LMR33620: Regulator short circuited as soon as powered ON

Hello

The input EMI filter may be causing an overvoltage on the input.  

The ceramic input capacitors have very low ESR, and will ring with the EMI filter L.

Try connecting a 22uF or 47uF Al. electrolytic capacitor across the input of the device,

as an experiment.  The ESR of the Al capacitor will provide damping to the filter.

Also, please provide PDFs of your PCB layout for us to review.

Keep in mind that the LMR33620 is only rated to 36V

Thanks

Dear Frank De Stasi,

     Actually we have given provision for EMI filter in the design, while testing the proto PCB we have not populated the inductor L1 (shorted pads with thick wire ) of EMI filter, remaining all input capacitors are mounted. Note that C1,C2,C4 & C5 are AL.Electrolytic and only C6 is a ceramic cap. 

Using bulk caps like 22uF or 47uF can create high inrush current. Moreover WEB BENCH tool did not suggest to use bulk caps on input side. And doing multiple experiments is not possible because of shortage of samples and high lead time. So please try to give an exact solution.

I will send PCB layout file in a while.

Hello

Please also send waveforms of the voltage on the VIN pin and the SW pin of the device during you testing.

Thanks

Hello Frank,

We have tested the board by adding 47uF AL-EL cap on the input side of the regulator and observed the following waveforms on VIN pin and CBOOT pin. Channel 1 is input supply at power connector. Channel 2 is VIN(2) pin of regulator. Channel 3 is CBOOT pin(7) of regulator. Unfortunately we have not measured SW pin, at the time of power on. 

This time device is not getting short circuited but the device is not generating the output voltage. Observed a 5V at CBOOT pin.

Channel 1 is input supply at power connector. Channel 2 is VIN(2) pin of regulator. Channel 3 is CBOOT pin(7) of regulator.

SW pin is measured 500mV as shown below. Output is also 500mV instead of 5V.

Hello

It looks like you have some very bad waveforms at the input to the regulator.

You should test with a stable bench supply and see if you get better results.

That would be the best starting point to troubleshoot this issue.

We also need to see your PCB layout to help further.

Thanks

Hello Frank,

     Plz refer the PDF file of my layout.

LMR33620.pdf

Hello

The component designators do not correspond between the schematic and your layout.

Please provide a schematic that has designators that correspond to your layout silk screen.

Thanks

Hello Frank,

     Plz refer the PDF file of my layout.

33620.pdf

Thanks

Hello Frank,

    Thank you for reviewing the PCB layout.

1. Vias under the DAP of the device to connect the ground plane, are already there in the layout. They are not highlighted in the pdf attached.

2. You are suggesting to add extra ceramic cap on both input and output side of device. Refer to your comment on jul 7, there you have mentioned that usage of ceramic caps will lead to ringing. So why it is not considered here.

3. I accept that, changing the layout according to your suggestions may increase the device performance, but I don't think device will get damaged with those reasons.

4. All 5 boards that I have made assembled, got damaged. I have replaced new device in one board and tested with 10uF Ceramic on input side and output side also, even though device got damaged. 

5. I have used different buck converters from TI, but never faced any issue like this. I have seen others posting the same issue in the forum.

No one got the correct solution.

Hi Khyarui Arham,

I have seen your post earlier and given a try removing the resistor. But it did not help me.

Hello

Did you try placing the 220nF and 10uF close to the device VIN and GND pins?

Also, did you try with an external supply?

Thanks

Hello Frank,

    Yes I have placed 220nF closed to Pin 1&2 of the LMR33620 and also placed 10uF cap. Changed input power source.Even tough facing the same problem.

Desoldering from one board and soldering it again in another board is also causing the damage to device without switching the power ON.

Procured 12 devices totally, out of them 10 got damaged during all these exercises. I am really disappointed with this device.

Thanks

Hello

Have you tried using an external clean input supply connected to your board?

Thanks

Hello Frank,

    Yes our LAB supply is certified and working fine and clean regulated output. We are using lab supply  from Scientific and Adapter we used before is from Sony .

Thanks

Hello

Did you see an issue when connecting the lab supply directly to the input of the regulator ?

Thanks

Hello Frank,

Yes I have seen the issue using the lab supply. Tried two ways, one is hot plugging and the other is switching on after connecting the supply. In both cases device got damaged.

Thanks

Hello

I still believe that it is some kind of overvoltage event or severe noise on the input.

Can you take a scope shot of the input voltage when you are using the bench supply.

Also, what load are you using during the test?

Thanks

Hello

I will close this post due to inactivity.

Thanks

Hello Frank De Stasi,

       I could not give reply because I am out of station these days. The issue is not solved yet. I did not see any over voltage event with my supply. I used the same supply for another board which contains TPS54302 regulator from TI and it is working fine. Find the captured images of bench supply output voltage. It is clean, no over voltage and a ripple of 540mVp-p. Load on the device during test is 1.5A maximum. I believe that only LMR33620 is creating some issue when powered with this supply. 

Hello Sreehari,

I am of a similar opinion to Frank. 

From your layout I observe that the path from +Ve input terminal of the board to the VIN pin of the device must travel from the top side of the board, down to the Layer 8, and back up to the top side of the board. This is a large input loop which will result in larger than necessary parasitic inductance which will result in ringing on the input node. This ringing may be exceeding the absolute maximum voltage rating for the input pin which is resulting in the device being damaged.

Could you please take a scope shot of the input voltage when using the bench supply when turning on the device? 

Ideally what I am looking for in the scope photo is as follows: Input voltage at the pin of the LMR33630 device, the SW node voltage, and the input supply at the power connector. 

Regards,

Harrison Overturf

Oh I didn't found it on the short description of LMR336xx, may be I missed it. Thanks for the information...