The Amplifier of 100 Transistors - Blameless using small signal devices as outputs! - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State
Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

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
Thread Tools Search this Thread
Old 7th June 2016, 10:53 AM   #1
diyAudio Member
Join Date: Nov 2006
Location: Adelaide, South Australia
Blog Entries: 45
Default The Amplifier of 100 Transistors - Blameless using small signal devices as outputs!

The Amplifier of 100 Transistors is exactly that. Or as my original blog was titled, "Now for something completely different." Refer here:

The idea of an amplifier comprised completely of small signal devices struck me, and the thought stuck - anything so silly just had to be tried. Unfortunately the idea did not leave my head until I had designed the layout, and then it was just too late and it had to be built!

Not only did the implementation work, but on testing it actually worked quite well: --> not world beating, but creditable.

I have had a few comments on the BLOG for this, and I figured it might be of passing interest to "Solid State Amplifier people". This is solid state in the purest way! What other amplifier has so many output devices - and yet outputs only 10 Watts? That said however, it does so with rather unique style and in a way that I have not seen before. (Nor, may you think, should we ever see again!)

For those of you looking for the "car chase":
- The amplifier comprises ONLY BC548/558 (and in the input 549/559
- There are 54 pairs of output devices
- The amplifier runs off +/-15 Volts (VCEo for BC558 is 30 volts)
- The amplifier is a very standard blameless design
- The amplifier includes 2 slope SOAR protection, so is reasonably safe into pretty much any load
- The amplifier can happily drive 4 Ohm loads without any protection kicking in

Distortion is surprisingly low, even though I have not gone to any super effort to tune anything.

Freq No load 8R
200Hz 0.0003% 0.0003% (noise floor)
500Hz 0.0003% 0.0003% (noise floor)
1000Hz 0.0003% 0.0006%
2000Hz 0.0006% 0.001%
5000Hz 0.0015% 0.0025%
10000Hz 0.0025% 0.0044%

So what is it? Lots and lots of very low cost transistors, really. Oh, and NO heat-sink! NOTE: This schematic includes two pole compensation. The one I built has only the 120pF cap - this is the only difference between these schematics and what I built.

******* Watch out the 4K7 on the emitters of the LTP is rubbish and should be deleted or made quite a low value.


And the rest of the "output stage"


Which once laid out on the PCB out looks a bit like this:


Note 1: I was making this PCB in my shed, and considering that the board is 245mm long by 70mm wide, I did not want to be doing a full double sided PCB. For this reason I went to some effort to get the bulk of traces onto the back of the board. If I were to get some made professionally, I would double up on the power rail and output traces. As it turns out, going single sided (with a few links) did not have a serious impact on performance.

Note 2: I went to a fair bit of trouble to minimise the loop inductance between the supply rails and combined outputs coming back - this was due to my paranoia over parasitic oscillations in the output stage. For this reason I have put the power rails down between 1206 size 10 Ohm emitter resistors. This means that the "power" rails are kind of high resitance, but the maximum current on each of these will be less than an amp, and on 8 Ohm loads only 500mA peak. The compromise on track width for the supply rails was on I was willing to make.

And once built (look for the wire links that are the "top layer" of the PCB):


Note: that dummy load led me on a wild goose chase on distortion measurements!

As noted, distortion was a lot better than I had expected. Indeed, in my initial blog, I noted concern that Douglas Self would kick me somewhere delicate for defiling the blameless topology with this implementation!

Distortion measurements, plots are (1) loopback on the distortion test set (2) no load distortion of the "Amplifier of 100 Transistors" and (3) distortion into 8 Ohms.


Final -3dbNoLoad_distortion.jpg

Final -3db8r_25W_WW.jpg

There is a reasonable amount of discussion in my blog about some of the reasoning as to what I did and why, feel free to read that.

I am more than happy to post the ALTIUM files for those wishing to give this a spin. Comments though:
- I can't guarantee you won't go mad soldering on the 104 emitter resistors (SMD), the 104 base stopper resistors (SMD), 104 output devices (TO-92)
- I had no trouble with stability. Was I lucky? You will need to try and make your own call on this.
- I used BC548/558's from a single batch. (A bit like a single vineyard Shraz?) I would recommend you try not to mix and match output devices too much in a single amp.

Your challenge:
- Tweak that two pole compensation
- Try different "Output Devices" - the USA seems to favour 2N3904/6 I have not tried these, and you might want to take a look at the HFE before you jump in too far!
- If you find higher voltage devices, push the limits... Remember to tune the SOAR protection though!
- Simply build one "because you can". 10 Watts is actually enough to be useful.

Anyway, I hope this inspires someone to try something completely different for a change... have fun!

Last edited by googlyone; 9th June 2016 at 09:54 AM.
  Reply With Quote
Old 7th June 2016, 11:05 AM   #2
diyAudio Member
Join Date: Sep 2009
Elektor did something similar with a bunch of 5532 OPAMPs which might interest you

5532 Opamplifier Deel 1 ontwerp, filosofie en schemas

in Dutch but schematics are understandable in any language
  Reply With Quote
Old 7th June 2016, 11:41 AM   #3
Bigun is offline Bigun  Canada
diyAudio Member
Bigun's Avatar
Join Date: Jan 2009
Location: Waterloo, ON or Herefordshire
Blog Entries: 2
Great job - I'm glad somebody finally did this as I've had a similar idea for several years but never actually made it. Mind you, I was thinking a heatsink and Class AB would be handy to allow higher power level. I hope you are going to put this in a box and press it into service.
"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.

Last edited by Bigun; 7th June 2016 at 11:44 AM.
  Reply With Quote
Old 7th June 2016, 12:03 PM   #4
diyAudio Member
Join Date: Nov 2006
Location: Adelaide, South Australia
Blog Entries: 45
I am not sure how I would heatsink this other than use free air... I guess I could have used TO-126, but then that would have been cheating on the original design brief.

I do need to put it in some sort of a box. Given the TO-92's are essentially the heatsink themselves, and in a real sense are the "point" of this thing, I think they will need to be either external or rather exposed.

I have not actually listened to these yet - the measurements do suggest that it will be rather blameless in sound though!
  Reply With Quote
Old 7th June 2016, 01:11 PM   #5
diyAudio Member
vzaichenko's Avatar
Join Date: Jan 2010
Location: Moscow, Russia
Send a message via Skype™ to vzaichenko
Cool! I like non-standard approaches
If you can't explain it simply, you don't understand it well enough (c) Albert Einstein
  Reply With Quote
Old 7th June 2016, 01:51 PM   #6
diyAudio Member
Join Date: Apr 2010
Location: Coffs Harbour
Originally Posted by gideon1990 View Post
Elektor did something similar with a bunch of 5532 OPAMPs......
No prizes for guessing the designer of that amplifier either....Mr Blameless, himSelf

On the topic of multiple discrete transistor amplifiers - I think Nelson Pass has taken the prize for crazy numbers of T092 parts with this one using more than 2000 JFETs! Now that's a monument to a vanishing device type:
  Reply With Quote
Old 7th June 2016, 02:27 PM   #7
xrk971 is online now xrk971  United States
diyAudio Member
xrk971's Avatar
Join Date: Oct 2012
Location: Metro DC area
Pass did a class A with 1000 TO92 JFETs.

Edit : above post already caught this.

I like this because 100 is much easier to solder and the parts are not as exotic as the ones Pass uses.

Last edited by xrk971; 7th June 2016 at 02:30 PM.
  Reply With Quote
Old 7th June 2016, 02:44 PM   #8
diyAudio Member
Join Date: Jan 2010
nice! that's real audio madness
  Reply With Quote
Old 7th June 2016, 03:02 PM   #9
diyAudio Member
Join Date: Mar 2007
Location: California
Originally Posted by gideon1990 View Post
Elektor did something similar with a bunch of 5532 OPAMPs which might interest you

5532 Opamplifier Deel 1 ontwerp, filosofie en schemas

in Dutch but schematics are understandable in any language
If this is Douglas Self's Op-Amplifier (two-part article in Elektor) then this is exactly what I was thinking of when I saw this thing. Self's project was pretty darn good and used op-amps (64?). This inspired me to do the same in my own 5532 headphone amplifier project and I measured very low THD, etc. with many devices in parallel into a 32R load (min headphone load).
Visit my Audio Web Page <<--CLICK TO LEARN MORE-->> Get my LADSPA plugins
  Reply With Quote
Old 7th June 2016, 03:31 PM   #10
DPH is offline DPH  United States
diyAudio Member
Join Date: Dec 2008
Location: Portland, OR
Even with the base stoppers, your output stage ft should be pretty high. I'd bet you can pull down that 20k thd with some tweaking of the dominant pole compensation.

Fun, ridiculous project, I approve. :-)
Happy DIYing, Daniel
(This space is occupied by the artist formerly known as Derfnofred)
  Reply With Quote


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
Seeking Japan small signal transistors Real_Macgyver Parts 1 23rd May 2011 02:15 PM
small-signal (small-power) fast switching complementary transistors Lola Luna Solid State 3 8th April 2010 07:51 PM
Safe use of unknown small signal transistors mashaffer Parts 8 29th July 2008 12:35 AM
Using Many Small Signal transistors for the output stage Update!! John Biles Solid State 10 14th August 2005 04:58 AM

New To Site? Need Help?

All times are GMT. The time now is 03:34 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2