After getting back into HiFi a couple of years ago, after a offspring enforced absence, I ended up with a system that included, amongst other things a 6550 based push pull amp built from a world audio design kit. I was very happy with this, and did several of the usual modifications common to this type of circuit. Black Gate's on the cathode resistors, changing the coupling caps, that sort of thing. Much time was spent reading books on valve circuit design and wandering around the various DIY valve areas on the Internet.
I was at this time somewhat cynical about the sonic results some were claming for single ended circuits, especially those using directly heated triodes with no global feedback. Coming from a engineering background I was somewhat skeptical about this. However after going to a few shows and hearing a couple of 300b-based amps, I did begin to realise that there just might be something in all this after all. The turning point was going to a get together of valve amp enthusiasts, and hearing a 300b amp (WAD PSE 300b) defiantly made me start reconsidering my position. However whilst I could see/hear what all fuss was about with the 300b midrange, there was softness at the frequency extremes that I wasn't so sure about. At the same get together, I got a chance to talk at length to Paul Barker, who for some time had been extolling the virtues of SET's, just not ones made with 300b's.
A short time later, I had the chance to visit Paul for another valve get together, and this time get the chance to hear SET's built around more interesting valves. The highlight of the day in many ways was Pauls 212 amplifier, which as you can see from some of the photo's taken on the day, is still very much a major work in progress, but the sound of this amplifier, whilst in many ways held back by the lack of matching output transformers, confirmed my suspicions, that the parts of the SET sound I was unsure off was a product of the 300b output valve, and not the single ended topology. This was confirmed, by hearing a commercial 300b amplifier on the same day, it was very nice, but just not as great as the sound the 212 made.
If the only amplifier I had heard that day were the 300b and 212, I would have gone home to my 6550 PP, and sulked, but done nothing further. There was no way I could see myself surviving (1400v B+), or affording to do anything with a 212 (or even get hold of a pair). But there was another couple of amplifiers there that day, both built by Darren Dean, who had gone though the process of discovering SETs a year or so in advance of me. He had two amps there, one a 6b4g based one that was very nice, but somehow, the amp that really made me think I could get involved in this stuff, was a little parallel 6em7 amp, producing 3w, or with two of the valves removed about 1.5w. But despite the lack of power, this little amp seemed to have all the qualities that I found so attractive in the 212 amp, but one big advantage, that of lower cost.
The 6EM7 is a interesting little octal valve. Looking into it a bit more, I discovered it was a dissimilar dual, triode, that is, it contains two triodes in one envelope. Originally designed to be used in TV's, they were produced in there thousands in the early 1970's, and are still readily available at low cost. Paul, I think described them best, as being similar (though not identical, so beware) as a 6SN7 and 2A3 in one valve (which makes them a ideal audio combination). For reference, I have put a copy of the 6EM7/6EA7 data sheet here; this includes the curves for the second triode. The first triode is referred to these curves for 6DR7. Both these sets of data come from the excellent 7JNP Tube Database search
This valve looked ideal for a first time scratch builders like myself (I had built a phono stage before that, but this was my first attempt at a power amplifier). As always, money was tight, so it became a case of slowly assembling the bits and pieces to build the first try. A pair of output TX's from eBay, a Maplin choke, and so on. Anyway, eventually I had enough to build something that stood a chance of working so this was the first attempt. (clink on the images to see full size).
From building the phono stage, I had discovered how well NiCad batteries worked to provide bias for the input stages, so I used that. For the coupling cap, I used a pair of Teflon caps that I had got ome of to use in the phono stage. The power supply was cobbled together, from w hat I had.
To my surprise, after a bit of fiddling with the power supply and output stage bias, I got the thing working on a breadboard (well actually a blackboard I found in the kids old toy box), and producing music. Ok, I was biased as I had just created it, but I was somewhat taken aback by just how good the music it produced was. Ok, it only produced 1.5w, but it was driving my ProAc Studio 125's loud enough to listen to. I was hooked. I found I wasn't using my 6550 PP, but the little SET instead.
For the power supply, I used a hybrid bridge, I didn't have a center tapped transformer of low enough voltage to get the 300v I needed, but I wanted to use a valve rectifier if I could, this produced the circuit shown above. The actual power supply isn't a thing to be proud of, as it can't provide much current to the output stage because of the dropper resistors used to get the voltage down to the 180v I needed for the output stage. If you want to produce a keeper 6EM7 amp, I would look at something like this By Gary Kaufman.
The output transformers also were a weak point, and the amp was only flat down to about 100hz, dropping of quite quickly below that, -3db being at something like 80hz, it was however flat way above 20khz, -3bd being at about 40k.
Despite all the faults of this little amp, it just sounded so good, and I pushed it on as many people as I could, and they all commented on just how nice it was. This was the start of a long and slippery slope...