1kW Gen2 - Page 3 - 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 17th December 2007, 02:20 PM   #21
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
Aloha ...nice pictures everywhere...
OK, I'll also send some pictures.
I kicked my lazy xxxx during the weekend and converted the naked half bridge to something like a class D amp.
Triangle, LM160 comparator, discrete level shifter, multiple shut down functions.
As always a breadboard not as sexy as most a proper PCBs , but Choco-handmade
Attached Images
File Type: jpg modshift.jpg (68.7 KB, 1914 views)
  Reply With Quote
Old 17th December 2007, 02:21 PM   #22
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
...a short look to the clipping behavior.
I am supplying all auxiliaries from my small flyback, so all supplies remain stable no matter how much clipping you drive. Nice to get blameless clipping bahvior. The capacitive couplings of my flyback are small (15pF), so easy fast floating for the high side supply is no issue.

Here a screen shot in low frequency clipping with 2.5Ohms load.
Ok, Ok, for my first examinations I am using a weak power supply which is dropping from +/-50V down to +/-28V...

Please note I used a 1:10 probe, so we have 10V/grid.
Attached Images
File Type: jpg lfclip.jpg (49.9 KB, 1672 views)
  Reply With Quote
Old 17th December 2007, 02:23 PM   #23
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
.....a 1kHz sine wave short below clipping.
Again with 2.5 Ohms load.

1:10 probe ==> 10V/Grid
Attached Images
File Type: jpg sin1k.jpg (58.1 KB, 1677 views)
  Reply With Quote
Old 17th December 2007, 02:24 PM   #24
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
My output filter is 5uH, 3uF and a damper with 10uF/1.7 Ohms.
Feedback is splitted partially from halfbridge and partially behind the filter, this helps to optimize the transient response.

Here a screen shot of running a 1kHz rectangular signal into 2.5 Ohms load.

Probe 1:10 ==> 5V Grid
Attached Images
File Type: jpg rect1k.jpg (46.2 KB, 1645 views)
  Reply With Quote
Old 17th December 2007, 02:34 PM   #25
Eva is offline Eva  Spain
diyAudio Member
 
Eva's Avatar
 
Join Date: Oct 2003
Location: Near the sea
Send a message via MSN to Eva
Hey, this is looking very good, but why such a low output inductor value? 10uH should be good enough for 2 ohm operation.
__________________
I use to feel like the small child in The Emperor's New Clothes tale
  Reply With Quote
Old 17th December 2007, 04:17 PM   #26
diyAudio Member
 
Join Date: Nov 2007
ChocoHolic:
Post like yours are what makes diyAudio such an incredible forum.
Discussing candidly your design merits and tradeoffs, providing full schematics, and a running project history....

Thanks for your efforts, and congratulations on a design which shows lots of promise....

Your square wave test shows a very well behaved circuit. Perhaps I would only use a faster timebase to better inspect the edges.
  Reply With Quote
Old 18th December 2007, 10:32 AM   #27
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
Hi Eva,

yes I have several reasons to use such a low L.
1. How about driving ESL speakers? Most class D output filters will go to dramatic mistuning (and quite low roll off frequency), when loaded with several uF capacitive load. A design with low L and large C can serve such loads much better.
2. Distorsions.
3. Hardswitching only during really high power passages.
4. Low output impedance even at high frequencies.

With roughly +/- 8A filter the ripple half bridge will almost always operate in inductive load condition.
Means I can systematically avoid the dead time distorsions during 99% of my listening time.
The ripple is still low enough to keep losses uncritical.
Overall idle power consumption is 6W.
Whereof I know that already 2x1W are the snubbers, may be 0.5W.. 1W in my flyback, roughly 1W in the output filter...
Of course you have to make sure that your rail e-caps are OK for the resulting effective cap ripple current of roughly 3.3Arms, but that's no problem. I am using 2x8x100uF nonames. No self heating visible. Also the specs for low cost series of global brands allow 500mA..600mA for a 100uF/100V cap (if I remember right...).
My filter choke is a gapped ferrite design with HF litz wire. With a metal powder core you might run into issues for this filter design. At 320kHz the losses of metal powder would allow just very low flux densities.

Hi Fernando,
thanks. It's nice to hear that you appreciate my postings.
...ah well, of course I had a look with faster time base...
And of course also with higher frequency. Frankly speaking my previos post was more for 'marketing' and making people cuorios.
If you look at the next screen shot with 20us time base and a 10kHz / 24Vpp square wave, then it becomes obvious that these results are good, but not perfect.
Attached Images
File Type: jpg rect10k.jpg (52.3 KB, 1288 views)
  Reply With Quote
Old 18th December 2007, 02:08 PM   #28
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
...and here the schematic without shut down circuits. In fact I am using the NI IN pins for shut down functions....
So this schematic is showing just the core of the amp.
I am pleased about the propagation delay of just 120ns from comparator input to halfbridhe output. On the other hand it turned out that this is only helpful for the feedback from HB and this feedback is limited much ealier by the requirement that the resulting dv/dt at in1 must be less than the dv/dt from the triangle at in2 of the LM160. From simulation it looks like you can get almost identical results no matter if your over all propagation delay is 100ns or 300ns. So the typical integrated halfbridge drivers would probably be OK as well.
For the feedback from behind the filter you can neglect the influence of some ns propagation delay, here the LC is dominant.
In fact the layout is always a concern in class D and nobody really wants to see my messy breadboard. But I will go on with some optimizations and tests, before designing a nice PCB.
Attached Files
File Type: pdf gen2discrete.pdf (28.7 KB, 546 views)
  Reply With Quote
Old 18th December 2007, 02:18 PM   #29
diyAudio Member
 
Join Date: Nov 2007
Quote:
Originally posted by ChocoHolic

Hi Fernando,
thanks. It's nice to hear that you appreciate my postings.
...ah well, of course I had a look with faster time base...
And of course also with higher frequency. Frankly speaking my previos post was more for 'marketing' and making people cuorios.
If you look at the next screen shot with 20us time base and a 10kHz / 24Vpp square wave, then it becomes obvious that these results are good, but not perfect.
I'm still very impressed with the squarewave performance.
I can't read the scope photo very well, but it seems that the amp's slew rate is ~ 24/5E-06 or roughly 4.8 volts per microsecond. Not bad, not bad at all!

Also the overshoot indicates minimal ringing, it settles within 2 or 3 switching cycles, very close to critical damping. NICE! You got it very close to optimal.
However if you have the time and are very perfectionist, you could try some snubbering at the output to dampen it further. Don't know if this would provide any sonic improvement, perhaps only in percussive sounds.
  Reply With Quote
Old 18th December 2007, 03:01 PM   #30
fumac is offline fumac  China
diyAudio Member
 
Join Date: Oct 2006
Send a message via MSN to fumac
ChocoHolic
ur design is diffrent with the old one
i like halfbridge more than full bridge

singing now ?
good job

fumac
  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
Modding the Gen2 Sonic Impact loninappleton Chip Amps 2 8th May 2008 01:21 AM
Sonic impact T-amp gen2 problem andywhale Class D 1 21st April 2008 11:54 AM


New To Site? Need Help?

All times are GMT. The time now is 03:42 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