Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Class D Switching Power Amplifiers and Power D/A conversion

what D amp give best sound.
what D amp give best sound.
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 27th November 2017, 10:21 AM   #61
kees52 is offline kees52  Netherlands
diyAudio Member
 
Join Date: May 2009
Location: Sprang-capelle Holland.
Quote:
Originally Posted by Grizlek View Post
Hi,

I don't think this can be stable. you have to put square signal at input and see if output is stable in time domain.

Low THD is good, but who cares about 0,1 or 0,001 % THD (probably cant be heared), my opinion, but unstable amplifier can be heared I would say.

What about "russian design", uses different feedback ? (other threads there), forum member SOUS

full differential UcD modulator with extra integrator

He use HYPEX topology with extra integrator, but special circuit that "kills" additioanl poles to get amplifier to be stable... hmm
Afcouse we try to diy our own amps, that is the challence. making amps from other members do not.

You say 0.0 or 0.001 do not hear, this is dependent of the kind of distortion, if it is even harmomics the amp get warmer with it, if it is other stuff it sounds bad, I try high end, so need very low distortion, I try to make the modulation liniar with poles, I have that from a paper, the power comparator is a existing one who is stable.

Thanks for the tips, I try a square, but then I get a lot of mudulation interfacts, who as I presume is needed to see how it is with stability.

regards
  Reply With Quote
Old 27th November 2017, 01:26 PM   #62
Grizlek is offline Grizlek
diyAudio Member
 
Join Date: May 2010
Quote:
Originally Posted by kees52 View Post
Afcouse we try to diy our own amps, that is the challence. making amps from other members do not.

You say 0.0 or 0.001 do not hear, this is dependent of the kind of distortion, if it is even harmomics the amp get warmer with it, if it is other stuff it sounds bad, I try high end, so need very low distortion, I try to make the modulation liniar with poles, I have that from a paper, the power comparator is a existing one who is stable.

Thanks for the tips, I try a square, but then I get a lot of mudulation interfacts, who as I presume is needed to see how it is with stability.

regards
MOST basic UCD (Putzeys patent version) gives about 0,1 % thd, and I measure mostly even harmonics.

But it is completely sable with square input... if you add more poles/zeroes.. it is not..
I would say that stable simple ucd must sound better then unstable one with very low simulated thd..
  Reply With Quote
Old 29th November 2017, 09:25 AM   #63
kees52 is offline kees52  Netherlands
diyAudio Member
 
Join Date: May 2009
Location: Sprang-capelle Holland.
I have to try it, I do soon, now I am busy with work. I do not now if with simulation the square test will work properly, I do have to look at output.

I had also try a model from @analogman however with a dead time generator with xor ports I still get shoot throughs, for what concerns the high end distortion, these people has such demands, otherwise you can not sell it to them, if you hear it or not. Class d is less simple as we think, the deadtimes needs to be very tight, and as such we need a very fast transitions or speed, I did try a emittor followed totem driver and a single totem driver, the first did very well driving big mosfets with high gate capacitances..

Adding more poles do not go without trouble, there is a border for how far you can go, but it do work very nicely, I do also use a fase network on the output and a 12 dB octave bessel low pass..
  Reply With Quote
Old 29th November 2017, 09:31 AM   #64
kees52 is offline kees52  Netherlands
diyAudio Member
 
Join Date: May 2009
Location: Sprang-capelle Holland.
I did try a deadtime on this schematic, however it do not cancel out the shoot through, I think the IR2011 model is not oke, also because I get less negative swing as positive and the square do not look oke.

I have now a model of a IRS20954 who has dead time in it, and a single input, that is more easy to use, did hear it is more sensitive for pcb design, but that is with all high speed switch designs.

Last picture is without the dead time and the first with, it looks like things are reversed so the dead time do not work, it even give more shoot through.
Attached Images
File Type: jpg ScreenHunter_524 Nov. 27 20.44.jpg (294.7 KB, 315 views)
File Type: jpg ScreenHunter_525 Nov. 27 20.47.jpg (270.4 KB, 310 views)

Last edited by kees52; 29th November 2017 at 09:40 AM.
  Reply With Quote
Old 7th December 2017, 06:02 AM   #65
Grizlek is offline Grizlek
diyAudio Member
 
Join Date: May 2010
Quote:
Originally Posted by kees52 View Post
I did try a deadtime on this schematic, however it do not cancel out the shoot through, I think the IR2011 model is not oke, also because I get less negative swing as positive and the square do not look oke.

I have now a model of a IRS20954 who has dead time in it, and a single input, that is more easy to use, did hear it is more sensitive for pcb design, but that is with all high speed switch designs.

Last picture is without the dead time and the first with, it looks like things are reversed so the dead time do not work, it even give more shoot through.
I have to find schematics that I use.. Completely remove output stage it will be faster simulation.. and instead of fets use voltage controlled voltage sources, dead time can be generated using RCD circuit (any dead time you want), and delay can be manually added by transmission line. (any delay you want) ... Remember me to post in next days...

Dead time will not prevent shout through and instability.. dead time is 50- 100 nS.. instability lasts at least 2-3 periods of pwm signal..

Like I said "Russian metthod" they probably copied original hypex from Bruno is better (but havent tried it) ...
  Reply With Quote
Old 7th December 2017, 03:45 PM   #66
kees52 is offline kees52  Netherlands
diyAudio Member
 
Join Date: May 2009
Location: Sprang-capelle Holland.
days...

Dead time will not prevent shout through and instability.. dead time is 50- 100 nS.. instability lasts at least 2-3 periods of pwm signal..

Like I said "Russian metthod" they probably copied original hypex from Bruno is better (but havent tried it) ...[/QUOTE]

You did give me an idea.

I have done the discrete comparator, this was from internet and with dead time I get this, and low distortion. The choise of the opamp integrator is very important, speed is needed for a clean output square.

regards
Attached Images
File Type: jpg ScreenHunter_536 Dec. 07 16.42.jpg (257.3 KB, 166 views)
File Type: jpg ScreenHunter_537 Dec. 07 16.43.jpg (158.1 KB, 164 views)
  Reply With Quote
Old 8th December 2017, 01:55 PM   #67
Grizlek is offline Grizlek
diyAudio Member
 
Join Date: May 2010
Quote:
Originally Posted by kees52 View Post
days...

Dead time will not prevent shout through and instability.. dead time is 50- 100 nS.. instability lasts at least 2-3 periods of pwm signal..

Like I said "Russian metthod" they probably copied original hypex from Bruno is better (but havent tried it) ...
You did give me an idea.

I have done the discrete comparator, this was from internet and with dead time I get this, and low distortion. The choise of the opamp integrator is very important, speed is needed for a clean output square.

regards[/QUOTE]

Forum member Sous (dont know if russian or indonesian) there was a link to russian forums. huh..

all in all feedback is advanced integrator with special circuit for preventing selfoscillation and instability.



This method full differential UcD modulator with extra integrator

Use single (non differentical) at the begining.. Try to simulate, you will get very low thd, but it will be very unstable if modulation is 60V peak and supply is 80 V peak.. only low pwm modulations are stable. Special circuits that shorts some pole/zeroes of integrator are needed. The same method uses Bruno Putzeys and hypex.

Maybe Sous schematics and design is free ?! But there was link to russian forums there full schematics give.. try to find and post here if you can (I had no luck)

Here you can see how transistor network shorts additioanl poles/zeroes of integrator. Have not tried with this in spice. This method is better and more stable then you posted. thd is below 0,001 (all powers) etc..

http://www.diyaudio.com/forums/attac...d_mdr-4_09-gif

here is it

Automotive UcD high-quality power amplifier 2x100W

????????????? ????????????????? ?????????????????? ?????? ????????? 2?100?? - ???????? 3

Viewer for layout and schematics lay files in zip folder

KOSTENLOSE DATEI-VIEWER, ELECTRONIC-SOFTWARE-SHOP

Last edited by Grizlek; 8th December 2017 at 02:25 PM.
  Reply With Quote
Old 9th December 2017, 10:47 AM   #68
kees52 is offline kees52  Netherlands
diyAudio Member
 
Join Date: May 2009
Location: Sprang-capelle Holland.
I did this one also, but is from somewhone else, it is like with normal amps, the ideas are look alike, the combination is what matters.

I am a analog amp designer, combiner, class D not yet done really, but it is nice to play.

I had a idea also to do a open loop version full bridge, but I did see that it is not easy, likely very unstable also, I do not now that a ucd with feedback before the low pass can be considered open loop? I think I can not compare the feedback loops with analog amps where feedback do inpack audio depth, that is why I do use current feedback there, independent of bandwidth is a very good thing, special for imd, Designers going digital to try to do no feedback slowly get furter to working one, the designs who work do not sound wel yet as I did read on internet.

Last picture is a discrete power comparator with under 100nS speed with the right parts, special the output mosfets, a same one I did replace the upper and lower driver for a emmittor followed totem pole, it makes even nicer square on higher frequency, see the low pass filter who has multiply poles and a fase correction network on output I had this idea from the internet, it is not complete mine idea, as I say earlyer everything is alrweady tested by university people or designers from big electronic companys, switching amps was already present in 1950.

As what I do now now, speed is mandory for a low distortion, the poles needed for stability, or even to much works against us afcouse, it is a modulated carrier so as with FM radio it needs much attention for liniarity.
Attached Images
File Type: jpg ScreenHunter_538 Dec. 07 22.27.jpg (283.5 KB, 75 views)
File Type: jpg ScreenHunter_540 Dec. 08 00.31.jpg (299.3 KB, 8 views)
File Type: jpg ScreenHunter_528 Dec. 01 22.56.jpg (289.3 KB, 8 views)
File Type: jpg ScreenHunter_536 Dec. 07 16.42.jpg (257.3 KB, 9 views)
File Type: jpg ScreenHunter_539 Dec. 08 00.30.jpg (74.6 KB, 9 views)
File Type: jpg ScreenHunter_537 Dec. 07 16.43.jpg (158.1 KB, 13 views)

Last edited by kees52; 9th December 2017 at 10:50 AM.
  Reply With Quote

Reply


what D amp give best sound.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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Does anyone with a high level sound science background want to give me the words to ag8908 Multi-Way 24 13th April 2014 03:24 AM
High end sound - does any specfic approach give up the goods more than another? Racket Scientist Multi-Way 46 31st August 2010 06:49 PM
Can someone give me informs about Cyrus amplifiers sound quality destroyer X Solid State 11 20th July 2006 03:52 PM
Can anyone give more info on this amp? mg16 Tubes / Valves 6 8th May 2003 01:43 AM
What PSU will give the best sound for an Ono (regs) ? Blitz Pass Labs 0 23rd February 2003 03:27 PM


New To Site? Need Help?

All times are GMT. The time now is 03:46 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio
Wiki