TPS62067 feedforward capacitor narrows the bandwidth

Hongjia-Wu

Other Parts Discussed in Thread: TPS62097

In theory, feedforward capacitor will introduce a zero and a pole, therefore can boost both the bandwidth and phase margin. However, in this paper, bandwidth narrows with increasing Cff, how to understand this result?

Optimizing TPS6206x External Component Selection.pdf

over 9 years ago

0 JohnTucker over 9 years ago

TI__Guru**** 191445 points

I'm curious about that. Let me look into it later this week.

0 Hongjia-Wu over 9 years ago in reply to JohnTucker

TI__Expert 6680 points

Hi John, Thanks for your reply. Do you mean that if the pole is lower than the original bandwidth without Cff, then the new bandwidth with Cff added will be narrowed? I draw the picture as below and according to the picture, the bandwidth is increased, so where did I go wrong?

0 JohnTucker over 9 years ago in reply to Hongjia-Wu

TI__Guru**** 191445 points

I want to think about this a bit. As you can see I retracted my earlier answer.

0 Hongjia-Wu over 9 years ago in reply to JohnTucker

TI__Expert 6680 points

Many thanks John! Looking forward to your reply!

0 Chris Glaser over 9 years ago

TI__Guru**** 150670 points

The TPS6206x family is a bit older and limited in the allowable output filter. This is the purpose of the app note: to show the measured stable range. Reference 1 has a basic discussion of the Cff.

Can you give an example of what you are referring to in the app note? The Cff is changing, but also the LC filter and its response.

Finally, the TPS62097 offers the same 2A step-down converter in a 2x2mm package but with higher efficiency, better output accuracy, a vastly more flexible control topology, and a ton more features. It is recommended.

0 JohnTucker over 9 years ago in reply to Chris Glaser

TI__Guru**** 191445 points

Yes reading that app not carefully, You have to compare table 1 to table 2. The BW always increases (as it should) when Cff is added.

0 Hongjia-Wu over 9 years ago in reply to Chris Glaser

TI__Expert 6680 points

Hi Chris, Thanks for your reminder and insight. I have neglected in table 2 that when Cff is increasing, Cout is also increasing.

0 Hongjia-Wu over 9 years ago in reply to JohnTucker

TI__Expert 6680 points

Thanks, John. Sorry for my carelessness.

0 Hongjia-Wu over 8 years ago in reply to Chris Glaser

TI__Expert 6680 points

Hi Chirs,

When I am doing loop response test, I have made a mistake. I found that if 50 ohm resistor is first in series with the upside resistor then in parallel with Cff, then the loop response is wrong (with Cff added, bandwidths narrows rather than broadens, see as the table below). But I do not know how to analyze this phenomenon with very solid theoretical analysis . I would like listen to your opinion. Would you please kindly help me? Thanks

Cff	bandwidth (khz)	phase margin(degree)	gain margin(dB)
22pF	45.6	82.5	-17.7
no Cff	181	4.2	-1.3

Best Regards,

Emma

0 Chris Glaser over 8 years ago in reply to Hongjia-Wu

TI__Guru**** 150670 points

The first schematic is not valid since it does break the entire loop. This is required by the theory: the entire control loop must be broken/measured in a single place.

Power management

Power management forum

TPS62067 feedforward capacitor narrows the bandwidth