe was connected between two banana plugs with "GND spring".

When I tried to did similar test with regulated power supply(below) the transient is different. Because power supply is not so stiff power supply as battery and also used cables maybe reduce this peak. Be inform that during this test the board was connected to regulated power supply via 20cm cables and the mínus (gnd) was permanently connected. I only connect and disconnect the possitive VCC rail.

So, in MZ opinion the problem is with hotswaping to battery connector and I need some solution how to reduce it. TVS is not a solution because 18v TVS don't exist with Vclamp around 20V.

Max power of my board is 7.5W.

Thanks

Martin

0 Martin Likavec 7 months ago in reply to Martin Likavec

Prodigy 180 points

I have two idea. What will be better?

1. Add small resistance in series with VCC?

2. input RC snubber?

0 Manish Gautam 7 months ago in reply to Martin Likavec

TI__Expert 8990 points

Hi Martin,

Input RC snubber will be better for removing that spike.

Regards,

Manish

0 Martin Likavec 7 months ago in reply to Manish Gautam

Prodigy 180 points

Can you help me how to calculate it correctly according to first scope screenshot?

Thanks

Martin

0 Manish Gautam 7 months ago in reply to Martin Likavec

TI__Expert 8990 points

Hi,

I am attaching a link, please follow that link.

https://www.daycounter.com/Calculators/Snubbers/Snubber-Design-Calculator.phtml

That link briefly discuss about the snubber calculation also you will find the snubber calculator in that link.

Regards,

Manish

Power management

Power management forum

TPS562212: Voltage transient on the input VIN side