A SOZ Transconductance Amplifier - diyAudio
Go Back   Home > Forums > Amplifiers > Pass Labs

Pass Labs This forum is dedicated to Pass Labs discussion.

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 28th June 2005, 05:01 PM   #1
gl is offline gl  United States
diyAudio Member
 
Join Date: Dec 2003
Location: Sierra Foothills - California
Default A SOZ Transconductance Amplifier

This is a SOZ transconductance amp I began building last Aug/Sept. and first turned on about Feb. 1. It's primary goal was to be a learning experience on current amps. For the most part this is an embodiment of all the "crumby hints" dropped by NP on this site as well as the good stuff included in the Zen articles. For my part there has been a great deal of cut-and-try experimentation and attendant headscratching. The amp was first turned on at beginning of February. At this point it plays music quite well although there is still a bit more distortion present than I think should be there.

There has been a relatively small and diminishing amount of interest in transconductaance amplifiers on this site over the last 8 months but that seems to have changed suddenly. For that reason I have decided to post this project such as it is.

This is not a copy of the F1. It was designed to use existing junk box parts. The heat sinks for instance are old Ampex TBC-1 sinks and the transformer is a Signal 88-8 from a long-gone A40. The sinks weren't up to F1 levels of heat dissipation so the CCS's were dropped from 1.75 amps to 1.4. This let me use .47 ohm 3W resistors, which I had on hand, to set the current. The 1.4 amps meant I could use single resistors in the places where NP had to parallel two parts in the F1. It all worked out quite nicely. I figure I'm getting about 6 to 8 watts - more than enough.

I currently use this piece with a pair of Fe166E's on quick and dirty MDF open baffles. By my calculations it produces approximately 15dB of gain with an 8 ohm load. I drive it from a single ended source and so lose 6dB. Even with 9dB of net gain into a high efficiency driver the gain does not seem as high as I would have expected.

If NP was not going to publish the F1 service manual shortly then I would keep on experimenting. I have a definite list of things to try. As it is I am going to stop work and wait for the F1 details to emerge. This has been a fun project. And although I don't consider it to good enough to rate stickers it has certainly met its goal of teaching me a lot about transconductance power amps.

The photo is fuzzy because the shiny metal confused the camera autofocus. It's a good thing because this partially obscures the electronic hairball on the perf board. The two visible transistors at the top of the right heatsink are the SFH9240 CCS's.

Thanks everyone for their encouragement - you know who you are - and NP in particular for the "crumbs" and the support. If there is sufficient interest I would be happy to clean up and post the schematic with more details about the circuit.

Best Regards,
Graeme
Attached Images
File Type: jpg g1.jpg (34.0 KB, 1210 views)
  Reply With Quote
Old 30th June 2005, 03:35 PM   #2
steenoe is offline steenoe  Denmark
diyAudio Member
 
steenoe's Avatar
 
Join Date: Mar 2004
Quote:
If there is sufficient interest I would be happy to clean up and post the schematic with more details about the circuit.
Graeme, I for one would like to see the schematic you arrived to If it isnt to much trouble to "clean" it up and post, that would be great These Current source amps are quite interesting.

Steen
  Reply With Quote
Old 30th June 2005, 03:49 PM   #3
diyAudio Member
 
promitheus's Avatar
 
Join Date: Jan 2001
Location: Germany
That thing looks gigantic !
Great stuff there.
I would also like to see the schematic when you have it ready.
  Reply With Quote
Old 30th June 2005, 11:27 PM   #4
gl is offline gl  United States
diyAudio Member
 
Join Date: Dec 2003
Location: Sierra Foothills - California
Hi Steen and Promitheus,

It is not gigantic really. The heatsinks are about 10 or 11 inches square. The caps are only 2 inches in diameter. Use the speaker binding posts as a gauge of the size.

OK, here is the schematic. The power supply schematic is not included. It is similar to those used in the Zen amps. The transformer is a Signal 88-8 which has two 22V-0-22V secondaries. Each channel has its own fullwave rectifier followed by a 30,000uF@50V/CL-60/30,000uf@50V filter section. The signal grounds for each channel are separate stars and each of these is connected to chassis ground through a CL-60.

The amp is essentially a SOZ using a pair of IRFP240's for gain. The drains of these transistors are connected to two CCS's based on Fairchild SFH9240's (IRFP9240's). These CCS circuits are configured to supply 1.4 amps each. This bias value was chosen because of the heatsinks I had available.

The lower CCS is based on Fig. 1 of the Zen V7 article. I used the 15k resistor/5K pot combo because I didn't have any 10K pots. Otherwise I would have used 10K/10K. This circuit works but I'm still not happy with it. I have tried a number of experiments here including using the center point of the 47 ohm resistors as the DC supply for the pot. The key here is to provide enough adjustment to bias the outputs, and the gain transistor source operating points, and also minimize bias drift. This part of the amp has been problematic and needs more work. The lower CCS circuit in the F1 appears to be very interesting and it is the one thing I want look at first when the service manual is posted.

The gain of the circuit was set by using the formula on page 3 of the "Balanced Zen Line Stage" article. I decided to follow the example of the F1 and shoot for approximately 14dB of gain or a gain of 5. The F1 manual implies that the 14dB of gain is calculated with an 8 ohm load in place. The result of this indicated that resistance needed to be added to the sources of the gain transistors. The .47 ohm resistors aren't quite the right value but as you can see that's what I had on hand. Actually I might have overestimated the transconductance here since I only have 1.4 amps of bias.

The input circuit is based on recent hints by NP. He uses 1 uF coupling caps on the F1 and has described it's DC bias circuit in general terms. The 330K values seem to work OK although this too is an area for further tinkering. I was also planning on trying 470K values here as well. I experimented with connecting the 330K resistors to the center point of the two 47 ohm resistors.

To set the operating point of the amp the pot P1 is turned up until the two upper CCS's are fully on (.66 volts across the .47 ohm resistors), At this point the pot is turned up further until 1) the DC value on the outputs is roughly 55% of the V+ supply and 2) there is about 5V on the output of the lower CCS. All the other resistor values are chosen so that the one pot can set all of these operating points.

As I stated in my earlier post there is still some audible distortion so either I'm not doing something right or I've toasted one or more of the components. So far I've only burned up a couple of 1/4 watt resistors in spite of all my experimenting. If anyone sees anything strange in any of this please point it out.

Best Regards,
Graeme
Attached Files
File Type: pdf g1-00.pdf (71.7 KB, 369 views)
  Reply With Quote
Old 1st July 2005, 03:56 AM   #5
The one and only
 
Nelson Pass's Avatar
 
Join Date: Mar 2001
Quote:
Originally posted by gl
As I stated in my earlier post there is still some audible distortion so either I'm not doing something right or I've toasted one or more of the components. So far I've only burned up a couple of 1/4 watt resistors in spite of all my experimenting. If anyone sees anything strange in any of this please point it out.
There is a subtle point about bias stability between the Drains
of two set of Common Source devices. When the schematic is
released shortly, you'll find it easy to make a correction, if
you can wait that long....
  Reply With Quote
Old 2nd July 2005, 10:36 PM   #6
gl is offline gl  United States
diyAudio Member
 
Join Date: Dec 2003
Location: Sierra Foothills - California
Aha. Now that is a seriously major crumb. How can I wait with that sitting there?

I'm figuring that the two 47 ohm resistors must need to become drain loads at audio frequencies. That means that a large cap needs to dropped from the junction of these to ground. That must be the large cap I see in the F1 internal photo now on the First Watt site. These are busy resistors - they do many things.

I understand why the 47 ohm drain load resistors don't break the transconductance property of the amp connected to gound as opposed to V+ or back to the gates. But I am going to have to think about why they're necessary for AC stability. Sort of reminds me of all the discussion on VAS load resistors. I presume that the 15 ohm resistor on the F2 output performs a similar stabilizing function when that amp is being operated with no load.

Thank you again sir.

Regards,
GL
  Reply With Quote
Old 2nd July 2005, 11:41 PM   #7
steenoe is offline steenoe  Denmark
diyAudio Member
 
steenoe's Avatar
 
Join Date: Mar 2004
Quote:
How can I wait with that sitting there?
Neither can we I mean all the other guys....... I do believe though, that the F amps are simple enough! I mean, take a seriuos listening to a ZenV1.... It just sounds amazingly good to my ears

Steen
  Reply With Quote
Old 3rd July 2005, 03:35 AM   #8
The one and only
 
Nelson Pass's Avatar
 
Join Date: Mar 2001
Quote:
Originally posted by gl
I'm figuring that the two 47 ohm resistors must need to become drain loads at audio frequencies. That means that a large cap needs to dropped from the junction of these to ground. That must be the large cap I see in the F1 internal photo now on the First Watt site. These are busy resistors - they do many things.

I understand why the 47 ohm drain load resistors don't break the transconductance property of the amp connected to gound as opposed to V+ or back to the gates. But I am going to have to think about why they're necessary for AC stability. Sort of reminds me of all the discussion on VAS load resistors. I presume that the 15 ohm resistor on the F2 output performs a similar stabilizing function when that amp is being operated with no load.
Very astute about the capacitor. Now consider that it really is
the DC voltage on the Drains that requires control. Two current
sources butting heads is not generally very stable, and there
is no common mode load on the balanced output stage.

The F2 gets it's DC stability from a different mechanism, and it
looks more like the original Zen, but the DC bias networks are
very high impedance.

  Reply With Quote
Old 3rd July 2005, 05:20 AM   #9
gl is offline gl  United States
diyAudio Member
 
Join Date: Dec 2003
Location: Sierra Foothills - California
Hmmmmmm.

The only thing I can think of is that a feedback mechanism is required to keep the DC on the drains stable. Like taking the DC supply for the lower CCS gate from the junction of the 47 ohm resistors like in the Zen V7. As the DC drain voltages float upward the lower CCS gate voltage would be pulled up and that would counteract what the output points were doing. Vice versa for drift in the other direction. I suspect that the network on the lower CCS gate would need to be as low an impedance as practically obtainable.

The "butting heads" situation with the CCS's has been a concern from day 1 for me and I may have even said so in a post. In the process of wrapping my head around the concept of a transconductance "power" amp I didn't think much about anything common mode. The simplicity of the circuit topology puts you off guard. Much food for thought here.

Thank you.

Regards,
GL
  Reply With Quote
Old 3rd July 2005, 08:15 PM   #10
The one and only
 
Nelson Pass's Avatar
 
Join Date: Mar 2001
Quote:
Originally posted by gl
I suspect that the network on the lower CCS gate would need to be as low an impedance as practically obtainable.
You forgot the cap.
  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
transconductance power amplifier for current-drive loudspeaker? alex0 Solid State 30 26th May 2008 08:15 AM
LM3886 as transconductance amplifier? squadra Chip Amps 4 23rd April 2008 10:16 PM
A Transconductance Amplifier Duo Solid State 27 19th February 2007 07:59 PM
[transconductance] error amplifier section of UC2526 zilog Power Supplies 0 21st January 2007 11:40 PM
commercial high-power transconductance amplifier! taloyd Subwoofers 10 5th August 2005 04:02 AM


New To Site? Need Help?

All times are GMT. The time now is 12:41 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