TPS50601A-SP: Output Instability

Hi Albert,

I ran a quick simulation of your setup and my simulation had similar results to yours, showing stability.

What is your switching frequency and load? I assumed 500kHz and 6A for my simulation.

I am trying to understand the topology of your setup. If I am understanding correctly, there is a 12V to 5V converter (5V PoL?) that provides a 5V source to the inputs of 1.2V, 1.8V, and 3.3V converters. The 1.2V and 3.3V converters are working well, but you are seeing oscillations in the output of the 1.8V converter, and it seems to be causing regulation issues with the upstream 12V to 5V rail? Please let me know if that is correct.

It is curious that the voltage on the 1.8V converter would oscillate between 2V and 3.5V, as those are both well above the target 1.8V. Can you confirm that your feedback divider is configured correctly? What does the voltage look like at the device feedback node?

You mentioned that, "The switching node of the 5V PoL seems to be impacted by the instability of the 1V8 PoL". How is the switch node of the upstream (if it is upstream) 5V PoL impacted? Does the duty cycle change? Does the on-voltage decrease?

Would it be possible for you to post oscilloscope screenshots of some of the device nodes? I would be interested in seeing PVIN, VOUT, COMP, PH, and VSENSE (or a subset) if possible.

As for your compensation setup, we do not typically recommend using a CFF cap, so it can be removed to avoid potential problems. Additionally, it does look like your CCOMP and CHF caps are smaller than what I calculated using datasheet equations (11) and (12). Your selected RCOMP value lines up well with a target 9kHz Fco. Page 22 of the device datasheet contains equations for calculating compensation values.

Thanks,

Andy

Hello Andy,

Thanks for the recalculation.

The switching frequency is 1MHz and the current load is around 2A. I think it is unstable, since the cross over frequency is 10kHz which I didn't realize when I first calculated the values.  I will remove the CFF cap and I have recalculated some values where the cross over frequency is higher, more in the 50kHz range.

Thanks!

Albert

ok so I updated the compensation loop, removed CFF, and it looks like its not regulating properly.

Here are something other things that I checked:

1. The feedback resistor is set correctly, so even though the output voltage is oscillating, the oscillating voltage of the feedback divider is the correct ratio.

2. I checked the compensation loop, and while there was a cold solder joint, re-soldering the comp pin resulted in the compensation pin and the compensation passives to see the correct voltage. Around 2V.

Then when I measure the switch node, it looks exactly like output voltage. A triangle wave that oscillates between 2.5V and 3.5V, which you can see below. The switching node is not switching properly. I haven't had the chance to swap the component, but I wanted to double check to see if there is anything else that I am missing. the RT resistor is set to the right resistor value so Im not sure why I would see this kind of switching behavior unless there was an issue with the internal driver.

Thanks!

Albert

Hi Albert,

Thanks for the additional details.

Can you confirm that the voltage at the feedback pin is greater than 0.804V? It should be if your feedback divider is correct.

Can you also confirm that you are seeing ~2V on the COMP pin? If so, that is extremely high and could be a cause / additional symptom of whatever is happening. If you are seeing such a high voltage on COMP, can you please check for tin-whiskers, solder bridges, schematic mistakes, or anything else that could be causing external current to be making its way onto the COMP node?  

Does the COMP pin also show this slow triangle wave pattern?

Thanks,

Andy

Hi Albert,

Glad to hear it's working now!

Please reach out again if you have any other questions, or need any further support.

Thanks,

Andy