TPS53513: TPS53513RVET is not starting

Ivan Mitic21

Part Number: TPS53513

Hi,

We use TPS53513RVET for quite a few designs up to now, for generating 5V (from 12V) on our boards. They were always operated in FCCM mode @250kHz (Mode tied to PGOOD).

In the newest design the chip broke immediately (afterwards, there was no voltage on Vreg). In the first half of second after startup, some buzzing could be heard indicating some oscillations somewhere.

I switched the mode to SKIP MODE by pulling the Mode pin to GND. After that, the chip did not get broken upon startup but did not start properly either.

I recorded the waveform on the output and noticed that the output raises to 5V and then slowly drops back to 0V. After that, the output voltage remains ~0V and I see that the coil is being switched. I also noted that Vreg is dropping periodically (at a period of 1.5mS).

I can send you the oscilloscope screenshots of the recorded waveforms?

Do you have any idea what can cause it to shutdown like that?

Thank you,

Ivan Mitic

over 7 years ago

0 JohnTucker over 7 years ago

TI__Guru**** 191445 points

Ivan,

Yes, please post your schematics and applicable waveforms.

0 Ivan Mitic21 over 7 years ago in reply to JohnTucker

Prodigy 140 points

Hi John,

Thank you for the quick answer. Here we go:

Vout:

Vout (zoomed):

Waveform of Vreg pin:

Waveform on SW pin:

And finally, the schematic:

We have checked that all components are populated properly and that the layout has been manufactured properly. The only difference is that now the MODE pin is tied to GND.

Thank you,

Ivan

0 JohnTucker over 7 years ago in reply to Ivan Mitic21

TI__Guru**** 191445 points

Ivan,
VREG is obviously dropping out of regulation in a repetitive manner. Does the circuit try to restart during each interval when VREG returns to regulation? Do you ever get any bursts of switching pulses?

0 Ivan Mitic21 over 7 years ago in reply to JohnTucker

Prodigy 140 points

John,

I think it is restarting since the SW and Vreg have the same frequency. Vreg is the yellow plot and SW is blue:

I noted bursts in switching pulses in FCCM mode before the chip broke.

Do you think that can have something to do with control loop stability? Can changing the R-C TIME CONSTANT have some influence? Right now its 60uS (0kOhm Rmode with 250kHz switching freq).

0 JohnTucker over 7 years ago in reply to Ivan Mitic21

TI__Guru**** 191445 points

Ivan,

How many out of how many total boards show this symptom? The VREG signal is going low every msec is caused by the UVP fault. The SW node should be switching continuously during the start up time. The soft start time is 1 msec. During that time teh output should ramp up linearly from 1V to 5V During that time the SW node should switch regularly. You are showing Vout to jump up to a high (~4 V) level immediately. Perhaps your inductor is damaged? Can you check it?

0 Ivan Mitic21 over 7 years ago in reply to JohnTucker

Prodigy 140 points

Hi John,

Only the latest (one) prototype board shows these symptoms. I have checked the coil and it is good (6.8uH).

When the coil was removed, I have measured the same voltage spike on 5V line when I started the board, showing that there is some cross coupling between 12V input and 5V line.

One thing which confuses me is that after the spike, the 5V line is 0V and the IC still wont start. The onboard controller is trying to reset the 5V IC every 10 seconds (by toggling the enable line) and it still wont start although the 5V line is solid 0V and 12V input line is stabilized. Do you know why would it behaves like that?

0 JohnTucker over 7 years ago in reply to Ivan Mitic21

TI__Guru**** 191445 points

Ivan,

If only one board is behaving like that, i would recommend you replace the TPS53513.

0 JohnTucker over 7 years ago in reply to JohnTucker

TI__Guru**** 191445 points

Ivan,
Did you get your board up and running?

0 Ivan Mitic21 over 7 years ago in reply to JohnTucker

Prodigy 140 points

Hi John,

Thank you for your inquiry.

Not yet, I am waiting for the package since we have ran out of those chips.

I will report as soon as I have something...

0 JohnTucker over 7 years ago in reply to Ivan Mitic21

TI__Guru**** 191445 points

Ivan,

I'm going to mark this "TI Thinks Resolved". If you still have issues after you get the new replacement IC, just re-open this thread.

0 Ivan Mitic21 over 7 years ago in reply to JohnTucker

Prodigy 140 points

Hi John,

I have replaced the IC (3rd time) and the IC still has the same behavior.
Do you have any more leads why it can be behaving like that?

0 JohnTucker over 7 years ago in reply to Ivan Mitic21

TI__Guru**** 191445 points

Ivan,

I'll need to see the layout. I need to also see the start up waveform from the first time power is applied and the EN signal is asserted. Please show SW and FB on the same screen.

0 JohnTucker over 7 years ago in reply to JohnTucker

TI__Guru**** 191445 points

Ivan,
Did you get a chance to get the layout and waveforms?

0 Ivan Mitic21 over 7 years ago in reply to JohnTucker

Prodigy 140 points

Hi John,

Sorry for delay. Here are the photos. Due to time constrains we will have to mount external 5V chip and discard this one for now.

But I would still like this solved in for the future designs...

Enable vs Vout:

Enable VS Vreg:

FB vs Switch:

Enable VS Switch 2:

Vin VS Enable:

Thank you,

Ivan

0 JohnTucker over 7 years ago in reply to Ivan Mitic21

TI__Guru**** 191445 points

Ivan,

There is not enough information in those waveforms to make any kind of definitive resolution. I would still like to look at the PCB layout.

0 JohnTucker over 7 years ago in reply to JohnTucker

TI__Guru**** 191445 points

Ivan,

Will you have time to continue troubleshooting this? If not I will close this thread. Please let me know.

0 Ivan Mitic21 over 7 years ago in reply to JohnTucker

Prodigy 140 points

Hi John,

I am sorry for the delay.

No, dont have any more time to trubleshoot this. I have replaced the whole circuitry with 5V DCDC converter block (which doesnt require any aditional components) and solved the problem for now.

We will make one more board to check if the fault was in the board it self and I will get back to you (you can close the thread for now).

Thanks,
Ivan Mitic

Power management

Power management forum

TPS53513: TPS53513RVET is not starting