LM5140-Q1: no output

Hirotsugu Kobayashi

Part Number: LM5140-Q1
Other Parts Discussed in Thread: TMDS64EVM,

Tool/software:

Hi team,

The LM5140 on TMDS64EVM is breaking down.

VDDA and SS2 are working, but no pulse is being output from channel 2 side.
I am turning the power on and off as per the procedures in the user's guide, and am not performing hot plugging.
Despite this, failures are occurring frequently, so I am investigating the cause.
When SW1 is turned off, a negative voltage is applied to the EN2 terminal, and I suspect that this may be the cause of the failure. What do you think?

7 months ago

0 Timothy Hegarty 7 months ago

TI__Guru 66182 points

Hello,

Please send the schematic and a completed quickstart calculator file for review. The EN pin has a -0.3V negative limit. You can maybe go a little more negative, but the internal ESD diode should not be allowed to conduct.

Regards,

Tim

0 Hirotsugu Kobayashi 6 months ago in reply to Timothy Hegarty

Mastermind 9255 points

Hi Tim,

The schematic is here.

https://www.ti.com/tool/TMDS64EVM?keyMatch=TMDS64EVM&tisearch=universal_search&usecase=hardware

There is no output from HO2 and LO2 on the channel 2 side.
I send waveforms SS2, COMP2, VDDA and RES.

The soft start circuit seems to be working, but there is no gate output. SS2 also rises to nearly 5V and then drops, and then it doesn't work.

The Channel 2 side is broken, so we are investigating the root cause.
In the process, we confirmed that a negative voltage exceeding the rated voltage is detected when SW1 is turned off.

Could this negative voltage be the cause of the damage?

Best regards,

0 Mari Tsunoda 6 months ago in reply to Hirotsugu Kobayashi

TI__Expert 4920 points

Hi Tim,

Linking the thread on the Sitara MPU side so that you are aware of the background of this inquiry and discussion on MPU side.

https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1390668/tmds64evm-don-t-work/5371815#5371815

Best regards,

Mari Tsunoda

0 Hirotsugu Kobayashi 6 months ago in reply to Mari Tsunoda

Mastermind 9255 points

Hi Tim,

Channel 1 works normally.

0 Timothy Hegarty 6 months ago in reply to Hirotsugu Kobayashi

TI__Guru 66182 points

Hi Kobayashi,

Are both phases similar here? If so, please check all the connections on phase 2. Is this behavior occurring on one board or multiple?

A short negative transient on EN shouldn't cause an issue, but you can put a cap on that pin to filter it out. Also, check if COMP2 voltage is high.

Regards,

Tim

0 Timothy Hegarty 6 months ago in reply to Timothy Hegarty

TI__Guru 66182 points

Is the schematic here corresponding to that shown in page 37 of the "PROC101D_SCH_With_Design_Updates" PDF?

Please check the boot diode and ensure that the boot cap is charged to ~4.5V. Otherwise, there may be inadequate gate drive to the high-side FET.

Have you tried installing the LM5140-Q1 on a known good board to check if it’s an IC or a board issue?

Regards,

Tim

0 Hirotsugu Kobayashi 6 months ago in reply to Timothy Hegarty

Mastermind 9255 points

Hi Tim,

The board is TMDS64EVM. The circuit diagram is PROC101C(004)_SCH.pdf in SPRR462C.ZIP.
It worked at first, but after using it a few times, it stopped outputting 5V.
I have several units that have broken in the same way, so I am checking to see if there are any design issues.

L5, C69, and C71 have been changed in RevD. Do they relate to the failure in RevC?
The boot diode is charged to about 4.5V, but it does not start switching.

0 Timothy Hegarty 6 months ago in reply to Hirotsugu Kobayashi

TI__Guru 66182 points

Hi Kobayashi-san,

Maybe you can do an ABA swap to confirm whether it's an IC or a board issue. Also, you can put a Schottky diode on EN to clamp the negative transient.

Regards,

Tim

0 Hirotsugu Kobayashi 6 months ago in reply to Timothy Hegarty

Mastermind 9255 points

Hi Tim,

This board was designed and provided by TI.

Five boards have already failed. The failure mode is the same - 5V is not output.

I think there is a fundamental problem.
Is there a problem with the LM5140 circuit?

Best regards,

0 Timothy Hegarty 6 months ago in reply to Hirotsugu Kobayashi

TI__Guru 66182 points

Hi Kobayashi-san,

We are looking into if VCC is transiently seeing an overvoltage when the input is hotplugged. This could be solved with a resistor divider on EN to set the VIN on/off thresholds.

Our FAE is currently investigating this. I will respond once we have completed testing.

Regards,

Tim

0 Hirotsugu Kobayashi 6 months ago in reply to Timothy Hegarty

Mastermind 9255 points

Hi Tim,

Thank you for your investigation and we look forward to your test results.

0 Timothy Hegarty 6 months ago in reply to Hirotsugu Kobayashi

TI__Guru 66182 points

Thank you.

0 Hirotsugu Kobayashi 6 months ago in reply to Timothy Hegarty

Mastermind 9255 points

Hi Tim,

Another one has broken. It is not outputting 5V.
The board revision is Rev.C.
I have not performed hot plugging.
Have you been able to reproduce this issue on your side?

Best regards,

0 Timothy Hegarty 6 months ago in reply to Hirotsugu Kobayashi

TI__Guru 66182 points

Kobayashi,

Try connecting a 17.4kΩ resistor from EN2 to GND. This, along with the existing 100kΩ pullup to VIN, will setup a UVLO on channel 2 where it turns off if Vin goes below 7V.

Also, you mentioned EN2 is going negative. Having the EN divider here may help that. You can also add a Schottky diode from EN2 to GND to prevent its voltage going negative.

Regards,

Tim

0 Timothy Hegarty 6 months ago in reply to Timothy Hegarty

TI__Guru 66182 points

You can install the 17.4kΩ resistor where C406 is here:

0 Hirotsugu Kobayashi 6 months ago in reply to Timothy Hegarty

Mastermind 9255 points

Hi TIm,

The operation process is as described in the user's guide.

Plug the plug into the EVM

Plug the AC adapter into the outlet

Turn the SW on

Turn the SW off

Unplug the AC outlet

Unplug the plug from the EVM

I don't think there's any point in changing the VIN UVLO threshold.

0 Hirotsugu Kobayashi 5 months ago in reply to Hirotsugu Kobayashi

Mastermind 9255 points

Hi Tim,

You stated the following:

”Our FAE is currently investigating this. I will respond once we have completed testing.”

Were you able to reproduce the issue of 5V not being output?

0 Timothy Hegarty 5 months ago in reply to Hirotsugu Kobayashi

TI__Guru 66182 points

My understanding is we were not able to reproduce, only postulate as to a possible cause.

Power management

Power management forum

LM5140-Q1: no output