Designing and Building
Audio Amplifiers.

50 Watt Per Channel EL34 Ultra Linear Amplifier.

Special Notice.

After putting the amplifier into operation I found a few changes were necessary to the power supply. These changes appear in red so you won't have to read through the entire text searching for them. These are what a manufacturer calls post production changes. Believe me, it happens frequently.

All photographs were taken immediately upon completion of the project. The photo of the power supply will show 5 zener diodes instead of 4. Neatness fell by the wayside as these post production changes were made. Also I have used a little touch-up paint on the scratches on the output transformers.

Doing it the right way.

 Photo of completed amplifier and power supply.


On other pages of this site I have talked about designing amplifiers the right versus the wrong way. This one was definitely designed the right way. I spent several months and more than a little money finding the best output transformer. Then I investigated various output tube types and configurations. Some I didn't write up and post. Then I spent many days looking into various phase inverter and driver configurations, again many I didn't post. This amplifier circuit has distortion of less than 1 percent before the global feed back is applied. When the feedback is connected the distortion almost becomes too small to measure with the equipment I have.

In the breadboarded version of the amplifier the figures came out exactly as reported on the Ultra Linear Monoblock page.


Although you may think of a quartet of EL34s as overkill it was the only way I could get the low distortion figures I wanted. Frankly I don't see how some manufacturers of commercial gear got what they claimed back in the 60s. Maybe they didn't and were just claming they did. But that's a discussion for another page.

This circuit appears to many people to be a Williamson. The first amplifier stage is not directly coupled to the split load phase inverter but that appears to be the only difference. The phase splitter is followed by an amplifier and the output tubes are operating well below maximum power. Also, I have used high mu triodes instead of the low mu ones used in the original Williamson circuit. This permits the use of a lot of cathode degeneration which reduces distortion to a very low level.

I have broken up the article into several pages because of the size of some of the graphics files. The links below will permit you to navigate around only seeing what you want to.

A foolish man dreams of wealth,
a wealthy man dreams of wisdom,
and a wise man dreams of tubes aglow all in a row.

Amplifier Smorgasbord.


Or use your "Back" button to return to where you were.

This page last updated April 16, 2006.