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

Class D investigation.
Class D investigation.
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 4th January 2017, 11:39 PM   #11
Pafi is offline Pafi  Hungary
diyAudio Member
 
Join Date: May 2004
Location: Budapest
Quote:
Originally Posted by kees52 View Post
I did sim with your setup Pafi

see the result, is this what you mention.

regards
No, I explicitely wrote NOT to use windowing on single period. I wrote this 2 times.
  Reply With Quote
Old 5th January 2017, 12:37 PM   #12
kees52 is offline kees52  Netherlands
diyAudio Member
 
Join Date: May 2009
Location: Sprang-capelle Holland.
Quote:
Originally Posted by Pafi View Post
I don't know what you changed and what do you expect.

I did change the feedback parts, so oscillation did change, now it was very high, this way I do learn some things what happens, that is the fun of simulation. some papers has very nice tests, but I get negative offset in the discrete comparator.

Do you now a very fast output stage? for example with ic comparator and ic gate drivers, there are gate drivers who go quite high I did want to test on 800 Khz carrier. Did ask also if soft switch is possible, but I think this do not is the case because we have to demodulate audio and with extra coils this go wrong.

The use of a triangle wave is the old system, it let stay the carrier frequentie stable, when use a very high quality triangle and analog feedback this can give a nice stable amp, the self oscillating versions did have the problem with the lowering carrier when output get higher and drops when get near supply rails, here can a transistor, two resistors on both sides make a nice protection, shut of the amp when current peaks, with a hold for 30 seconds before the amp switch on again this kind of electronics is not that difficult.

I have a clip detector who shut of the amp for 1 minute (it is tube hybrid) and yjis works nicely.

For ltspice, I have design the allfet circlotron but when I measure in real time I get quite high harmonics, when simulation I get verly low distortions, possible because I did measure with the pc and the floating circlotron output things get messed up, I do need to build a differential to single output converter, then I can do it better, agree?,. But Maybe I did sim wrong, but measuring just one half of the amp is also not a real nice way, so I go build a converter or maybe one on ebay.

regards
Attached Images
File Type: jpg ScreenHunter_ Dec. 18 16.43.jpg (197.4 KB, 325 views)
File Type: jpg ScreenHunter_ Dec. 21 15.39.jpg (196.6 KB, 325 views)
  Reply With Quote
Old 5th January 2017, 02:57 PM   #13
kees52 is offline kees52  Netherlands
diyAudio Member
 
Join Date: May 2009
Location: Sprang-capelle Holland.
This delta sigma one is also a lot on the internet, however the high pulse rate is a problem, we need more 0001 to set 0011 or even 0111 to get lower rate with wider pulses, maybbe feed in in a pwm who make pulses wider etc etc, complicated stuff.

nice stuf there is going on.

But I have the circlotron parts so I go make it on a new PCB, the D I do keep watching and I play with it in spare time.

regards
Attached Images
File Type: jpg ScreenHunter_ Jan. 05 15.38.jpg (350.7 KB, 322 views)
File Type: jpg ScreenHunter_ Jan. 05 14.57.jpg (209.5 KB, 308 views)
  Reply With Quote
Old 5th January 2017, 07:15 PM   #14
Pafi is offline Pafi  Hungary
diyAudio Member
 
Join Date: May 2004
Location: Budapest
Quote:
Originally Posted by kees52 View Post
I did change the feedback parts, so oscillation did change, now it was very high, this way I do learn some things what happens, that is the fun of simulation. some papers has very nice tests, but I get negative offset in the discrete comparator.

Do you now a very fast output stage?
TAS... power stages. 50 ns delay, negligible dead time. There are many fast comparators.

Fast gate driver: UCC27200 and similar.

Quote:
For ltspice, I have design the allfet circlotron but when I measure in real time I get quite high harmonics, when simulation I get verly low distortions, possible because I did measure with the pc and the floating circlotron output things get messed up, I do need to build a differential to single output converter, then I can do it better, agree?
Yes, certainly a differential measurement amp is needed.

Quote:
,. But Maybe I did sim wrong, ...
This is also a possibility. Many things can go wrong.

When simulating, don't focus only 1 parameter! switching freq and distorition is 1 thing, but always check if it really work in switch mode, and doesn't make cross conduction! I think this is not the situation here. I asked you to change R22 and 24, but no response.

I can't help you in switching freq. This feedback is too complicated to analyse in head, and I don't have the simulation file and no time to play with it.

Chocoholic made very through research, worth to read his posts. The others in ex-UcD 1200W topic don't understand anything from electronics.
  Reply With Quote
Old 5th January 2017, 09:56 PM   #15
kees52 is offline kees52  Netherlands
diyAudio Member
 
Join Date: May 2009
Location: Sprang-capelle Holland.
Thank you Pafi for your responds.


I do not need much help and you have more things to do then help me, I am already thankfull to give me some hints about the D amps, she can be very complicated.

I do find 105 102 nS for the powercomparor quite fast, as say someone that it is old, the most amplifier designs are old, even D come from 1952.

Oke, Last quenstion R22 and 24 from what schematic? I did not read it because I have not seen it, that is why I have this tread, so things get not everywhere .

As for the Tas power modules these are noce but I like the challence of discrete stuff, and special try to lower distortions, so I di read a lot.. for discrete power modules I do set dead time with the gate resistors, this works fine and is simple.

Maybe this gate driver is better, because of higher voltages., IR2011S

regards

Last edited by kees52; 5th January 2017 at 10:05 PM.
  Reply With Quote
Old 6th January 2017, 04:16 AM   #16
sgpu is offline sgpu  United Kingdom
diyAudio Member
 
Join Date: Jan 2012
Location: Nottingham, UK.
One to try

OPAMP ---> LT1671---

---> NC7SZ08 ---> HI --> DRIVER
&
---> NC7SZ02 ---> LO -------^

Just before the output inductor feed back to the opamp in front of the LT1671.

http://cds.linear.com/docs/en/datasheet/1671fs.pdf

Last edited by sgpu; 6th January 2017 at 04:39 AM.
  Reply With Quote
Old 6th January 2017, 12:38 PM   #17
kees52 is offline kees52  Netherlands
diyAudio Member
 
Join Date: May 2009
Location: Sprang-capelle Holland.
Quote:
Originally Posted by sgpu View Post
One to try

OPAMP ---> LT1671---

---> NC7SZ08 ---> HI --> DRIVER
&
---> NC7SZ02 ---> LO -------^

Just before the output inductor feed back to the opamp in front of the LT1671.

http://cds.linear.com/docs/en/datasheet/1671fs.pdf
Thanks mate

I have test with a comparator who is 2.9 nS but has max 5 volts supply, reason for such fast one was the triangle generation together with a fery fast opamp, for sound integrator I did use the opa365, I have this from a datasheet and distortion with it can be 0.009 while goal was 0.1, without feedback, with feedback we get pretty low distortion.

I did read balanced output did cancel also some distortion special even ones.

Why delta signa do not get used more, I did read here also that digital D is dead end, and that self oscillating versions get more and more feed on the groud.

For as the gate drivers, I see that these ones are for smps supplys, and here things are different so is in smps not swinging signals like modulated pwm signals, and so chips can introduce harmonic distortion.

in picture one and two an weld smps with resonant zvs action build for mine welder who has a burnt transformer, and last picture a schematic of a class b like clipped amplifier used as a output stage who is from diyaudio somewhere. I also did question of a circlotron like output stage testing with dopuble coils and a feedback opamp for balanced to unbalanced..

dead time if for a audio amp also a issue because of it has to be very small, so need of very fast risetimes to get it small.

NC7SZ08 has small current so gate charge uncharce is a issue we need IR types of drivers.


regards
Attached Images
File Type: jpg ScreenHunter_924 Jan. 24 13.54.jpg (367.6 KB, 281 views)
File Type: jpg ScreenHunter_923 Jan. 24 13.53.jpg (279.8 KB, 91 views)
File Type: jpg ScreenHunter_ Jan. 03 20.42.jpg (309.5 KB, 92 views)

Last edited by kees52; 6th January 2017 at 01:01 PM.
  Reply With Quote
Old 6th January 2017, 01:00 PM   #18
sgpu is offline sgpu  United Kingdom
diyAudio Member
 
Join Date: Jan 2012
Location: Nottingham, UK.
Here's a useful fast drive. PWM input with deadtime adjustment via resistor. 8ns progagation delay. 1ns second rise and fall.
http://www.psemi.com/pdf/datasheets/pe29100ds.pdf

Last edited by sgpu; 6th January 2017 at 01:05 PM.
  Reply With Quote
Old 6th January 2017, 01:04 PM   #19
kees52 is offline kees52  Netherlands
diyAudio Member
 
Join Date: May 2009
Location: Sprang-capelle Holland.
Quote:
Originally Posted by sgpu View Post
PWM input and deadtime adjustment via res. 8ns progagation delay.
http://www.psemi.com/pdf/datasheets/pe29100ds.pdf
This is a very fast driver? quite new, and maybe the gallium mosfets are a choise in future because speed is a problem with class D amps.
  Reply With Quote
Old 6th January 2017, 01:39 PM   #20
sgpu is offline sgpu  United Kingdom
diyAudio Member
 
Join Date: Jan 2012
Location: Nottingham, UK.
For my class-D amp project I've been looking at the following isolated design.

PWM -> NVE isolator -> dead time circuit -> driver -> fet -> speaker

Isolated surface mount PSU 12Vin, 9Vout - PDS1-S12-S9-M

Before Isolation
LDO - 12V down to 5V to power primary side of NVE isolator

After Isolation
LDO - 9V down to 5V for secondary side of NVE isolator
LDO - 9V down to 6V to power the fet driver

A little bit on the crazy side but I'm always up for doing something "different".

Last edited by sgpu; 6th January 2017 at 01:54 PM.
  Reply With Quote

Reply


Class D investigation.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
Cathodyne investigation - stop the presses! exeric Tubes / Valves 40 29th June 2016 07:02 AM
Leaving class A how to make this amp leave class A to AB when volume is increased? rhythmsandy Solid State 3 14th June 2016 02:16 AM
Contest: Linear Power Amp in a mint tin (class Aa, class AB, or class B) danielwritesbac Chip Amps 206 19th January 2013 09:52 PM
Sony system failure - full investigation thread dmx2020 Solid State 11 19th October 2012 06:59 AM
My folded-cascode VAS linearity investigation GK Solid State 96 19th July 2010 11:01 PM


New To Site? Need Help?

All times are GMT. The time now is 08:17 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio
Wiki