TPS543C20: The output is oscillated.

Hi Alec, 

The following is output waveform. Sorry for wrong information, EN is about 2V, not low.

I have disconnected most of the load, 

Thanks

Peter

Hi Alec,

This is start up waveform.

Thanks

Peter

Hi Peter, 

If EN is left floating, it will be pulled up to BP internally. The waveforms you showed look like the device is current limiting. Based on your Ilim resistor selection, the current limit should be around 34A. Would you be able to capture the SW node by chance during the startup event? Also, I know you mentioned most of the load was removed, but is there any low impedance path to ground the output sees?

Also, I can take a look at the layout to verify if the routing and component placement may be impacting what you are seeing.

Sincerely,

Alec Biesterfeld

Hi Alec, 

The following is output and BST(green) waveform. I measured the output impedance (not powered up), and it's more than 10K. Short to GND shouldn't be the cause. But if you think it's overcurrent, I will check other non-passive components.

There are only a few supporting components for this chip. I don't know which one i should change for debug.  Is Rramp worth trying from 187K to 51.1K for control loop? or trying connect a electrolytic cap? 

I will send layout after check.

Thanks

Peter

Peter,

I think a few things may be worth looking at in this scenario. I would first examine if a lower ramp setting helps correct the issue (51.1 kohm and 8.66kohm are two options you could look at given the large inductance). I would also try a lower inductor value as our design tools recommend 0.500uH- 1.5uH for this application. Changing the inductor value will likely create a need to change RAMP. Please let me know if this is a feasible option and I can recommend a RAMP value based on the inductor selection.

A final debug option, that I do not suspect is the issue but is nonetheless something to rule out is pulling out the LED pulled down by PG. This provides a load on BP, which should be sustainable given the rating of the part, but I would like to rule it out as an issue if the RAMP and inductor changes do not succeed in your case.

Regards,

Alec Biesterfeld

Hi Alec, 

This 5V is for a daughter. After I added the daugter card, this DC/DC is working well, which means TPS543C20 is not stable under light load or no load. Is there any way to get stable under light load? 

By the way, two boards using TPS543B20 can boot up under same light load despite there is a short time oscilation.  If there is a daughter card, both C20 and B20 has no oscilation.

Thanks

Peter

Hi Peter,

How are your RSP and RSN pins routed? Does the daughter card need to be connected to complete the feedback?

Regards,

Alec Biesterfeld

No, as you see in the sch, whole required components are local, then provide 5V to the daughter.

Thanks

Peter