TPS57140-Q1: phase cross Zero at low frequency when measuring bode plot

Hello,

As you can see from the datasheet, the above loop response does not match the expected curve from the Applications Section:

How are you measuring the loop response?  Please see below the typical set up.  Between Vout and Vout' should be a 10 - 50 Ohm resistor.

Also, what are the test conditions for these measurements (VIN, Load conditions)? 

Thank you,

Katelyn Wiggenhorn

Hi,

My second image did not attach:

Best Regards,

Katelyn

Hi Katelyn,

Thanks for your reply. Yes. we measure the bode plot with same configuration as you shows. Iout=496mA ( normal load)  and 800 mA( full load) . Thanks.

BRs

Given

Hello,

What is the VIN during for the shared test results?

Also, can you share what is the dB level of the injecting signal? The injected signal may not be large enough at lower frequencies to accurately measure the loop response. Could you try to increase the signal to 30dbm, and see if you see this change?

Thanks,
Katelyn

Hi,

Vin is 6V+/-10%. design target is 6V to 16V.

it seems the phase gain curve is similar when using -30dbm. see below.   In my opinion, since it happens at low frequency, I think it will not affect the loop stability. And from the phase margin, it's stable.  If so, it will not affect the application and I think customer's main concern is on this. Please kindly add your comments.

one more thing is that, if the AC injection is increased to 5dbm, there seems no such thing. But I think 5dbm is too large that curve at high frequency is not correct.

Thanks.

Hi Given,

I am not sure if your network analyzer can vary the injection signal amplitude at difference frequency. We also normally do this on our AP instruments bode plot analyzer.

Another way to check for stability is to do some load transient and see the load transient response whether it is ringing or not.

Thanks
-Arief