TGM Amplifier ? - Page 33 - 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 5th June 2009, 11:33 PM   #321
Bigun is online now Bigun  Canada
diyAudio Member
 
Bigun's Avatar
 
Join Date: Jan 2009
Blog Entries: 2
Back to the Lender VAS,

with some help from OS it is morphing into something and the student (I) am liking it more as I learn about common base amplifiers.

Common base has a low impedance input which is fine because the LTP in TGM2 has plenty of drive for that. It also has excellent voltage gain - potential for lots of swing. This is an important attribute for the VAS.

I also don't see Cdom being such a problem in common base configuration, so good for HF and good isolation between input and output so not too much instability to worry about.

Trying this out on my pcb is a bit tricky. I'll have to make some changes, pull out a resistor (R4) and short it, add fly wire + resistor to connect up the other diff device collector to the VAS, maybe flip around the VAS device on the pcb since it's e-c-b can make it b-c-e. hmmmmm, will have to check if this is possible without making a mess.
__________________
"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 6th June 2009, 12:01 AM   #322
Banned
 
Join Date: Apr 2005
Location: Stockholm
Gareth,
now you are on the right track.
  Reply With Quote
Old 6th June 2009, 12:06 AM   #323
Bigun is online now Bigun  Canada
diyAudio Member
 
Bigun's Avatar
 
Join Date: Jan 2009
Blog Entries: 2
Quote:
Originally posted by Lumba Ogir
Gareth,
now you are on the right track.
Good, the student needs positive feedback (and guidance from Lumba, OS, MJ, Carlos etc.) to stay on track

and I will make the Master happy, to rid my amplifier of the Cdom
__________________
"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 6th June 2009, 12:48 AM   #324
AKSA is offline AKSA  Australia
diyAudio Member
 
Join Date: Sep 2001
Location: Melbourne, Australia
Hi Gareth,

You made some interesting observations here. Colin and I have done some work on this topology and can contribute a bit more here. (Colin is in Vancouver, BTW). You said this:

Quote:
Common base has a low impedance input which is fine because the LTP in TGM2 has plenty of drive for that. It also has excellent voltage gain - potential for lots of swing. This is an important attribute for the VAS.
As we know, common base holds base at a reference voltage, signal input to emitter, and signal output at collector. Input impedance is very low, as the injection point is effectively a virtual earth, and very similar to the folded cascode input situation.

However, output impedance is very high, as you'd expect from a reverse biased junction. It is high for all collector outputs except the CFP, so there is no advantage there.

You are right on the money with voltage gain, see here. The increased voltage gain will improve feedback factor for the complete amp, but because input and output phase are identical, there is less overall phase shift than a common emitter amp, and this increases the pole frequency (essentially it's a faster configuration, commonly used to drive RF amplifiers), making Nyquist criteria for stability a little easier to achieve - that is, smaller Cdom.

However, Cdom is difficult to implement because there is no phase reversal, so a simple cap from collector to base is not possible. Shunt compensation to ground is possible, certainly this works, as Colin and I have found, but it's not as effective and you need more of it. You can also use phase lead to the feedback node, which is problematic, but another LC trick is to string a cap from output to the antiphase side of the LTP, which shoud be resistively loaded to give a reverse phase signal.

If you use the Lender configuration, you can still use standard lag compensation, but enjoy the benefits of common base AND common emitter. Incidentally, in simulations I found that there is no gain advantage with the Lender over the standard single ended common emitter, since the emitter degeneration needed for the Lender scotches any increased gain you might otherwise derive from differential drive. In simulations the OLG appears to be identical either in diff, Lender drive or SE common emitter drive. However, the Lender does combine some of the characteristics of a good current mirror, as it very effectively forces current balance in the LTP.

Hope this is helpful,

Cheers,

Hugh
__________________
Aspen Amplifiers P/L (Australia)
www.aksaonline.com
  Reply With Quote
Old 6th June 2009, 01:40 AM   #325
Bigun is online now Bigun  Canada
diyAudio Member
 
Bigun's Avatar
 
Join Date: Jan 2009
Blog Entries: 2
Hugh,

Thanks for helping, it is always appreciated, and as always, you are way ahead; I had not stopped for a minute to consider how to achieve effective compensation once Cdom was slain.

TGM already incorporates some HF feedback from the VAS to the feedback node. I see what you mean by 'problematic' - both ends of that cap will now be at the same phase. I like your suggestion of compensation by finding the 180 degrees phase difference from the other side of the LTP.

With compensation I worry less about stability but more of how to ensure good sonics, no doubt because I haven't fried my amp with oscillations (yet). From your experience with these things, what is going to preserve / enhance the sonics that I am already enjoying ?

I will try something on a sim which I can run at work next week - die hard Mac fan at home, no Spice (plus I can see what OS does with the RCA 'back to the future amp' mods). I don't think I can bear to mess up my current pcb now that I'm listening to it.

You have me worried about the original Lender that I started with, I don't like the idea that it has 'current mirror' tendencies; some prejudices are useful in constraining the scope of this project
__________________
"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 6th June 2009, 01:58 AM   #326
AKSA is offline AKSA  Australia
diyAudio Member
 
Join Date: Sep 2001
Location: Melbourne, Australia
Gareth,

I cannot answer your question, but in typically confucian mode I can suggest questions...

Music is a subjective thang, and different topologies seem to suit different folks. Roender likes differential folded cascode, Scott Wurcer likes fully complementary input stages with diff VAS, I like single ended input stages (yes, even singleton's, which work very well!), but from a measured standpoint some are better than others.

However, from a sonics POV, the one which matters to me as I've never sold an amp to a customer with an AP distortion meter in his listening room, I favour single ended designs, and I like the simple common emitter amp, I like the Lender, and I like a buffered VAS. I've had some success commercially with very simple amplifiers using the old RCA Lin topology and honing the dimensioning and component choices.

So, while the sexy part of audio design involves imaginative topology choices, careful simulations and other highly mathematical investigations, I tend to think that the sonics lie, as they do in many technologies, with careful, dimensional fine tuning and synergistic component choices. I can ruin the sound of an amp with a bad choice of compensation cap, for example, as you've just found. Move it up or down by more than about 15% and you lose all the magic, as you put it. I'd even suggest that 70% of the design game lies with this one cap.

The unsexy aspects are component choice and dimensioning. Since we are working towards marginal improvements, it's precisely these areas which take lots of time, and huge iteration, which is mind numbing. Perhaps because of long years of military service during peace, I find this not too difficult.... however, my ears are now so bad I depend more and more on others with good ears, like Colin, Huy, and several others who hear absolutely everything.

The technical approach, laden with math and sharp observations about stability criteria, have their place, but rarely do academic designs sound marvellous. I'm pretty sure I know why, but it's taken years to confirm it, and the reasons are not popular in these parts.

Sorry, no easy answers...

Cheers,

Hugh
__________________
Aspen Amplifiers P/L (Australia)
www.aksaonline.com
  Reply With Quote
Old 6th June 2009, 02:54 AM   #327
andy_c is offline andy_c  United States
Banned
 
Join Date: Apr 2003
Quote:
Originally posted by AKSA
The technical approach, laden with math...
Math is only a burden for those who do not understand it. For those who do, it is an enabler.

Quote:
...and sharp observations about stability criteria, have their place, but rarely do academic designs sound marvellous. I'm pretty sure I know why, but it's taken years to confirm it, and the reasons are not popular in these parts.
I don't take issue with any claim along the lines that an incompetent hack, randomly changing components, can make a really good-sounding amplifier. I only take issue with your claim that only an incompetent hack who randomly changes components can make a great-sounding amplifier. This is not something that only occasionally comes up either. You repeat it so often it deserves "big lie theory" status.
  Reply With Quote
Old 6th June 2009, 04:49 AM   #328
SY is offline SY  United States
diyAudio Moderator
 
SY's Avatar
 
Join Date: Oct 2002
Location: Chicagoland
Blog Entries: 1
OT and personal posts removed. Keep to technical discussion.
__________________
You might be screaming "No, no, no" and all they hear is "Who wants cake?" Let me tell you something: They all do. They all want cake.- Wilford Brimley
  Reply With Quote
Old 6th June 2009, 09:07 AM   #329
AKSA is offline AKSA  Australia
diyAudio Member
 
Join Date: Sep 2001
Location: Melbourne, Australia
Quote:
I only take issue with your claim that only an incompetent hack who randomly changes components can make a great-sounding amplifier.
This is darkly false. I do not make that assumption. I say that a competent technician with reasonable equipment, a thorough knowledge of Ohms and Kirchoff's laws and a good ear can design an exceptional amplifier, quite the opposite of your contention. A strong knowledge of the history and theory of amp design is helpful too. Your laughable assumption shows sloppy thinking of a subjective, non-scientific kind - precisely what one would expect of someone with a very large axe to grind and arguably a vested self-interest in promoting the glory of mathematical ability.

Besides, where are the double blind tests which clearly show 100% correlation between low measured, single tone THD and subjective listening pleasure?

Have a nice day!
__________________
Aspen Amplifiers P/L (Australia)
www.aksaonline.com
  Reply With Quote
Old 6th June 2009, 02:47 PM   #330
Bigun is online now Bigun  Canada
diyAudio Member
 
Bigun's Avatar
 
Join Date: Jan 2009
Blog Entries: 2
Default Back on topic

Guys, you have my grey matter in over-drive trying to learn about common base.

I am thinking more about the common base. And I see that the simple implementation is now quite a problem for the LTP balance. My first set of simulations showed an increase in distortion and I think this is the cause.

And as I established a constraint of no current mirrors, so it leaves me with few options. I tried to find some ideas from the tube guys, who seem to like common-grid for phono stages [http://www.tubecad.com/2007/08/blog0116.htm]

One option is to go more symmetrical. I make two common base VAS devices, driven by each side of the LTP. This way the LTP is loaded the same on both sides. Sketch attached (you can see Lumba's schematic showing through from the other side of my sketch!).

Not sure yet if this common base has enough drive capability for the output. Self shows several VAS improvement schemes but points out that the non-linear input impedance of the output stage requires buffering.
Attached Images
File Type: jpg dscn1946f.jpg (62.5 KB, 757 views)
__________________
"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

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 07:55 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