Fault of TPS562200 circuit

Your resistor divider values are too high  I would recommend 10.0k for the lower, then calculate the upper from that.  C420 should not be used.  Are you hot plugging the input voltage?  Layout is also critical.

Thank you John.

Can you explain how these changes make the IC burn?

BR

Gil

BTW, I followed TI's Webench recomantations for this circuit design. Including resistors values and Cff.

Webench has been updated to correct this issue.

Thank you john.

It will take me few days to get more board to validate the solution. Hope it will work for us. I will keep you updated.

Thanks again

Gil

Hi John,

1. After replacing FB resistors to 10K/33.2K on 7 defected boards, and testing each one 15 times - 2 devices burned again.

2. When saying "hot plugging", do you mean applying V_in and GND at once or powering the circuit with no load?

3. Can you please advice?

Attached layout:

"Hot plugging"  Is directly applying Vin to your circuit by connecting Vin to the circuit while Vin is present vs connecting your Vin supply when Vin = 0 V, then allowing Vin to ramp up from 0 V to Vin at some normal ramp up time, usually in the range of a few msec.  Hot plugging can cause voltage transients up to Vin x 2 which can exceed the abs max and damage the IC.

I also see that you have a long return path from your input cap to ground.  That may be your issue.  Try connecting the ground side directly to your ground pin 1.  You can temporarily jumper it with a short heavy gauge wire.

Thanks a lot for such a prompt response during weekend!

We have only one source of external Vin. When circuit is not powered we may only have the residual voltage of input capacitors.

Cin has vias directly to GND in layer #2 (PCB has 4 layers).

BR

Gil 

The blue ground copper wraps around to get from the input cap to pin 1.  It should be a shorter path.  On the reference design, the input caps are connected to VIN and GND directly on layer 1, while the SW node is routed on the bottom.  You might want to try grounding the cap on the top side with a short wire.

Can you take a waveform of Vin when you apply power?  look for any over voltage conditions.

Attached plot of Vin (14V nominal, 24V peak for few usec) and Vfb (1ms ramp up).

Connecting Cin ground to pin #1 by short wire gives the same results.

I think that 24 V peak may be enough to damage the IC.  You will need to do something to clamp it or possibly reduce the rise time.

I tested 40 circuits after changing resistors and cap. All board worked fine with and w/o protection diode (16V Zener).

I was not able to reproduce the fault, even by getting the old resistors and cap back. Since not all of our board had this issue from the beginning, I am not 100% sure that we pointed the root cause. 

Do you have experience with TPS562200 that burned due to high resistors and/or Cff ?

Dear John,

We followed your recommendations: 10K resistor, no cap between VFB and SW, VIN overshoot < 17.5V.

We have much less fault incidents, but we still get some ICs that got burned few seconds (or minutes) after power-up.

Vin = 12V, Vout = 3V, Iout < 50mA.

My questions are:

1. Is it possible that touching the device with bare hands is the root cause?

2. Do you have any other suggestion?

3. Is it better to use TPS563200 or TPS562209 instead of TPS562200?

Thank you in advance and happy new year

Gil

Hello,

I am having the same issues with this part.  I originally thought it was an ESD issue.  I burned out 5 boards in a row.  The board would work fine sitting on the desk, but then if I touched anywhere on the board it would fry.  After replacing the parts it seems fine now.  We have a couple dozen boards running.  Maybe there's an intermittent oscillation happening.  I notice that if I touch the circuit w/ my finger I can get the part to fry, even if I touch the FB pin with my 1M scope probe.

Here's my circuit below, power is ORed through two schottkey's from either 7.4V from batteries or 15V from a boost converter.  I've tried unpopulating the phase cap, and tried increasing to 10pF.  If I hold my finger or probe at certain points in the circuit, I can audibly hear the inductors start to hiss, not sure if this is a clue.  

Another behavior I  noticed is that the regulator would shut down intermittently, seeming depending on the stray capacitance/resistance it was presented by the case.

Temperature environment is normal room temp, but it's been hot in NY the past week, so maybe that's affecting something?  

We bought 1 month ago.

Device Marking: