MONOLITH MAGNETICS

150W SE AMP BLOG: the mean machine

I will use this blog page to share my findings on the development of a very powerful, yet refined single ended amplifier. I started this project several years ago and have been working on it since. For the last two years, I have been using this amplifier almost exclusively to drive my main speakers. Keep in mind that it is not the typical nicely crafted good looking DIY amplifier, but rather a highly experimental machine which is very user unfriendly and very dangerous to operate. A friend of mine called them ‘the mean machine’ upon seeing the amplifiers. I fully agree with him … the intensive orange red glow of the anode structure has a seductive attraction to people visiting my listening room. This tube would be my favorite device to power up the electrical chair in case it was needed for me . However, this is not likely to happen in the near future. Let’s get serious now!

The early days

I have designed quite a lot of push–pull and single ended amplifiers in the past that made use of amorphous output transformers, various directly and indirectly heated power tubes and I experimented with different kinds of tube stabilized power supplies (series as well as shunt). After quite a while, I noticed that single ended amplifiers did the right things for me. However, the low power aspect of this concept proved to be a bottleneck on some extreme recordings in my very specific case. I have a very large dedicated listening room at my disposal (lucky me) and my speakers have a rather low sensitivity of around 88 dB/W/m. I got interested in the idea to develop a high power single ended amplifier probably running on one tube. Personally, I am a great fan of the WE300B as is Dick van de Merwe which we all know as ‘triodedick’. The most easy and logic way would have been to make a double or a quadruple SE 300B amp.

The challenge

But, as usual, I needed a challenge to proceed. The basic idea was to design and build a >100W SE triode tube amplifier featuring one directly heated tube as I like simplicity. But simplicity comes at a price. After some research on the internet, I did read about the famous 833 based Wavac amplifier that was covered in a Stereophile article.

Wavac 833 amplifier	silvercore 833 amplifier	glowing red @ 300W

I started to work on an amplifier that runs into class A2 and used the well-known directly heated 833A transmitter tube. I had never been working with "class A2 operation" and the high voltages required to run the 833A. I also got in touch with Stéphane Puechmorel who already had a lot of experience with these tubes and even higher powered versions such as the GM-100 monster tube. He advised me to look for RCA or GE NOS tubes.

Gathering key components

I made a call to Michael Wiams from Halfin and I was pleased to hear that NOS 833A tubes were readily available from stock at a very reasonable price. I immediately ordered a pair of RCA tubes.

my RCA 833A tube	ooh, these red glowing tubes ...

Upon arriving home, the first idea that crossed my mind was to light the filament as this was some sort of lighthouse tube. I did not have 10V -10A at hand and I also noticed that the terminals were not really compatible with the standard stuff we all know. After some time, I got connected to a Dutch guy, Reinhout Devries. He was a few years ahead of me with a similar project. His 833A amplifier was already operational and very well built using an impressive amount of magnetic components. This must have been the heaviest amplifier I have ever seen. During a visit at his place, I noticed that he had very nice filament connectors. I was able to buy a pair of these (new in original box!) and went back home with a satisfied feeling.

833A filament connector out of the box	833A filament connector close-up

Next problem to be tackled was the output transformer. I had very good experience with the use of amorphous output transformers for SE as well as PP operation. Amorphous metals were also the main topic of my Ph.D. research many years ago. I actually produced my own amorphous metal strips as well as quasi-crystal ribbons. In the picture, you can actually see the ribbon coming off the rotating polished copper wheel.

from melt to ribbon	the machine	the result

I decided to do myself a favor and I bought the largest amorphous double C cores (for domestic use) that were available from Metglas. These cores had a net cross section area of 46 cm² which should do the job. At this point, I had the iron but not yet the transformer. I had the tube but not yet the driver scheme or the power supply. It's time to investigate the 833A datasheet! Enough for now, … I will keep you posted as this is quite a long story to cover :-)

to the neXt page

//WDC