Author Topic: Low-Noise Buck For Bias Supply  (Read 976 times)

0 Members and 1 Guest are viewing this topic.

Offline Evan.CornellTopic starter

  • Regular Contributor
  • *
  • Posts: 189
  • Country: us
Low-Noise Buck For Bias Supply
« on: August 21, 2020, 12:48:05 am »
I'm struggling a bit identifying appropriate buck. Here's the requirements:
VIN=10-40V, nominal 24V
VOUT = 1.8V
IOUT = 20mA max, 5mA nominal

I've currently got this implemented with TPS54061, but the problem is that I originally designed for the full 200mA capability of the part in DCM, but under real load conditions (much lower, more like 5mA), I get pulse-skipping. This is creating noise in band of interest (10-40kHz) that is coupling into sensitive signal conditioning circuitry. Using TPS54061 design spreadsheet, I can't solve for DCM solution at this light load and avoid pulse skipping. There seems to be solution for CCM, but then again, TPS54061 doesn't have all that low IQ to begin with.

Does anyone have any suggestions for a high conversion ratio buck, low output current, that isn't going to generate all sorts of skip/pulse/ECO mode noise in the 10-40kHz region?
 

Offline pardo-bsso

  • Regular Contributor
  • *
  • Posts: 235
  • Country: ar
Re: Low-Noise Buck For Bias Supply
« Reply #1 on: August 21, 2020, 01:26:36 am »
I presume that going linear or adding a stable minimum load is not suitable for your application?
 

Offline Evan.CornellTopic starter

  • Regular Contributor
  • *
  • Posts: 189
  • Country: us
Re: Low-Noise Buck For Bias Supply
« Reply #2 on: August 21, 2020, 11:47:26 am »
I presume that going linear or adding a stable minimum load is not suitable for your application?

Linear, no, since that increases power by an order of magnitude. Minimum load, possibly. I am going to take a stab to get TPS54061 designed into CCM mode first and see if that resolves the problem.
 

Online KT88

  • Frequent Contributor
  • **
  • Posts: 334
  • Country: de
Re: Low-Noise Buck For Bias Supply
« Reply #3 on: August 21, 2020, 12:27:22 pm »
The ADP2360 could fit https://www.analog.com/en/products/adp2360.html
It uses PFM which could create issues as well but in a quick sim I got >90kHz out of it at 40V/5mA.
 
The following users thanked this post: Evan.Cornell

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17429
  • Country: us
  • DavidH
Re: Low-Noise Buck For Bias Supply
« Reply #4 on: August 22, 2020, 04:36:58 pm »
The problem is that the TPS54061 implements fixed current sensing internally so operating at low current is prone to noise.  A part with external current sensing could be adjusted to operate properly or a voltage mode controller could be used.  A higher output inductance will ensure CCM operation at low current.

A tapped output inductor could be used to raise the peak current and duty cycle but that requires external power switches and an expensive inductor.
 

Offline Evan.CornellTopic starter

  • Regular Contributor
  • *
  • Posts: 189
  • Country: us
Re: Low-Noise Buck For Bias Supply
« Reply #5 on: August 22, 2020, 10:28:25 pm »
The problem is that the TPS54061 implements fixed current sensing internally so operating at low current is prone to noise.  A part with external current sensing could be adjusted to operate properly or a voltage mode controller could be used.  A higher output inductance will ensure CCM operation at low current.

A tapped output inductor could be used to raise the peak current and duty cycle but that requires external power switches and an expensive inductor.

The internal sensing is the problem.. I tried modifying the circuit for much higher L, was still getting the noise. See TI e2e post with more details here:
https://e2e.ti.com/support/power-management/f/196/p/933931/3450237

TI recommended changing to one of their parts with forced PWM (FPWM) mode, which is the path I will take.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf