Global NFB with muitiple pole output filter - diyAudio
Go Back   Home > Forums > Amplifiers > Class D

Class D Switching Power Amplifiers and Power D/A conversion

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 29th September 2007, 12:27 PM   #1
GK is offline GK  Australia
Account disabled at member's request
 
Join Date: Jan 2006
Default Global NFB with muitiple pole output filter

I知 designing a 1200V half-bridge PWM amplifier as a plate modulator for a vacuum tube class C AM RF transmitter. Upper audio bandwidth limit is 3kHz and the switching frequency is 100kHz.
Switching frequency ripple voltage at the output needs to be rather low (than for speaker driving applications), so a multiple pole LPF will be required at the output. However, the ability to apply a fair amount of global NFB would be desirable.

Any suggestions?

Cheers,
Glen
  Reply With Quote
Old 29th September 2007, 01:27 PM   #2
GK is offline GK  Australia
Account disabled at member's request
 
Join Date: Jan 2006
Oh, here is a scope shot of a pulse test of a 2kV prototype of my half-bridge circuit. Tr=~200nS, Tf=~140nS.

Is that reasonably OK?
Attached Images
File Type: jpg 10us.jpg (44.6 KB, 256 views)
  Reply With Quote
Old 30th September 2007, 12:48 AM   #3
GK is offline GK  Australia
Account disabled at member's request
 
Join Date: Jan 2006
No takers?
FWIIW I値l restate my dilemma. I need to attenuate all harmonics of the 100kHz switching waveform by at least 60dB, but I can't see an easy way to apply any substantial amount of NFB from the output due to the phase shift incurred by the filter.
Maybe there isn稚 a workable solution, meaning I値l just have to run the thing open loop??

Cheers,
Glen
  Reply With Quote
Old 30th September 2007, 01:01 AM   #4
BWRX is offline BWRX  United States
diyAudio Moderator Emeritus
 
Join Date: Jan 2005
Location: Pennsylvania
I've heard others who use multiple pole filters mention that they take feedback from before and/or after the first LC filter section and leave the rest of the filter outside the feedback loop.

Your design requirements mean you'll need some pretty large inductors, in value and physical size; especially if you want to keep losses low.
__________________
Brian
  Reply With Quote
Old 30th September 2007, 01:10 AM   #5
GK is offline GK  Australia
Account disabled at member's request
 
Join Date: Jan 2006
Actually something along these lines is what I was thinking of, but picking off the NFB part way through the filter can lead to some pretty interesting pass-band effects.
I figured that there might be some kind of established design procedure to this kind of problem, but Google unfortunately turns up very little on class D filter design (at least with my searching abilities).

Cheers,
Glen
  Reply With Quote
Old 30th September 2007, 03:49 AM   #6
diyAudio Member
 
Join Date: Dec 2001
Location: SIUE, Illinois, USA
There is always some form of current mode control. While more complex, it does simplify things. It turns the inductor in the filter into a constant current source, which leaves a first order filter on the output. The added inner current loop can lead to issues with 1-cycle vs 2-cycle ("fast" bifurcation) stability.

IIRC there was a "leapfrog" filter of sorts that used voltage and current feeback from the shunt capacitors and series inductors of an output filter.
__________________
if only it could be used for good, not evil...
  Reply With Quote
Old 30th September 2007, 04:27 PM   #7
Tim__x is offline Tim__x  Canada
diyAudio Member
 
Join Date: Jun 2004
Location: Edmonton area, Alberta
Our (once) very own analogspiceman's Leapfrog is the first thing I thought of too.
There's an old thread started by analogspiceman with a fairly detailed paper attached.
  Reply With Quote
Old 1st October 2007, 12:16 AM   #8
GK is offline GK  Australia
Account disabled at member's request
 
Join Date: Jan 2006
Found it. Thanks!

The "Leap-Frog" Method Of Switching Amplifier Control Loop Design

Cheers,
Glen
  Reply With Quote
Old 1st October 2007, 01:26 AM   #9
GK is offline GK  Australia
Account disabled at member's request
 
Join Date: Jan 2006
Hmm.....that Leap Frog method is clever, but implementing a real life version at with the number of filter elements and operating voltages I need would be a bit of an exercise . I also really need a fixed frequency design.

Anyway, if anyone is interested, attached below is my 2kV+ half bridge circuit. MOSFETs are low capacitance high voltage units from Sanyo. D1-D6 are beefy transorbs to clamp Vds. Works very well to over 3kV, but the topmost DC-DC converters begin to break down eventually

Cheers,
Glen
Attached Images
File Type: jpg pwm.jpg (45.7 KB, 226 views)
  Reply With Quote
Old 1st October 2007, 10:59 AM   #10
GK is offline GK  Australia
Account disabled at member's request
 
Join Date: Jan 2006
Okaayyy.........
After considering my options I致e decided to run the LPF open loop. NFB is still mandatory though, so I guess I値l be feeding the PWM waveform directly from the MOSFET half bridge back to the input summer.
I figure that if I just make my opamp based input error amplifier an inverting integrator my dominant pole will be set here.
Lets say I apply 20dB NFB. I require only 3kHz audio bandwidth for (legal) voice communications, so the dominant pole can be set as low as 3kHz/20dB = 300Hz.
Since my fixed PWM clock frequency is 333.33 times higher than this at 100kHz, I figure that the fed-back PWM waveform will be adequately conditioned for me by the intergrating summer. Yes? No?

Honest question for the class D experts out there - Does it sound like I know what I知 talking about? Workable / feasible?

Cheers,
Glen
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
URL collection of Totem Pole and CSPP (Single Ended related) Solid State Output Stage tiefbassuebertr Solid State 7 10th May 2012 05:02 PM
Output filter Skorpio Class D 5 8th August 2008 06:44 PM
Output filter m-tech Class D 11 22nd September 2007 07:04 AM
BTL output filter ted12 Class D 16 26th April 2005 01:55 PM
output LR filter Helix Solid State 7 8th February 2002 12:08 PM


New To Site? Need Help?

All times are GMT. The time now is 04:36 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ゥ1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2