Simple Symetrical Amplifier - Page 242 - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

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 16th April 2012, 06:57 AM   #2411
diyAudio Member
 
Join Date: Aug 2009
Default curious....

Quote:
Originally Posted by hahfran View Post
I have an even faster topology which is an inverting current feedback but with that it is difficult to take advantage of the therm trak.
Hi Hafran,

about the amp.I have been following this as they have said that it is fast...Is this mean that this amp can process a signal faster than a typical amp.If so does it guarantee or have a relation that It will provide good sonics.and can a speaker cope up with the speed of this amp.?

regards
joel

Last edited by drowranger; 16th April 2012 at 07:01 AM.
  Reply With Quote
Old 16th April 2012, 08:35 AM   #2412
hahfran is offline hahfran  Germany
diyAudio Member
 
Join Date: Dec 2007
Quote:
Originally Posted by drowranger View Post
Hi Hafran,

about the amp.I have been following this as they have said that it is fast...Is this mean that this amp can process a signal faster than a typical amp.If so does it guarantee or have a relation that It will provide good sonics.and can a speaker cope up with the speed of this amp.?

regards
joel
the fastest electro-dynamic speakers have a rise time that equals about 20 Volts / µsec . This applies only for small excursions. And is valid for the most expensive tweeters . besides what is the max rise time that a digital sample with 96 kHz sampling frequency can reproduce? Rough estimate: One quarter of a full period takes 6 µsec. Assuming 10 Volts gives 10V/6 µsec= 1.7 Volts/µsec.
  Reply With Quote
Old 16th April 2012, 08:43 AM   #2413
mikelm is offline mikelm  England
diyAudio Member
 
Join Date: May 2003
Location: England
Send a message via Yahoo to mikelm
Yeah - 1.7V / uS - sad but true
  Reply With Quote
Old 17th April 2012, 04:15 AM   #2414
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
I was curious about an idea I had so I made a simulation based on the most recent reasonable schematic I could find in post 2383. I thought I would share. With some drivers for the outputs I think the GBWP could be increased. The open-loop gain is plotted without a resistive load. This was just for fun, I did not worry about thermal stability, so it needs thermal compensation before it could be built (probably facilitated by a CCS tempco).
Attached Images
File Type: png SSA_Keanmods.png (131.7 KB, 486 views)
  Reply With Quote
Old 17th April 2012, 09:11 AM   #2415
diyAudio Member
 
Lazy Cat's Avatar
 
Join Date: Sep 2007
Quote:
Originally Posted by keantoken View Post
I was curious about an idea I had so I made a simulation based on the most recent reasonable schematic I could find in post 2383. I thought I would share.
Hi keantoken, thanks for sharing

3 MHz/-3db CLG is a real world repeatative result in many measurements I performed.

Did you find Q9, Q10 feedback buffers to be as improvement here, cause feedback loop is low impedance anyway. R19, R25 are also not necessary to my opinion and yes drivers makes a difference also. Not to mention VAS cascodes ...

P.S. Are Ccomp+R better performer here than usual Miller cap compensation?

Last edited by Lazy Cat; 17th April 2012 at 09:15 AM.
  Reply With Quote
Old 17th April 2012, 10:16 AM   #2416
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
I copied the schematic as quick as I could, I could have left out those resistors including R14 but it didn't concern me.

The feedback network may have low impedance but it got eaten up quickly with the output stage capacitances. The Cjc of the cascodes and of the outputs are the biggest limits of my schematic.

I've found that RC shunt compensation like this does a much better job at stabilizing the amp, not just this one but most amps I've simulated. This is because Shunt compensation absorbs HF energy (like a snubber), not just relocating it, potentially creating pole-mania. If you look at Miller comp and take the VAS away, you can see the miller cap injects positive feedback into the output stage. If the VAS and OPS were perfect this would not be a problem, but it is. You can exaggerate the effect if you degenerate the VAS heavily. Shunt compensation loads the input stage less at HF and is therefore also more immune to RFI. Miller compensation as it is usually applied to the VAS also decreases PSRR; Shunt compensation usually avoids this problem.

But there are some good reasons this type of compensation is not used often. #1, most people don't understand it and it's more tedious to adjust than Miller compensation. #2, it causes more distortion. With high-impedance loads it actually distorts less than Miller, but with speaker loads it distorts more. #3, Miller compensation gives the appearance of making a faster amp. In the simulator AC Analysis it looks like it but we already know Miller compensation can have trouble with RFI, and the input stage will saturate with any significant signal at those frequencies. #4, a gain hump near the -3db point is an unfortunate side-effect of shunt compensation. This is not ringing, just a hump, and it will show up as overshoot on the square wave test with it's own unique shape.

VAS shunt compensation has many practical benefits, but it's a bummer to the distortion specs. It works great but it's an "ugly duckling". Even so, I prefer it sonically over Miller compensation. It seems to have more realism, better voice legibility, and more soundstage depth. Go ahead and try it.

If you want to see a hidden example of shunt compensation, look at Symasym.

Attached is the simulated response of my circuit. Again without load. A load changes things significantly here, because the lack of buffering of the output stage. Still, it's an indicator of what can be achieved.

PS. The PSRR caveat with Miller comp could be solved by running the Miller caps to the IPS cascode emitters rather than the VAS bases. It is risky but it could perhaps be made to work.
Attached Images
File Type: png SSA_keanmods_Response.png (130.8 KB, 431 views)
  Reply With Quote
Old 17th April 2012, 01:10 PM   #2417
diyAudio Member
 
Join Date: Dec 2008
Location: Ljubljana, Slovenia
LazyCat, any progress with GB for boards?
  Reply With Quote
Old 17th April 2012, 01:46 PM   #2418
diyAudio Member
 
Lazy Cat's Avatar
 
Join Date: Sep 2007
Yes I know about this kind of serial RC compensation at input nod of the VAS. I used it already in my amps before and at a start I always put both R&C as trimmers, find the best spot in between lower and upper frequency oscillation, measure the values and solder fixed values RC parts on PCB. This kind of calibration is always fun and very acurate job, no problem to find optimum trimmers positions.

But you said that distortions are higher by this type of compensation, how much higher is that going to be?

Quote:
Originally Posted by keantoken View Post
Attached is the simulated response of my circuit. Again without load. A load changes things significantly here, because the lack of buffering of the output stage. Still, it's an indicator of what can be achieved.
Very nice response for both gain and phase shift, good enough for you to make it in real someday hehe


Quote:
Originally Posted by keantoken View Post
PS. The PSRR caveat with Miller comp could be solved by running the Miller caps to the IPS cascode emitters rather than the VAS bases. It is risky but it could perhaps be made to work.
Will try it next time, thanks for the tip, regards Andrej
  Reply With Quote
Old 17th April 2012, 01:59 PM   #2419
diyAudio Member
 
Lazy Cat's Avatar
 
Join Date: Sep 2007
Quote:
Originally Posted by igor0203 View Post
LazyCat, any progress with GB for boards?
Hi igor0203

Well the amp is fully functional, playing music nicely, but it is still in R&D phase, so some more improvements will be tested before to go to conclusion of the project and say, this is the final SSA version.

What's happening now is an upgrade for CCS current injectors to SSA feedback bridge, since it was realized that the distortion spectrum is much more clearer with the use of high performance CCSs. And these are ready to build and installed into the amps. Both positive and negative CCSs are located on PCB size 30 x 45 mm (attached) and will be soldered to main PCB as add-on module. Will look quite professionally I tell you.
Attached Images
File Type: jpg SSA CCS PCB.jpg (398.0 KB, 387 views)

Last edited by Lazy Cat; 17th April 2012 at 02:07 PM.
  Reply With Quote
Old 17th April 2012, 02:11 PM   #2420
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
Quote:
Originally Posted by Lazy Cat View Post
But you said that distortions are higher by this type of compensation, how much higher is that going to be?
At 10KHz, which I think is a frequency representing compensation quality, distortion increases from .031% to .064%. I don't know whether output drivers will bring these figures closer. Distortion at 1KHz changes only barely from .0305 to .31. This is at 12V into 6R.

Quote:
Very nice response for both gain and phase shift, good enough for you to make it in real someday hehe
That sounds good to me, but right now I am invested in my own design.

Quote:
Will try it next time, thanks for the tip, regards Andrej
I found 100R resistors in series with each Miller cap damped a resonance around 50MHz; could be much worse in real life.
  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
Symetrical schematics are alike plague in Brazil, do you like them? destroyer X Solid State 151 1st July 2010 02:09 PM
Symetrical out low Zout karsten21 Tubes / Valves 5 1st February 2010 10:19 PM
going balanced/symetrical - what benefits? weissi Solid State 15 20th October 2007 08:06 AM
Symetrical field, is this? Raka Multi-Way 6 14th September 2003 01:21 PM
Non-symetrical SMPS output cm961 Parts 4 21st August 2003 10:27 PM


New To Site? Need Help?

All times are GMT. The time now is 11:32 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