I must have built nearly a thousand preamps and maybe a hundred power amps when I was doing Supratek.
Most of the power amps were push pull amps , the single ended amps I built were mostly using 6C33C-B Russian "Mig" tubes that are able to output 15 watts with low output impedance.
15 watts is a good compromise for the majority of speakers . The lower power DHT tubes like the 300B, 2A3 , 6B4G, etc will drive a lot of speakers but with serious compromises in regards to dynamics , response , etc when asked to drive speakers that need some oomph.
If you are going to run these low powered amps you need speakers with very high efficiency , at least 90dB , preferably higher .
Now that I've retired I've found myself with a lot of NOS tubes and some very good output transformers , and time on my hands.
So here is a description of what will be most likely my last build of a Single Ended amplifier- an amplifier I'm going to throw everything at , and will reflect my many years of tube audio.
I'm still very much in awe of SE amplifiers, the hype about their superior sound quality is quite true, and there are good technical reasons for it.
BUT, you need a very good high-efficiency speaker , that is accurate, transparent and is an easy load for the SE amp.
Or maybe use a low power SE amp to drive a high efficency mid or full range driver in an active 2 or 3 way system.
(Use a high power tube or sold sate amp to power the bass drivers, as I do with my Open Baffle speakers.)
An 85-90dB speaker will not bring out the best features of a SE amp - you will be better off long term with a DHT push pull amp built to the same level of quality .
The heart of this new amp build is the Tango X-10S output transformer (OPT)
I swapped some original 1930's Western Electric 300B tubes for a pair of original Hirata Tango X10-S opt's many decades ago with Nobu Shishido , a highly regarded Japanese tube guru .
The X-10Swas designed to be used with 211 and 845 DHT tubes, and I built 3 or 4 versions using these tubes and the X-10S
Then the opt's sat on a workbench for maybe 20 years . I was going to sell them- you see them advertised for ridiculous prices , but I kept them as I regarded them as state of the art devices even though they were +40 years old.
The original plan was to use them with Australian manufactured 1930's VT25's, which weren't VT25s at all , but very close substitutes to British DET 25 and CV1025 tubes, able to take 1000 volts on the plates and probably a very good match with the X10-S ( there are no datasheets for the ancient Australian VT25)
However, much to my dismay I discovered that the filaments of these tubes were all open circuit , and no amount of careful surgery could fix them.
There's not many tubes that work with the Tango X10-S primary impedance of 10Kohms.
The 211 is the ideal tube to use with it, but I dont really fancy working with 1000+ again volts at my age.
Most DHT tubes use 2kHz - 5Khz opt's . But the equally ancient 10, 10Y, 801A and the American VT25 all need a high impedance opt , so I decide to use 801A tubes .
In Class A1 an 801A can give a maximum of 3-4 watts output with a plate dissipation maximum of 20 watts.
However, with interstage transformer coupling between driver stage and output tube , and with fixed bias in this circuit, it can provide around 6 -7 watts deep class A2 , at the onset of clipping .
Plate dissipation goes above the maximum at 22watts, but these values tend to be conservative, and I have plenty of 801A tubes.
I tend to prefer single tube driver stages for my SE amps - they have a purity that seems to become veiled with multiple tube driver stages.
You need a high gain tube , but it also has to have the swing and current to drive the 801A tube.
The later generation frame grid tubes such as the 7788, D3a, E280F, 417A, 437, 6E6P, and others are all suitable tubes to use with transformer coupling , and I have used all of these over the years with the Sowter interstage transformer.
But the finest driver tube I've used with the Sowter 8423 interstage transformer (IT) is the very rare and very expensive STC 3A/167 tube.
The 3A/167 has enormous gain, tremendous drive and works perfectly with the 5K:5K impedance of the 8423.
The measured 1kHz square wave is just about as good as you will get with a SE amp , and even more impressive as there is zero feedback used in the amp.
A word about measurements , and square waves in particular:
Measurements do not define the sound of an amplifier in absolute terms, but they do give a very good indication of how an amplifier is performing in a technical sense , and provide real clues to the sound quality.
A 1khz square wave will give you an enormous amount of information if you have the experience to understand the wave form.
A perfect square wave does not exist , but its not too hard to get it looking almost perfect with a solid state amp loaded with lots of feedback , like this.
Its a bit trickier to get a tube amp design with a perfect square wave, and in truth you really wouldnt want to.
Tube amps have lots of resonances from the transformers , the capacitance , inductance and reactance interactions, and these all add up to bend the square wave out of shape , or more correctly into the shape of the response.
Have a look at this square wave:
Its taken from the measurements of a $350,000 dollar amp reviewed in an American audiophile magazine.
Frankly it's a shocking result , and I would be ashamed to sell an amplifier with this sort of response- it indicates a severe bass problem , and I would imagine it would be just about unlistenable if you were half interested in an accurate presentation.
In fact, its so bad I suspect the amplifier may have been damaged , or the measurer made a huge mistake - I cant imagine a third of a million dollar amp could be this bad.
Again, you can "fix" some problems with feedback, but there's a heavy price to pay for that.
Back to the square wave response of the 3A/167-801A amp.
No feedback , there's a bit of ripple on the top leading edge , which indicates a little bit of peaking at high freq >20Khz , but it wont be audible , and can be adjusted to suit the ears with a damping resistor on the interstage secondary.
Its nice and square , and indicates a very flat response from well below 15Hz to above 20khz.
Its not perfect, but a perfect square wave usually indicates some , or a lot of feedback in a tube amp.
This amp will have extended treble and a punchy , controlled bass . The Tango's are renowned for a nice midrange so that should be a given.
Power supply:
The high voltage B+ supply uses UF5408 diodes in a CLC configuration, followed by a GZ37 tube rectifier used as a damper for a slow start. All polypropylene caps, no electrolytic caps.
The filament supply for the 801A is a voltage regulator , followed by a constant current source regulator for the 7.5V DC @1.25A . I find this configuration to be very quiet and it sounds better than a straight voltage regulator for DHT filaments.
The bias supply uses a voltage doubler CRC configuration , followed by a 75 volt gas regulator , which is then in series with a pot for fine bias adjustment .
The 6.3V DC for the 3A/167 is a conventional voltage regulator.
There's nothing revolutionary or special about the design - it's a good solid design that uses high quality components , is well built , performs very well on the test bench , is built to last a long time and be very reliable. Each monobloc weighs 20kg.
The components and tubes are worth over $5000 ( the Tango X10-S and tubes are very expensive) so to build this as a commercial venture would not really be viable .
I don't subscribe to the ridiculous prices amateur and AI guided unprofessional builders are trying to extract from ignoramuses these days.
If your going to buy an expensive amplifier make sure its built by a person or company with many years of experience , and the cost reflects the components and engineering in it.
Ask for a set of measurements , ask lots of questions - use the same thinking you would use to buy a car . Dont rely on forums or AI for unbiased facts- you'll only get opinions or rehashed info that means nothing.
Sound of the 3A/167-801A
It sounds very much like a 211 SE amp , with a lot more grip and power than its 7 watts would suggest.
Its rich and organic sounding, very good bandwidth , but still very realistic and natural sound.
They have turned out pretty much as I hoped for , and I'm enjoying them very much.
They have Supratek badges on them, but I'm retired now and not making any more power amps , these are for my personal use only.
