SystemD_2kW, any interest for an open design? - 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 24th November 2012, 07:45 PM   #21
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
Quote:
Originally Posted by Workhorse View Post
Just uploaded the sim file, i hope you have LTC
Yes, I am also using LT Spice.
Thanks for putting more detailed.
0h, the current mirror at the output.... I have to think deeper...
  Reply With Quote
Old 24th November 2012, 07:56 PM   #22
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
I think the main issue regarding propagation delay of your comparator is that you are chasing Q5 and Q9 into saturation.
Could be improved by giving the output stage a negative rail and clamping the output voltage before Q5 or Q9 will saturate.
....have to think about asymetric slewing more detailed...
  Reply With Quote
Old 24th November 2012, 08:11 PM   #23
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
Quote:
Originally Posted by Workhorse View Post
I think a Hysteresis modulator with a CCS control [i have seen this approach somewhere] is best option in order to mitigate any effects of voltage fluctuations. This will give more precision and control over how it actually performs.
If you find the link it would be great.
From your short description, I am not fully getting the idea behind.
  Reply With Quote
Old 24th November 2012, 09:03 PM   #24
pergo is offline pergo  Italy
diyAudio Member
 
pergo's Avatar
 
Join Date: Dec 2008
if you want DIL8 fast comparator, try LT1016 (LTspice has the model) or AD8561

In your opinion, better the discrete solution?
  Reply With Quote
Old 24th November 2012, 09:45 PM   #25
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
Hi Pergo,
yes in my opinion for class D it is better to go for a discrete comparator rather than ultra fast high end comparators.
The issue is noise.
The combination of speed, high gain and integrated mini hysteresis (usually not specified) form an extremely efficient HFjunk-to-NFnoise converter.
I did not measure the LT1016, but I experimented with the LT1711, which is an amazing beauty, but simply not the right thing for class D audio amps.
Somewhere here in the forum I had thread about jitter measurement...

I am considering the LM160 as integrated solution, because it is easily available almost everywhere, easy to use, medium speed.
In all my measurements better suited for class D than LT1711.
In order to see it in a real life situation which allows comparison vs a discrete solution I modified my SystemD_MD and got still pretty good results with noise figures that were similar in the lower frequency range and increasing over the audio band, going to approx 10db worse than my discrete solution at the upper audio band.
Overall the noise also with the LM360 the noise is not audible on the speakers (also when putting ears close to the tweeter).
I have to plug my headphones directly to the 600W output in order to hear it.
  Reply With Quote
Old 24th November 2012, 10:15 PM   #26
pergo is offline pergo  Italy
diyAudio Member
 
pergo's Avatar
 
Join Date: Dec 2008
for easy jfet solution, you can use LSK389. It's a dual matched 2SK170 in SOIC or TO71
No complementary
  Reply With Quote
Old 25th November 2012, 08:51 AM   #27
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
Right. Do do you have real life experiences with them?
I remember someone reported issues of that dual version with the common floating substrate....
For dual P-Chanel the 2SJ109 could be considered.
On the other hand I am pretty confident that matching can achieved in a
sufficient way with the single types.


Back to the impact of the comparator to the noise figures...:

Attached some comparative measurements with LM360 and SystemD_MD.
In classD it is important to watch idle noise and operational noise, because with increasing load currents the HF disturbances increase and will also affect the NF noise.

First picture is the operation of the original SystemD_MD with discrete comparator, running 60W into 2 Ohms.

Second picture is the comparative operation of original version and implanted LM360 both running 60W into 2 Ohms. Don't panic about increased distortion, I think this could be readjusted by the shaping network. Also noise can be inproved by approx 5db with a 47pF cap directly across the comparator input.

Third picture is the idle noise of both versions. Shown decibel relate to 60W again. This is fitting to the findings with the headphone that both versions are pretty silent.
Attached Images
File Type: jpg SystemD_MD_original.JPG (89.6 KB, 953 views)
File Type: jpg Orig_vs_LM360.JPG (89.8 KB, 897 views)
File Type: jpg Noise_orig_vs_lm360.JPG (89.9 KB, 825 views)

Last edited by ChocoHolic; 25th November 2012 at 08:55 AM. Reason: typos & grammar
  Reply With Quote
Old 25th November 2012, 09:19 AM   #28
pergo is offline pergo  Italy
diyAudio Member
 
pergo's Avatar
 
Join Date: Dec 2008
I don't have any experience in jfet input for comparator stage in class D.
I've used them only for linear amplifiers. So i cannot help you more.

I will try to simulate some discrete comparator by myself, than i will post you and we can discuss about them, if you want.
  Reply With Quote
Old 25th November 2012, 05:32 PM   #29
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
Hi Pergo,
I think a discrete comparator plugin for this amp is a second step for upgrading, if desired. Similar like using High End OpAmps, or discrete OpAmps.

Hm, what do you (and all) think?
Should I add an overvoltage protection, which shuts down the amp in case any rail exceeds 95V?
Most likely my set up will not need it. Going to use a Hypex 1200 with synchronous rectifiers. Rail pumping will be no issue. Also between +/-82V towards the critical +/-100V there is enough margin to cover normal fluctuations of the mains voltage in Germany.
But depending on the psu design and country such a shut down might be a good idea.
The additional efforts for a precise shut down would be:
Two voltage refs, one dual Op amp, some small resistors and caps, one small NPN.
It could be an option on the PCB, which can remain unused without impact on the amp itself.
  Reply With Quote
Old 25th November 2012, 05:44 PM   #30
pergo is offline pergo  Italy
diyAudio Member
 
pergo's Avatar
 
Join Date: Dec 2008
you can use some zeners in series from Vrail to a NPN base.
When the voltage exceed Zener_total_voltage + Vbe_on, the NPN conducts and shutdown everything you want.
For example 3x33V or 1x100V.
It's a rugged circuit, not for purist
  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 05:23 PM
Probing interest in My inrush limiting design nattawa Power Supplies 2 19th November 2011 01:08 PM
Custom open baffle 12" woofer - any interest? SimontY Group Buys 2 4th March 2009 01:53 PM
Open baffle design lawbadman Multi-Way 1 17th January 2008 12:15 AM
Open call for suggestions on Open Source DIY Audio Design gfergy Everything Else 1 15th April 2007 07:33 AM


New To Site? Need Help?

All times are GMT. The time now is 08:30 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