[tomb]

   
Tubey. It has rather high distortion but mostly 2nd and 3rd, so it sounds rather euphonic. See http://www.pmillett.com/starving.htm
 
 
That's the price for running valves at lowish voltages but it makes for an amp you're never tired to listen to.

Fred's version in this thread uses the 12AU7.  I don't think it's been tested the same as Pete did with the 19J6.  It may have different performance results.

 

[00940]

  Fred's version in this thread uses the 12AU7.  I don't think it's been tested the same as Pete did with the 19J6.  It may have different performance results.

 
Due to the topology, the distortion pattern will be largely similar. The exact values might differ but probably not by much (and ltspice agree). I didn't want to sound derogatory. Those are actually pretty good figures for an amp running open-loop and at those voltages. Heck, my monster mu-follower actually has grossly similar thd specs... into 300R (at higher voltage output of course). I like it better than my previous tube amp, which had a lot of feedback, WCF output stage and low distortion.
 
Furthermore, such a thd pattern doesn't sound wrong to the ear. I had five minutes of fun with ltspice (simulating the 12au7 version). If you remove the cathode bypass capacitor, you cut the total thd from about 0.5% to 0.13%. You also cut gain by two (a benefit rather than a loss imho). However, this reduction goes with a massive reduction in 2nd harmonics (see attached fft, red with bypass caps, green without, 1Vrms into 32r, 20khz), due to local feedback. Some people might prefer that balance in between second and third harmonics but I'd guess that most would prefer the original.
 

 

[tomb]

   
Due to the topology, the distortion pattern will be largely similar. The exact values might differ but probably not by much (and ltspice agree). I didn't want to sound derogatory. Those are actually pretty good figures for an amp running open-loop and at those voltages. Heck, my monster mu-follower actually has grossly similar thd specs... into 300R (at higher voltage output of course). I like it better than my previous tube amp, which had a lot of feedback, WCF output stage and low distortion.
 
Furthermore, such a thd pattern doesn't sound wrong to the ear. I had five minutes of fun with ltspice (simulating the 12au7 version). If you remove the cathode bypass capacitor, you cut the total thd from about 0.5% to 0.13%. You also cut gain by two (a benefit rather than a loss imho). However, this reduction goes with a massive reduction in 2nd harmonics (see attached fft, red with bypass caps, green without, 1Vrms into 32r, 20khz), due to local feedback. Some people might prefer that balance in between second and third harmonics but I'd guess that most would prefer the original.

With respect, you sort of confirmed what I was trying to say.  You can guess certain flavoring from the topology, but with the tube providing the signal amplification, it's pretty much that - a guess - unless you listen to and test the actual article.  Fred's version may, in fact, be better ... or it could be worse.  I strongly suspect the latter is true with the 12AU7 and Fred's resourceful PCB, but it's more than I'm willing to state - not having heard it -  if someone is asking for a sound comparison relative to another amp, LTSpice notwithstanding.

 
Just an FYI, but you might have overstated the significance of the 3rd harmonic as well in Pete's original design - the 2nd harmonic was absolutely dominating.  I'm not trying to be argumentative, but it's important when statements are put out there about the quality of an amp's sound or in a comparison.  When you stated, "It has rather high distortion but mostly 2nd and 3rd ..." it could imply that the amp is harsh, when it's not.  That's because you make no distinction between the levels of the even and odd harmonics.  If I'm interpreting correctly, the 3rd harmonic (odd harmonic) is down by 75dB over the signal - that's inaudible with almost every headphone. 

 

[00940]

For the sake of clarity (not argument, I swear):
 
 
Quote:

   You can guess certain flavoring from the topology, but with the tube providing the signal amplification, it's pretty much that - a guess - unless you listen to and test the actual article.
 

 
I'd rather say that from the topology, you can determine a basic sound (the harmonics relative weight, a distortion rough ballpark range) and changing tubes (as long as they are suitable for the topology) and pcb will only change the flavor at the margin. Whatever the layout and the tube used, no ssmh will sound harsh nor have vanishing thd (unless you make a mess of things or bring it to oscillation).
 
But, hey, aren't we all striving for those changes at the margin ? So I certainly wouldn't dare to guess if Fred's version will sound better than Pete's. But I'd certainly be willing to say that both will sound quite different from an O2 (and similar one to another compared to the O2).
 

 Just an FYI, but you might have overstated the significance of the 3rd harmonic as well in Pete's original design - the 2nd harmonic was absolutely dominating.  I'm not trying to be argumentative, but it's important when statements are put out there about the quality of an amp's sound or in a comparison.  When you stated, "It has rather high distortion but mostly 2nd and 3rd ..." it could imply that the amp is harsh, when it's not.  That's because you make no distinction between the levels of the even and odd harmonics.  If I'm interpreting correctly, the 3rd harmonic (odd harmonic) is down by 75dB over the signal - that's inaudible with almost every headphone. 

 
More for the information of the public than mine...
 
My comment about distortion being mostly 2nd and 3rd harmonics is as plain as it can be : distortion in ssmh is mostly 2nd and 3rd. No intention to comment on respective levels. I don't see how I could have implied that the amp was harsh while stating in the same post that ssmh are "tubey", "euphonic", "amps you're never tired to listen to" (judgements actually based on the 2nd strongly dominating, as well as experience, having built such an amp).
 
If you want, we can rewrite it as "it has rather high distortion, mostly 2nd harmonic and a bit of 3rd, and nothing above".
 
Btw, 3rd harmonics aren't responsible for amps sounding "harsh". Otherwise almost all amplifiers with a push-pull output would fall in that category. You need higher order odd harmonics for things to go unpleasant (5th, 7th). Iirc, about a person out of three actually prefers amps with dominating 3rd (I could dig a reference from Nelson Pass about that).

 

[tomb]

OK - I give up.

 

[Fred_fred2004]

It's just a good fun build

 
and cheap

 

[tomb]

  It's just a good fun build

 
and cheap

 
Good statement, Fred!  Thanks for your contribution to the community!

 

[wakibaki]

I already have an O2, makes sense to still get this? Does this have tube like sound or more SS like sound?

Most people don't really need an amp, however I've got at least 6 different ones, so if it's not going to break the bank and it's maybe going to teach you a bit in the process of building it, why not? And then you're going to know the difference between SS and tube for yourself, if you can hear it.

w

 

[doors666]

In the schematics, I see a single triode being used for each of the channels, but in the build pics and pcbs, I see two twin triodes 12au7 being used. what am i missing here.

 

[Fred_fred2004]

I think you are looking at two triodes but they are in the same tube, have another look you'll see 2 anodes 2 cathodes etc in each tube
 
cheers
FRED

 

[doors666]

i expect to see one tube, but in the first post of this thread, the pcb images and completed build pics have two tubes.

 

[doors666]

Ok, I made a mistake, I see that both triodes are being used for a single channel and both the triodes in a tube are shorted to each other.. I dont understand why its done like this though.

 

[00940]

Ok, I made a mistake, I see that both triodes are being used for a single channel and both the triodes in a tube are shorted to each other.. I dont understand why its done like this though.

 
-1- in the original amp, the 19j6 is a dual triode with only one heater for both sections. So, as you use the heater as load, you need one heater by channel, so two tubes.
-2- a nice consequence is that running the triodes in // reduces the plate resistance by 2, which decreases distortion in the 33K plate load.
-3- When substituting with the 12au7, which has a pair of 6.3V heaters, you could in theory just use one tube, grounding the middle point of the heaters and using each half as load for one channel at 6.3V. But (and remember, there's nothing important in front of a but), you probably would increase crosstalk, increase interaction from heaters to cathode, increase thd (by increasing ra), increase dissipation in the mosfets by 1.8W each. And the layouts made for the 19j6 have to be redone completely.

 

[HIFIbox]

Are these still available?
PM sent
Thanks !

 

[RealSpark]

Just finished my build today. Here are the photos. 
 

 



 
 
Very wired that the sound on one channel is normal and the other is not. On the other channel the sound is delayed like echo and the base is totally wrong. I have checked all my circuit and did not find anything wrong. Tried to swap the two tubes and nothing changed. 
Do you guys have any clue about what I did wrong? Maybe I should use less wire? I used two GE 5814A to replace my 12AU7. Very wired that one of my 12AU7 does not work on this build... The heater just won't light up. They both works find on my another single 12AU7 Amp...
Please help me. Thank you.