TPS40422: Design review

Hello,

How many boards were damaged and how many are in the field? For how long the board is in the field? What is the circumstances i.e high temp, low temp?

Regards 

Hello, I tried to find the EMC13050S-R82M inductor spec on Google but did not appear. My question:

1- What is the load current on the 5V rail?

2- What is the EMC13050S-R82M saturation current and its DCR?

3- Are the temp sensors (BJTs) in different page?

Regards

Dear Mahmoud Harmouch,

Thanks for your reply.

The customer responses is as follows.

 - Failed q'ty : 12 / 20K, Running 6 to 12 months in the field, Room temperature

 1. Max 4.5A

 2. The inductor actually used is Pulse PG0077.801NL.

      Please refer to the attached datasheet PG0077.801NL.PDF

 3. Temp sensor is not used separately(TSNS1, TSNS2 pins are floating).

Regards,

JH

Hello JH and thank you for sending the info!

The schematic looks OK. Is that possible to send scope capture for the U2 gate drive signals? and SW node as well.

Regards

Hello Mahmoud,

1. T_G (PIN No.3)

2. B_G (PIN No.5)

3. V_SW (PIN No.6,7,8 Common)

Thanks,

JH

Hello JH,

Sorry I meant in the same screen with zoom in at the switching time. Please configure the scope in full bandwidth. I'm trying to see if there is cross conduction or shoot through.

Regards

Hi Mahmoud,

Please see the following images.

Regards,

JH

Hell JH, something does not look right in the waveform. I'm not sure if it is real or noise picked up due to parasitic. The low drive signal goes to 2V during the off time and suppose to be flat at 0V. The power FET threshold voltage is around 1V. So 2V is enough for cross conduction.

...

...

Did you have the chance to verify if the 2V is from noise or real? You can check by differential probe LDRV signal to GND

Hi Mahmoud,

Does the 2V you mention refer to the area marked in red in the picture below?

Regards,

JH

Hi Mahmoud,

Regards,

JH

Hello JH, 

the switch node shows negative spike of -4V and the limit is -1V. The positive spikes at switching is 21V and the limit is 30V. I would like to see the LDRV2 signal pin 23 by itself. I want to see if the 2V during off time. If it is real, there is cross conduction and this may damage the power FET and the TPS40422 as well. Please send same scope pictures the 1V rail. It is for comparison between the 2 rails signals.

Regards

Hello Mahmoud,

Here is the customer's responses.

1. LDRV2 (PIN23)

2. V_SW (Max=23.8V, Min=-4.2V)

3. V_SW (1V rail)

Regards,

JH

Hello,

The negative spike on SW pin and positive/negative spikes on LDRV2 during OFF time could be the root cause of cross conduction and damage. Is it a new design? can you change the design by adding a 5.1 between LDRV2 and power stage BG pin? 

Regards

Could you please tell us what '5.1' is and how to add it?

Regards

Sorry! 5.1 Ohms resistor between the controller gate drive LDRV2 pin and the power FET gate BG pin. The idea is to slow down the switching edge and lower the spikes.

Regards

The customer wants to summarize this review.

  1. There is a possibility that the FET may be damaged due to undershoot occurring on the SW pin of the FET.

  2. The undershoot is caused by overshoot/undershoot that occurs due to cross conduction during switching off time of the controller.

  3. To eliminate (or reduce) overshoot/undershoot, it is suggested to add a resistor between the LDRV2 pin of the controller and the BG pin of the FET.

Is the customer's understanding correct?

How was the resistance value of 5.1Ω determined?

Regards