The MONGREL (supersym II) - Page 102 - 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 20th December 2010, 07:18 AM   #1011
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
Hugh, I emailed you.

Perhaps the idea has merit, but the diode has only several uA flowing through it. Wouldn't this keep its voltage drop fairly low too? And how about trying a more capacitive, beefy diode? One could also use a third transistor, as per below. I also added a second variant for thought experiment - now the second transistor is the one with shorted C-B, the opposite extreme.

- keantoken
Attached Images
File Type: png CMVce.png (11.5 KB, 478 views)
  Reply With Quote
Old 20th December 2010, 08:27 AM   #1012
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
After some sims...

I tried to equalize Vcb for both transistors in the third example. Top graph is gain error. Bottom graphs are gain of the different mirrors.

Seemingly, Mirror 2 would give the lowest blanket THD, but Mirror 3 would give the lowest high-order harmonics. In fact, Some odd harmonics are even canceled in mirror 3. FFT last attachment.

So it appears that the performance of the CM will depend on what you put at the first transistor's collector.

- keantoken
Attached Images
File Type: png CMcompareSchem.png (70.4 KB, 478 views)
File Type: png CMcompare_graph.png (54.4 KB, 459 views)
File Type: png CMcompareFFT.png (47.9 KB, 449 views)

Last edited by keantoken; 20th December 2010 at 08:37 AM.
  Reply With Quote
Old 20th December 2010, 12:43 PM   #1013
diyAudio Member
 
ostripper's Avatar
 
Join Date: May 2008
Location: Albany , NY (smallbany)
Quote:
Originally Posted by keantoken View Post
After some sims...

I tried to equalize Vcb for both transistors in the third example. Top graph is gain error. Bottom graphs are gain of the different mirrors.

Seemingly, Mirror 2 would give the lowest blanket THD, but Mirror 3 would give the lowest high-order harmonics. In fact, Some odd harmonics are even canceled in mirror 3. FFT last attachment.

So it appears that the performance of the CM will depend on what you put at the first transistor's collector.

- keantoken
Thanks for that , KT. It looks like the wilson will be in the AX1.2/BX1.4 CM, one more $.11 device will add to the other "micro-improvements". The other factor is that Nico'e ELD also successfully incorporates the wilson.
"GK" .. way back ... used the widlar enhanced CM. I tried it and did not like the results.
OS
__________________
Mongrel website , always current and updated :
http://www.fidelityforce.com/ostripper

Last edited by ostripper; 20th December 2010 at 12:47 PM.
  Reply With Quote
Old 20th December 2010, 07:52 PM   #1014
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
Let me explain something I found about current mirrors. At low current excursions, the waveform peaks sharply because of the nonlinear diode characteristic (transconductance is low thus Dv/Dt is higher at the time input current changes direction). This peak induces pulse-like currents into the parasitics surrounding the device and can be the cause for some W_T_F moments (such as a "blip" that's always at the top of the output waveform). Interestingly, the simple two-transistor mirror is immune to this effect, unless you choose transistors with large Cbe (which should be avoided any high-performance mirror anyways).

For experiment, I tried adding degeneration to third transistor of the Wilson mirror. It still doesn't work quite as well as the diode variation. So it still stands that the harmonic profile and THD of the Wilson current mirror will depend on what you put at the first transistor's collector. In this case, a diode works better than anything else in simulation. See the attached plots for why. The orange trace is the one for the Vcb of Q6. As you can see, the diode distortion is curved the opposite direction... So apparently there is cancellation going on.

The lowest plot is the time derivative of the Vcb of Q6. This is representative of the current through any parasitic capacitor that sees this voltage. At 100KHz, full power, this can be a problem.

- keantoken
Attached Images
File Type: png CMCompare_Artefact.png (66.4 KB, 371 views)

Last edited by keantoken; 20th December 2010 at 08:14 PM.
  Reply With Quote
Old 20th December 2010, 08:09 PM   #1015
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
Hey, I'm finding all my silicon diode models show decreased transconductance compared to actual transistors. Is this normal?

- keantoken
  Reply With Quote
Old 21st December 2010, 02:26 AM   #1016
diyAudio Member
 
ostripper's Avatar
 
Join Date: May 2008
Location: Albany , NY (smallbany)
Default I WON'T give up on the CX.

Quote:
Originally Posted by keantoken View Post
Hey, I'm finding all my silicon diode models show decreased transconductance compared to actual transistors. Is this normal?

- keantoken
Of course , the standard diode models are very basic. Some of the MUR (ON) downloaded diode models I have behave very differently.

On to my recent FAILURE , I already have something better than the "goldmund killer" I just blew up (the cx) . Below is the best non-tmc'ed topology I have simulated. It is composed of a standard single ended LTP/cascode w/ current source. It THEN drives a full complimentary (leech) style differential set (like electrocompianet does) which drives a symmetrical dual differential.
WOW!! a lot of freakin' transistors. BUT , the performance matches a FULL TMC blameless and even exceeds it above 20K . over 100db loop gain and 125v slew. (below 1-3) PS , distortion is so low because BOTH even and odd are cancelled out in this one !

Stability is rock solid by either compensating the first differential or lead compensation from the VAS. I'm a gonna build it. A 100 x 100 mm VB (a little bigger).

OS
Attached Images
File Type: gif CXnew.gif (29.1 KB, 258 views)
File Type: gif CXnewFFT1k.gif (25.2 KB, 252 views)
File Type: gif CXnewFFT20k.gif (22.5 KB, 109 views)
__________________
Mongrel website , always current and updated :
http://www.fidelityforce.com/ostripper

Last edited by ostripper; 21st December 2010 at 02:28 AM.
  Reply With Quote
Old 21st December 2010, 03:23 AM   #1017
diyAudio Member
 
Lazy Cat's Avatar
 
Join Date: Sep 2007
Why bother? It has to be symetrical from input, double LTP and all that ...
  Reply With Quote
Old 21st December 2010, 03:31 AM   #1018
diyAudio Member
 
ostripper's Avatar
 
Join Date: May 2008
Location: Albany , NY (smallbany)
You don't understand , I stated why. single ended (first stage) cancels out even harmonics, 2nd stage (complimentary) cancels out odds. except for the vas , it is also mostly 14 cent transistors , just 4 to-126's , not much harder than a goldman/symasym. Why not ??, it's a hobby of mine .

OS
__________________
Mongrel website , always current and updated :
http://www.fidelityforce.com/ostripper
  Reply With Quote
Old 21st December 2010, 03:51 AM   #1019
AKSA is offline AKSA  Australia
diyAudio Member
 
Join Date: Sep 2001
Location: Melbourne, Australia
Pete,

You are building something with Halcro performance, with almost unmeasureable distortion. Great engineering achievement, to be sure, but will it sound good? I'm betting it will be dry, surgically accurate, uninvolving. A technical tour de force, a sonic let down......

But, have a shot, see if I'm wrong. I was wrong once before, took a corner too fast in 1963, fell off my bike.....

Hugh
__________________
Aspen Amplifiers P/L (Australia)
www.aksaonline.com
  Reply With Quote
Old 21st December 2010, 04:22 AM   #1020
49 - for the 18th time
diyAudio Member
 
c2cthomas's Avatar
 
Join Date: Oct 2006
Location: Near "Music City" (Nashville Tennessee)
I was wrong once before too – but I have a mental block that prevents me from remembering just when that was.

Wait a second – I’ll pop over and ask SWMBO………(imagine the sound of my footsteps)

Oh – never mind – it seems that I’m wrong about being wrong.
I would be more specific if I could – but she’s still laughing too hard to answer!!!!

I still find it remarkable how even the high and mighty can be rapidly put into place by SWMBO. Former President Bush recently remarked that Mrs. Bush has him drying the dishes.

Heh-heh -even OS - right in the middle of designing and building some great amps - needed to take a "time out" to clean up the mess he made in the Mrs. kitchen.
__________________
DIY audio can be expensive but getting to see things go up in smoke - that's priceless!!!! ..... "whatever - call it brainfart of Mighty ZM"

Last edited by c2cthomas; 21st December 2010 at 04:27 AM.
  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
Symasym - the next generation (supersym) ostripper Solid State 260 26th June 2014 09:23 PM
Supersym patent Rescue Toaster Pass Labs 6 22nd January 2008 09:49 PM


New To Site? Need Help?

All times are GMT. The time now is 12:21 PM.


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