bootstrapsCCS+T-TMC - Page 3 - 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 14th April 2011, 12:02 PM   #21
dadod is offline dadod  Croatia
diyAudio Member
 
dadod's Avatar
 
Join Date: Apr 2006
Location: Zagreb
Quote:
Originally Posted by jcx View Post
I must have been too indirect - that is nearly the opposite conclusion from how I interpret the research, theory, practice

there are several distortion mechanisms that rise with frequency - in fact they cause rising phase shift nonlinearity with frequency too - independent of loop gain slope - so for flat loop gain amps the distortion does rise with frequency
with higher loop gain the distortion rises to the same level that it gets to with the same amp "flat loop gain compensated" - but the distortion at all frequencies below the corner is lower

I would point to Self, Cordell and even our low/"no" feedback designers like Pass and Curl here to suggest that the output stage looks pretty feeble - even at 100 W parallel output devices are a good idea for multidriver dynamic loudspeaker with possible 2 Ohm impedance dips due to bad XO design

and triple EF seems like a requirement to avoid load impedance and Q beta nonlinearity from limiting VAS performance

when those measures are in place I think the "back emf"/complex load arguments are way less relevant
I have understood your argument about cutting loop gain at audio frequencies, and I followed all that discussion before, but I could not help myself but to try it and hear how it sounds.
I do not think there is any benefit that I can hear, but some higher distortion at lower frequences is no big deal. I was trying to get decreasing harmonics in the FFT(simulated only).
Triple EF, parallel output devices, that is all an another story. I wanted simple and good sounding amp.

Quote:
I would consider guarding/shielding the mirror out/VAS buffer end node just to maximize loop gain
Did you meant triple EF by this??
dado
  Reply With Quote
Old 14th April 2011, 12:18 PM   #22
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
I consider critical subjective evaluation too difficult – you absolutely need two complete amps with only the one tested variable different between them, level matching, A/B/X and a number of subjects improve the reliability – without that level of care I would dismiss even my own “just listen” opinions as worthless for informing engineering design


guarding is a layout technique that put a low impedance trace between a sensitive node and a "aggressor" node that has signal we want to keep out ot the sensitive node

the low impedance trace preferably follows the signal on the sensitive node to prevent its (larger) parasitic leakage impedances to the sensitive node from being a problem

given the high sensitivity of the 2-pole response to small pF C, MOhm "bridge" impedances I would consider guarding/shielding the mirror-VAS input buffer node from the VAS out/amp output node's large Vswing

as for Lateral Mosfets - what they are missing is Gain - gm
for reduced VAS load they only look good at low frequencies - at high audio frequency the incompletely bootstrapped Cgs becomes a significant load

Last edited by jcx; 14th April 2011 at 12:32 PM.
  Reply With Quote
Old 14th April 2011, 12:38 PM   #23
MiiB is offline MiiB  Denmark
diyAudio Member
 
Join Date: May 2005
Location: Denmark
having worked with speakers i know that our hearing is particularly sensitive to the phase content...so when it comes to base performance of tubes h2 may be the reason, but another could be the low end phase shift that starts an octave or two above the actual low end cutoff frequency...this being trafos or coupling caps.. I agree that good tube amps excel in base performance over most SS equipment
  Reply With Quote
Old 14th April 2011, 12:51 PM   #24
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
if the effect is real and its just frequency response of coupling networks then SS amps can be EQed - likewise output impedance can be padded to match

in fact that is what Bob Carver did in the Carver-Stereophile Challenge

with their own source, speakers, in their own listening room Stereophile's "Golden Eared" reviewers couldn't tell Bob's $600 SS amp from their own choice of "SOTA" tube amp after Bob tweaked the SS amp frequency response, output impedance for a deep null with the tube amp
  Reply With Quote
Old 14th April 2011, 01:02 PM   #25
dadod is offline dadod  Croatia
diyAudio Member
 
dadod's Avatar
 
Join Date: Apr 2006
Location: Zagreb
Quote:
Originally Posted by jcx View Post
I consider critical subjective evaluation too difficult – you absolutely need two complete amps with only the one tested variable different between them, level matching, A/B/X and a number of subjects improve the reliability – without that level of care I would dismiss even my own “just listen” opinions as worthless for informing engineering design


guarding is a layout technique that put a low impedance trace between a sensitive node and a "aggressor" node that has signal we want to keep out ot the sensitive node

the low impedance trace preferably follows the signal on the sensitive node to prevent its (larger) parasitic leakage impedances to the sensitive node from being a problem

given the high sensitivity of the 2-pole response to small pF C, MOhm "bridge" impedances I would consider guarding/shielding the mirror-VAS input buffer node from the VAS out/amp output node's large Vswing

as for Lateral Mosfets - what they are missing is Gain - gm
for reduced VAS load they only look good at low frequencies - at high audio frequency the incompletely bootstrapped Cgs becomes a significant load
I agree completely with you about subjective evaluation, and now I am in no possibility to do proper listening test.
I included PCB in this tread in the begining and I would appreciate your comments.
Quote:
given the high sensitivity of the 2-pole response to small pF C, MOhm "bridge" impedances I would consider guarding/shielding the mirror-VAS input buffer node from the VAS out/amp output node's large Vswing
I am not sure I understand this, do you mean it's better to use TPC instead of TMC??
One other thing, I downloaded SPICE models from Supertex, but LTspice doesn't recognize it. I used NJFET from LTspice librery for temporary test.
Here is my LTspice zip file.
dado
Attached Files
File Type: zip DX-TMC.zip (12.3 KB, 78 views)
  Reply With Quote
Old 14th April 2011, 01:11 PM   #26
MiiB is offline MiiB  Denmark
diyAudio Member
 
Join Date: May 2005
Location: Denmark
If you try to make speakers linear in the crossover region that area will stand out...so you need a suck-out in that area to have the speaker sound natural...my understanding of this is that our hearing is quite sensitive for multi source multiphase SPL...also testing raw drivers reveals this, as you can't actually tell the SPL curve by listening, as what you feel is loud actually measure quite low and the and the opposite what you hear as gentile can in fact have high SPL...much more to speakers than SPL linearity and distortion..
  Reply With Quote
Old 14th April 2011, 01:13 PM   #27
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Quote:
Originally Posted by jcx View Post
if the effect is real and its just frequency response of coupling networks then SS amps can be EQed - likewise output impedance can be padded to match

in fact that is what Bob Carver did in the Carver-Stereophile Challenge

with their own source, speakers, in their own listening room Stereophile's "Golden Eared" reviewers couldn't tell Bob's $600 SS amp from their own choice of "SOTA" tube amp after Bob tweaked the SS amp frequency response, output impedance for a deep null with the tube amp
This reminds me of Baxandall's test set up for comparing amplifiers.
__________________
regards Andrew T.
  Reply With Quote
Old 14th April 2011, 01:18 PM   #28
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
I prefer 2-pole compensation on theory/simulation grounds – I’ve not built comparable hardware - composite op amp circuits are more my current style - analyzing, simulating discrete power amps is more an intellectual exercise for me

I believe it was finally accepted in the Cordell Book thread TMC debate that my linear stability analysis sims showed no advantage to TMC vs 2-pole – and that the loop gain cross section for useful stability analysis has to measure the gain around the output Q inside the local TMC loop Bob Cordell's Power amplifier book

There also appears to be little to choose between them in simulated clipping response when values are adjusted for the same total loop gain enclosing the output Q

Since there is a fairly large linearizing effect on the input diff pair with 2-pole’s larger loop gain giving much smaller input error signal I would choose 2-pole over TMC with what I think I know about the two options

Last edited by jcx; 14th April 2011 at 01:30 PM.
  Reply With Quote
Old 14th April 2011, 01:31 PM   #29
Dave S is offline Dave S  United Kingdom
diyAudio Member
 
Join Date: Mar 2003
Location: Searchin' for Sweet Singletrack
Quote:
Originally Posted by AKSA View Post
David, this is semantics, superficial nonsense. You know as well as I do that damping factor is the ratio of load to source impedance.
Maybe my post was not clear - I was actually agreeing with your POV

Quote:
Originally Posted by AKSA View Post
Again, you know well that SETs inject a great deal of H2 and that this is indeed enjoyed by many audiophiles. Adding some output resistance simply hikes up output impedance, lowering damping factor and achieving precisely what I was alluding to.
I was trying to make the point that high damping factor might result in a badly damped system i.e. with Q >0.7.

Quote:
Originally Posted by AKSA View Post
You mustn't kick yourself for your apparent 'poor judgement' buying the AKSA, which you obtained by crafty negotation anyway. I would hope you sold it for a good price! This is a cheap shot, and completely unsolicited.
I don't remember any crafty negotiation, although my receipt from April 2002 does have the kit price of AU$420 crossed out and $213 written there instead. Sorry if you feel bitter about this. I don't have any feelings of guilt about the deal, but maybe I just have a poor memory.

I really enjoyed owning the AKSA and it met its promise of sounding "valvelike". I still own a P61 and although it's not as good as the 100N it shares the same characteristics of sounding warm and engaging. IMO where both these amps lose out is when trying to follow busy music and complex rhythms.
Please let me stress that this is only my opinion and readers are at liberty to assume I'm wrong/foolish/lying or whatever else they choose.
By the way I never got attacked by NP for suggesting my Aleph 30 is very boring to listen to. This really is an amp that is hopeless with rhythms IMHO.

Quote:
Originally Posted by AKSA View Post
More gratuitous criticism, really, why don't you use the forum for its intended purpose, to swap information of genuine use to other people?

Hugh
Sorry if you thought my "ported" comment was aimed specifically at your speakers. FWIW I also use ported speakers and I know that theoretically they have poor transient response at LF. I am working on a new project to sidestep this.

There seems to be an unwritten rule on this forum that nobody shall criticise the sound of any equipment. I actually value people's subjective comments (although it is nigh on impossible to judge if these people share your own perspectives) and regard reasonable criticism as of "genuine use".
  Reply With Quote
Old 14th April 2011, 02:40 PM   #30
dadod is offline dadod  Croatia
diyAudio Member
 
dadod's Avatar
 
Join Date: Apr 2006
Location: Zagreb
Default TPC now

I simulated the same schematic, but with 2-pole compensation.
Schematic, FFT, OLG with R//C and OLG without R//C.
distortion at 1kHz
Fourier components of V(vout)
DC component:0.00057103

Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 1.000e+03 2.134e+01 1.000e+00 -0.02° 0.00°
2 2.000e+03 1.189e-05 5.573e-07 100.11° 100.12°
3 3.000e+03 1.109e-05 5.195e-07 16.99° 17.01°
4 4.000e+03 6.457e-07 3.025e-08 95.53° 95.55°
5 5.000e+03 1.175e-07 5.508e-09 -71.57° -71.55°
6 6.000e+03 4.228e-07 1.981e-08 90.47° 90.49°
7 7.000e+03 8.708e-07 4.081e-08 22.50° 22.52°
8 8.000e+03 2.746e-07 1.287e-08 64.52° 64.54°
9 9.000e+03 9.066e-07 4.248e-08 27.06° 27.08°
Total Harmonic Distortion: 0.000077%

On next PCB I will make provision for both TMC and TPC and then I can compare!!
dado
Attached Images
File Type: jpg DADOt-TPC-T_NJFccs.jpg (245.8 KB, 437 views)
File Type: jpg DADO-TPC-T_NJFccs-FFT.jpg (109.1 KB, 429 views)
File Type: jpg DADO-TPC-T_NJFccs-OLG.jpg (122.6 KB, 398 views)
File Type: jpg DADO-TPC-T_NJFccs-OLGnobr.jpg (120.2 KB, 387 views)
  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
ThermalTrak+TMC amp dadod Solid State 244 8th July 2014 09:05 AM
MEMDIST POWER AMP + JC + TMC - help to set up padamiecki Solid State 8 10th January 2014 09:38 AM
Real World TMC/ETMC applications ostripper Solid State 23 14th December 2010 07:57 PM
What to do with TMC tube radio gear? diamondsouled Tubes / Valves 4 31st March 2009 01:30 PM


New To Site? Need Help?

All times are GMT. The time now is 04:00 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