Tube amp

[Pentium100]

I have always wanted to build a tube amp but could not afford the output transformers. Now I finally can, so I decided to built the amp. As I do not have a lot of skill in making a case, I first tested this with a simple and cheap phono preamp using a broken ATX power supply for the case. It turned out good, though I spent too much time shielding everything before finding out that the magnetic field from the power transformer couples into the cartridge and results in hum.

Anyway, with that done, it was time for a power amp. Because I wanted this amp for my computer room, it had to be medium power - the room is noisy and the speakers not very sensitive, so a 1W amp would not be enough, but the speakers are only 25W, so the amp did not have to be very powerful. I decided on triode SE, so it would be simpler to make (as I am still learning I wanted to start simple and maybe then build a big PP amp). The output transformer availability (at least for the cheaper ones) limited me to either 6S33S or 6N13S (6AS7G), so I tried both tubes and decided to use 6N13S (both sides in parallel) resulting in less power but smaller amp size (it had to be 30cm width or less to fit on the shelf I chose for it).

I also wanted a level indicator - got a 6AF6 magic eye for that and 6H2P double diode for detection.

The end result is this (the case is not very pretty, but I managed to make it and consider it an achievement):

Tubes used:
6?13? (6AS7 - both halves in parallel)
6?3? (2C51 - input stage)
6?1? (kinda like a higher voltage version of a ECC88 - drive stage)
5?3? (5U4G rectifier)
6?2? (double diode, for the level meter)
??-1? (voltage stabilizer)
6AF6 (level indicator)

Output power is about 8W. It sounds a bit different than a transistor amp and I like it

This amplifier does not have a single semiconductor device in it Not that it affects the sound, but I did it as a challenge - using only devices that were available when tubes were mainstream. I have also designed an input selector using thyratrons and relays, but did not include it in this amp becasue I have an external selector in this room, but it will be included in my next preamp/power amp.

[calexanian]

There you go.. BTW be careful with those ST bulb Russian (Sovtek) 5U4 tubes. Lots of catastrophic filament failures with those. They now make a 5U4GB which may not look as nice, but is far far more reliable.

[Roeland_R]

Looks great.

[T3sl4co1l]

Cool.  Any problems with getting enough grid drive voltage?

Tim

[Pentium100]

There you go.. BTW be careful with those ST bulb Russian (Sovtek) 5U4 tubes. Lots of catastrophic filament failures with those. They now make a 5U4GB which may not look as nice, but is far far more reliable.

I have NOS tubes - made by Svetlana in 1967. I do not know about the reliability, but these were cheaper.

Any problems with getting enough grid drive voltage?

Not really the 6N1P can get about 160 or 180 Vpp (IIRC) which is enough to drive the output tubes into clipping.

[Quarlo Klobrigney]

"Those tubes are no good because they have big black spots on them"

[AF6LJ]

It is a nice project, I love the magic eye indicator.....
Shame we can't get you some 6L6GCs or EL34s.
It is very cool.

[Pentium100]

It is a nice project, I love the magic eye indicator.....
Shame we can't get you some 6L6GCs or EL34s.
It is very cool.

Actually I have about 17 6P3S (which is the equivalent of a 6L6GA IIRC), some of them will be used some day

[AF6LJ]

It is a nice project, I love the magic eye indicator.....
Shame we can't get you some 6L6GCs or EL34s.
It is very cool.

Actually I have about 17 6P3S (which is the equivalent of a 6L6GA IIRC), some of them will be used some day

Very cool I had a tube amp that used a pair of 6L6GCs ages ago it was fun.

[Radio Tech]

Very nice! I like it very much. Dont worry about the chassis, looks good to me. I built a couple over the years, And the chassis are just sheet metal bent by hand. One of the last ones I built was a Fender Champ clone. All parts were from my vintage supply.


It was not quite finished at tihe time of this pics. Someone wanted it so I had to finish and let it go.

[oldway]

A tube amplifier is as good as its output transformers.

What do you use?

It becomes difficult to find good Hi Fi audio transformers.
I know the Dutch firm Amplimo that makes still high end audio transformers.
http://www.amplimo.nl/
http://www.ringkerntrafo.nl/shop/high-end-audio.html
Make a frequency response test of the amplifier, if your output transformers are not good, you might get a nasty surprise ... a frequency response limited to a few kHz ...

I prefer a class AB amplifier instead of a A class amplifier.
In the very late years of tube technology, Philips had developed a very good audio output pentode (EL503), but now, it is no more available.
I would choose EL34 as output tubes.

[Pentium100]

The output transformers are Indel TGL20/003. It is rated for 20W output, 200mA DC and 20-20k frequency range.

At around 2W power, I measured 18-20k at -1dB and 13-40k at -3dB. Seems good enough to me. I did not measure the frequency response at full power though.

As I was not sure of my ability to complete this amplifier, I chose to build one that's simpler and less expensive. As some point in time I will build a double amp for my media room (a preamp with a couple of watts power and a big ~120W power amp - that one will have to be Class AB).

[oldway]

Very good...
Check it with 1Khz square wave input to see if there is no oscillations.
And enjoy it...

[Quarlo Klobrigney]

Triode output and amplifier stages are always cleaner than a pentode output. Pentodes have more distortion.

[oldway]

If you look for low distorsion, you would choose a transistor amplifier not a tube amplifier.
Distorsion of a transistor amplifier is 10 x lower.
My choice is the oldschool Marantz 2285B, that's a true beauty...

[AF6LJ]

Triode output and amplifier stages are always cleaner than a pentode output. Pentodes have more distortion.

There were feedback circuits implimented to deal with that.

[AF6LJ]

If you look for low distorsion, you would choose a transistor amplifier not a tube amplifier.
Distorsion of a transistor amplifier is 10 x lower.
My choice is the oldschool Marantz 2285B, that's a true beauty...

This is exactly true.
The guy that founded South West Technical Products knew this, wrote several articles on the subject. Complementary Symmetry designs, with low stage gain.
http://www.tigersthatroar.com/

[dom0]

Distorsion of a transistor amplifier is 10 x lower.

Contemporary transistor signal level amplifiers exhibit harmonic distortion of less than -120 dB, while you can get below -100 dB using (semi-)integrated power amplifiers/drivers.

[oldway]

Triode output and amplifier stages are always cleaner than a pentode output. Pentodes have more distortion.

There were feedback circuits implimented to deal with that.

Feedback in tube amplifiers is limited to a low value to avoid oscillations and instabilities. (for reason of phase shift in output transformer)
But with some schematics, it is possible to succeed a higher rate of feedback and an harmonic distorsion as low as 0.33% in the full frequency range (20 to 20Khz) at full power.
For example, the Mac Intosh 40.

[AF6LJ]

Triode output and amplifier stages are always cleaner than a pentode output. Pentodes have more distortion.

There were feedback circuits implimented to deal with that.

Feedback on tube amplifiers is limited to a low value to avoid oscillations and instabilities. (for reason of phase shift in output transformer)
But with some schematics, it is possible to succeed a higher rate of feedback and an harmonic distorsion as low as 0.33% in the full frequency range (20 to 20Khz) at full power.
For example, the MacIntosh 40.

It is a nice amplifier but....
This shows the extent that one had to go to in order to attain harmonic distortion that is thirty or more times grater than solid state amplifiers of the seventies.

Tube amps are fun, but they have their limitations.
That Mac Intosh is like a vintage Rolls Royce, a fine example of excellence in design and the state of the art from the past.
I love the old stuff, and respect its limitations.

[oldway]

As you see, this amplifier use "cathode feedback" for the power tubes.
Quad II use more or less the same schematic.
This kind of special output transformer is still available but very expansive.
http://www.ringkerntrafo.nl/shop/high-end-audio/uitgangstrafo-s/specialist-range-opt/421/vdv2100cfb-h-balans-ugt--raa-2000-ohm.html

[AF6LJ]

Considering the labor involved that is a decent price for that transformer.
The cathode feedback did not escape me

[Pentium100]

If you look for absolute lowest THD then transistor amps are better. However, since the distortion if my tube amp is primarily second order (since it's SE and has no feedback), the THD can be higher and it may still sound OK.

If I turn the volume on my amp a lot, the waveform on a scope looks bad, but it still sounds kinda OK - somewhat different than at a lower volume but still OK. I tried to calculate the THD by using the FFT function of my scope (Rigol DS1052E) and got about 2% at ~8W (and 0.6% at 4.5W), however, I may have calculated it wrong.

[vk6zgo]

If you look for absolute lowest THD then transistor amps are better. However, since the distortion if my tube amp is primarily second order (since it's SE and has no feedback), the THD can be higher and it may still sound OK.

If I turn the volume on my amp a lot, the waveform on a scope looks bad, but it still sounds kinda OK - somewhat different than at a lower volume but still OK. I tried to calculate the THD by using the FFT function of my scope (Rigol DS1052E) and got about 2% at ~8W (and 0.6% at 4.5W), however, I may have calculated it wrong.

O.6% is not very good for a serious tube amplifier----2% is appalling!

The old "Type 3" Monitor Amplifiers we used did better than your 4.5w figure at 20watts out.
Of course,this was done the old way,with a resistive dummy load & a "Noise & Distortion Test Set".
I'm a bit dubious about the Rigol's FFT as an alternative.

Some of the fantastic results achieved with modern solid state gear raises the question-----can our ears do better than the amplifier?
I would suggest not!

[dom0]

Some of the fantastic results achieved with modern solid state gear raises the question-----can our ears do better than the amplifier?
I would suggest not!

Of course! But only for people with reference-grade ears.