SystemD_2kW, any interest for an open design? - Page 29 - 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 23rd February 2013, 10:46 AM   #281
diyAudio Member
 
Workhorse's Avatar
 
Join Date: Dec 2004
Quote:
Originally Posted by ChocoHolic View Post
At Workhorse:
20kHz clipping into 2R is not the most realistic situation, but the behavior there could point to troublesome short comings. Ok, check will come.
Yes, the behavior at 20kHz is what i need to see.
  Reply With Quote
Old 23rd February 2013, 11:35 AM   #282
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
...you 'need'...
Is it a deperate 'need'?

Here it is.
The 2R situation is boring. More interesting is the behavior with 8R, means the filter is not getting much damping from the speakers. All the filter gain and natural overshooting of the filter has to be catched by the amp.
Which of course lacks perfection, because during clipping the feedback is partially disabled...
All measurements with 1:10 probes.

- 20kHz clipping into 2R
- 10kHz clipping into 2R
- 20kHz clipping into 8R
- 10kHz clipping into 8R

P.S.
The 20kHz load situations are clearly beyond the allowed continuous load of the output MKPs, but during my short test times they remained cool.
Attached Images
File Type: jpg clipping20khz2R.jpg (632.8 KB, 352 views)
File Type: jpg clipping10khz2R.jpg (687.5 KB, 315 views)
File Type: jpg clipping20khz8R.jpg (694.8 KB, 276 views)
File Type: jpg clipping10khz8R.jpg (688.6 KB, 246 views)
  Reply With Quote
Old 23rd February 2013, 11:56 AM   #283
diyAudio Member
 
Join Date: Feb 2013
Location: brasil
oi bom dia é a primeira vez que posto aqui e queria que vocês me corrigi se se estiver fazendo errado.Pois tenho varios esquemas de amplificadores classe D de uma marca industrial que vai de 2x75 rms a 30kv .
  Reply With Quote
Old 23rd February 2013, 12:05 PM   #284
tinitus is offline tinitus  Europe
diyAudio Moderator R.I.P.
 
Join Date: Nov 2005
vilmar, our rules ask to speak english, please


moderation
  Reply With Quote
Old 23rd February 2013, 12:14 PM   #285
diyAudio Member
 
Workhorse's Avatar
 
Join Date: Dec 2004
Quote:
Originally Posted by ChocoHolic View Post
...you 'need'...
Is it a deperate 'need'?

Here it is.
The 2R situation is boring. More interesting is the behavior with 8R, means the filter is not getting much damping from the speakers. All the filter gain and natural overshooting of the filter has to be catched by the amp.
Which of course lacks perfection, because during clipping the feedback is partially disabled...
All measurements with 1:10 probes.

- 20kHz clipping into 2R
- 10kHz clipping into 2R
- 20kHz clipping into 8R
- 10kHz clipping into 8R

P.S.
The 20kHz load situations are clearly beyond the allowed continuous load of the output MKPs, but during my short test times they remained cool.

Yes ........

Thanxz,

Actually i wanted it to compare with mine.......

I will shortly conduct a 30kHz test in 2,4,8 ohm, open loads and then will post here.
  Reply With Quote
Old 23rd February 2013, 12:32 PM   #286
diyAudio Member
 
Join Date: Feb 2013
Location: brasil
hi beleza?é the first time that put here and if it is wrong I ask that me desculpem.é that I have several outlines of amplifiers class D that are going from 2x75rms to 30kv and if I am willing the fornecelos to who interests.
  Reply With Quote
Old 23rd February 2013, 01:24 PM   #287
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
It is always a matter of taste how one adjusts a class D amplifier.
THD, IMD, output filter/ fs/ carrier residuals at the output, step response.
It is my taste to balance them all in moderate way, without excessively focussing on just one property and ignoring all the rest.
I already have posted the THD, IMD and carrier residuals earlier, here the reproduction of rectangular signals.

At low and medium signal levels the rectangular reproduction is fast and blameless.
Limitations become visible only at large rectangles, which are not part of music program.
Despite being not part of music program also the CD format is not capable to deliver such signals, but for amp testing it is important to understand the limitations.
To speed up and catch back the filter fast, theory tells that the drive in front of the filter has to provide voltages which which are far larger (in both directions) then the filter output. Naturally the voltage that you can provide to the filter is limited by the supply rail of the amp. Means at high output levels there is no way to provide the theoretically needed huge signals at the filter input and consequently overshoots will happen – so you can chose to have the system generally slower, or fast and blameless at lower levels but with overshoot high levels with high load impedances.
Low load impedances will dampen the filter and avoid overshoot even at high levels.

Attached the screen shots with 1:10 probes.
Picture 1&2
Up to 25Vpp rectangles are reproduced blameless into all load impedances.

Picture 3&4
At 60Vpp the 2R is still driven blameless, but with 8R the overshoot is obvious.

Picture 5&6
Trying to reproduce a rectangle with 114Vpp the 2R situation has no overshoot, but the rise time has increased due to the fundamental limitations that we get from the filter and available rail voltage.
This 114Vpp with a 10kHz rectangle is close to the shut down of the used power supply, note that the amp delivers already about 1.5kW in this situation.
Pushing the amp with 8R to such levels of rectangular signals is resulting in massive overshoots, but still working in a controlled way.
Attached Images
File Type: jpg 25Vpp_2R.jpg (671.9 KB, 232 views)
File Type: jpg 25Vpp_8R.jpg (701.0 KB, 25 views)
File Type: jpg 60Vpp_2R.jpg (679.2 KB, 23 views)
File Type: jpg 60Vpp_8R.jpg (661.3 KB, 23 views)
File Type: jpg 114Vpp_2R.jpg (656.4 KB, 23 views)
File Type: jpg 120Vpp_8R.jpg (698.1 KB, 21 views)
  Reply With Quote
Old 23rd February 2013, 02:00 PM   #288
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
Hi Vilmar,
I am not sure what you are trying...
a) Asking me to do a custom design for you?
or
b) Offering to disclose your designs?
or
c) Selling your designs?

In case of a) we should check by PM.
In case of b) you should open a thread and show your designs.
In case of c) you should open a thread in the commercial section.
  Reply With Quote
Old 23rd February 2013, 05:57 PM   #289
Cresnet is offline Cresnet  Germany
diyAudio Member
 
Join Date: Nov 2012
@ChocoHolic

What about the power supply voltage? +-?

Why the amplifier suffer at the 8R ohms, while I see most amplifiers are suffering at 2R?

Are you going to make a group buy for this amplifier soon?

thanks
  Reply With Quote
Old 23rd February 2013, 09:05 PM   #290
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
The rails are pretty stable at high load impedances.
Different from what the people tell who want to sell you their scrap,
it is fact that all class D amps with such LC output filters do suffer in both directions.
Only at one load impedance the filter does something acceptable.
At higher load impedances the filter generates overshot and ringing.
At lower load impedances the filter will show slower and slower creeping of the signal. This can be catched by postfilter feedback.
The shown amp has a postfilter feedback with PID loop gain which is capable to speed up the step response about factor three vs the natural step response of the filter.
Furtheron this PID loop is free of overshoot at 2R up to no load as long as the transfer chain remains linear.
Different from what most control theory books make us believe, most real chains are nonlinear and have hard limits like supply rails. In order to avoid brute destructive oscillation modes you have to limit the action of the I portion during such internal clipping situations (have a look to Q205 and Q206). With this only uncritical overshoots are happening in load situations which cause pre filter clipping.
And even this could be fully avoided by adjusting the PID loop in a way that the rise time would end up around 8...10us.
Rise times around 10us are pretty slow for high quality audio and bear the danger to cover small fast nuances of the music. At the same time it is pretty obvious that no music is made of full amplitude rectangles. For this reason I am convinced that this adjustment is by far more musical than an adjustment which has a 10us rise time.
Regarding most class D amps in the market - you will find that most of them come along with rise times around 10us or longer and still show overshoot at most loads and even at low signal levels...
They just don't tell it to anybody and hope nobody knows how to use a scope.

Below two comprehensive screen shots.
Upper trace is the output signal with 8R load. Lower trace is the signal at the filter input.
This signal contains the information about the rail and the PWM.
Obviously the rails are stable.

In the first picture you see that the PWM is always pulsing and the PID loop can perfectly control the filter.

In the second picture you see that there are long times were the PWM is not pulsing any more but giving full power to speed up the filter and to catch back the over. Very clear pre filter clipping.
You also see that the PWM is switching to the opposite direction already before the output has reached the intended value. In order to avoid the overshoot you would need to shift this even earlier,
which means larger D portion in the feedback and slower step response.

All measurements with 1:10 probes.

P.S.
Group buy? I did not intend a group buy. I am a silly tech soul who has no idea about logistics.
But in special cases I can give support with one or the other component.
In any case I should prepare the builders package (schematics, BOM, PCB data, some hints) first and also for this I have to ask for patience.
Attached Images
File Type: jpg SmallRect_PWM.jpg (671.7 KB, 22 views)
File Type: jpg LargeRect_PWM.jpg (692.1 KB, 24 views)

Last edited by ChocoHolic; 23rd February 2013 at 09:06 PM. Reason: typo
  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
O2 Front Panel INTEREST POLL your own design Llama16 Group Buys 1 12th September 2012 06:23 PM
Probing interest in My inrush limiting design nattawa Power Supplies 2 19th November 2011 02:08 PM
Custom open baffle 12" woofer - any interest? SimontY Group Buys 2 4th March 2009 02:53 PM
Open baffle design lawbadman Multi-Way 1 17th January 2008 01:15 AM
Open call for suggestions on Open Source DIY Audio Design gfergy Everything Else 1 15th April 2007 08:33 AM


New To Site? Need Help?

All times are GMT. The time now is 01:17 AM.


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

Content Relevant URLs by vBSEO 3.3.2