TGM2 amplifier - Page 2 - 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 11th June 2009, 11:17 PM   #11
Account disabled at member's request
 
MJL21193's Avatar
 
Join Date: Mar 2007
Quote:
Originally posted by AKSA

I'd guess the reason is that you have a phase reversal before the OLG has come back to unity - that is, you have not met the Bode-Nyquist criteria.

Yes, if I may demonstrate on my Havoc amp (has the CFP input). Attached is the correctly degenerated loop gain plot. Emitter R's are 91 ohms each:
Attached Images
File Type: png 117.png (30.4 KB, 1106 views)
  Reply With Quote
Old 11th June 2009, 11:18 PM   #12
Account disabled at member's request
 
MJL21193's Avatar
 
Join Date: Mar 2007
And here's how it looks with no degeneration:
Attached Images
File Type: png 118.png (30.6 KB, 1078 views)
  Reply With Quote
Old 12th June 2009, 12:41 AM   #13
Bigun is offline Bigun  Canada
diyAudio Member
 
Bigun's Avatar
 
Join Date: Jan 2009
Blog Entries: 2
Hugh, MJL21193,

You explained it so well that it now seems obvious in hindsight

That 180 degrees kicks in where you don't want it !

So in fact, even if the sims. show no oscillations when using a resistive feed to the LTP, it's likely pretty marginal and guaranteed to ruin the party if I build it like that.

So I got a nice improvement with the CFP front end, better linearity (lower distortion), higher slew rate if I want to crank the current up.

Next stop will be to investigate:

Is there really any more to be gained out of this CFP front-end by adding a CCS ?
__________________
"The test of the machine is the satisfaction it gives you. There isn't any other test. If the machine produces tranquility it's right. If it disturbs you it's wrong until either the machine or your mind is changed." Robert M Pirsig.
  Reply With Quote
Old 12th June 2009, 12:52 AM   #14
diyAudio Member
 
ostripper's Avatar
 
Join Date: May 2008
Location: Albany , NY (smallbany)
Quote:
With emitter degenration of the LTP (33R - 47R wasn't enough, 82R was marginal, 100R OK)
Confirmed , I did the Step thing with a CFP LTP @47p Cdom

.STEP PARAM RDG 5 165 40 (starts at 5r , increases by 40 up to 165R)
Replace your LTP resistor values with {RDG}

As the plot shows , at both 5R and 45R , you have too high a UG
freq. (3.1mhz and 1mhz) with insufficient phase margin (almost 0 -180 degrees) with a 5R resistor. As you said 85R ( 600k UG -92 phase) is stable. with 68 Pf (47R - 68R is stable).
OS
Attached Images
File Type: gif phase.gif (25.6 KB, 1039 views)
  Reply With Quote
Old 12th June 2009, 01:14 AM   #15
Bigun is offline Bigun  Canada
diyAudio Member
 
Bigun's Avatar
 
Join Date: Jan 2009
Blog Entries: 2
Hi OS,

Thanks for checking into it, I can see I'm in good company.

I was surprised how much degeneration I needed when I was manually changing the resistor in the sim (I'll have to try this .step command).

OS, Hugh,

Thing is, I am reluctant to increase Cdom because my listening tests on the TGM1 resulted in me reducing Cdom to 44pF. Of course there's no guarantee that this is the best 'sound' for TGM2. This is an example of where the 'subjective' influences the design.

Another parameter I can look at is the gm of the input devices. Increasing the collector resistance on these would reduce the current and gm a little and give me more flexibility on Cdom/degeneration tradeoff maybe. I can arrange it to keep the total LTP current unchanged so that the reduction in current through the input devices is added to that through their 'slaves'. Does this make sense ?
__________________
"The test of the machine is the satisfaction it gives you. There isn't any other test. If the machine produces tranquility it's right. If it disturbs you it's wrong until either the machine or your mind is changed." Robert M Pirsig.
  Reply With Quote
Old 12th June 2009, 01:45 AM   #16
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
Attached is something that I think was suggested in your other thread, but that wasn't said again.

Still not a perfect CCS, but I would say it is a trade off between what you want and what others suggest.

Also, pertaining to turnoff thump - I am very surprised no one has said this yet.

Look at C1 and C2, and the accompanying resistors. The time constants are not the same. This will cause some "thump" if I'm not mistaken.

Look at this schematic. The gain is 10, so the voltage across the feedback network will be 10 times the input voltage. Also, the feedback network has 10 times less resistance (I do this because it increases stability and also decreases distortion by a very insignificant amount - however, there will be offset voltage at the output even on a fully balanced LTP). So multiply the input capacitance by 100 and you have the same time constants. With this configuration, the time constants are the same (unless someone corrects me). Before we try to destroy turn-on thump some other way, I think we should correct this.

Click the image to open in full size.

Hope this is useful,
- keantoken
Attached Images
File Type: png bootstrapccs.png (6.9 KB, 886 views)
  Reply With Quote
Old 12th June 2009, 02:33 AM   #17
diyAudio Member
 
ostripper's Avatar
 
Join Date: May 2008
Location: Albany , NY (smallbany)
Quote:
guarantee that this is the best 'sound' for TGM2.
With CFP , the best overall distortion was realized at "balance"
(equal current between slave / master) , play with that B-E resistor.
or step it between 470r - 1.5k to see.

Use this for your main directive :

Quote:
.param num_fft_pts 1024
.param freq 20k
.param timestep {1/((num_fft_pts-1)*freq)}
.tran 0 1m 0 {timestep}
.options plotwinsize=0 .four {freq} V(c)
Better AC source ... (0 {sqrt (.3)} {freq}) add this as your signal source value.

Make the output of your amp (C) "label net"

Add the .step param Rbe 400 1500 100

make your 2 slave resistors {RBE} ,RUN ... then the spice error log will show distortion for each step (all 11 of them). If you do a FFT of the output (C) , you will also see the harmonic "shift" as you step through the values.
REAL COOL
OS
  Reply With Quote
Old 12th June 2009, 02:39 AM   #18
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
If you use the two-transistor CCS, the CFP resistor will be roughly 4x the value of the CCS resistor for the mentioned equal current.

When configuring LTPs I look at all the resistor values relative to the CCS resistor. It makes things simpler.

- keantoken
  Reply With Quote
Old 12th June 2009, 03:38 AM   #19
Bigun is offline Bigun  Canada
diyAudio Member
 
Bigun's Avatar
 
Join Date: Jan 2009
Blog Entries: 2
Thanks guys, this will keep me busy at the Sims when I next get on it

I knew there was more to this CFP than I saw first time around. I need to upgrade my Spice skills and look at this design more closely. I did briefly look at a balanced master-slave current in the LTP but didn't see an obvious null in distortion. I'll try some of these Spice directives.


Talking of turn-on thump
------------------------------

First off, I don't have a noticeable thump with TGM1, but that doesn't mean there isn't any.

Keantoken - if I understand your point,

At turn-on the VAS and output are woken up as quickly as the psu can charge its caps. To avoid trouble we want the voltages at the bases of the LTP input devices to come up together so that the turn-on transients appear as common-mode input, which is rejected by the LTP before it reaches the VAS. That makes sense to me !

So I look at how quickly the voltages on the bases of the input devices can change.

The input device (Q1) sees the input cap (C1) and the input shunt resistor R1 (+R2, which we can ignore). RC = 0.47uF x 44k = 20.68ms (giving me an input filter with a -3dB at 48Hz).

The feedback device (Q2) sees the input signal ground via the feedback loop of R8 and C2. RC = 22uF x 2.7k = 103ms (suggesting that the feedback loop is not limited my bass response).

They don't match.

I did look at turn-on thump in my sim. I noticed that there was a transient at the output, will have to check but I think it was about 40mV at worse.

What you have said makes perfect sense, and I can modify the design to balance the RCs more closely.


p.s. Bootstrapping wasn't forgotten, but my current aim is to understand the CFP structure a bit more, and see how much improvement a CCS offers for a CFP-LTP.
__________________
"The test of the machine is the satisfaction it gives you. There isn't any other test. If the machine produces tranquility it's right. If it disturbs you it's wrong until either the machine or your mind is changed." Robert M Pirsig.
  Reply With Quote
Old 12th June 2009, 03:48 AM   #20
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
Yes, you understand perfectly!

Big, SPICE can be an amazing tool for learning electronics. If you want to know something, hook it up in spice - don't get lazy!

I learned practically everything I know about electronics from spice+DIYAudio, in I would say max 3 years.

So as far as learning about the LTP - I did it by simming numerous "Blameless" style class A amplifiers with various tweaks I had come up with. As you screw around you tend to figure out things that would have never occurred to you without all that time tinkering. Time has a strong correlation with experience (although, you must never assume this to be true in all circumstances).

Have fun,
- keantoken
  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
Wanted: DIY amplifier builder with experience. I need a 0.7 ohm stable amplifier Audiophilenoob Swap Meet 34 16th May 2005 11:58 PM
Questions about how to decrease the PASS A-40 amplifier into a class AB amplifier. novtango Pass Labs 2 21st October 2002 01:50 AM


New To Site? Need Help?

All times are GMT. The time now is 11:52 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