Tube Matching in a Power Amp?

[TexAus]

I'm very new to tubes and have ordered a ST 120 kit from Bob Latino.
I would like to try a set of the Mullard CV4024 an some point.
I have read conflicting information regarding the need to match these. Some say all tubes should be matched, others only the power tubes. What are your thoughts on this?
Also, any chance these will disappear from the market in the next 30 days or so? If so, I may need to go ahead and order some before they do.

[SY]

I think you mean "power amp," not "preamp." In any case, there's lots of good ECC81-types around- the Mullards are very good, there are many others as good or better. Presumably, these amps use a feedback loop which minimizes the differences between input tubes. Don't sweat it.

Good matching of output tubes within a channel can make a significant difference in noise and distortion. Between channels, not as important.

[TexAus]

You are right SY - power it is.
I was reading something on pre's and the article polluted my post.
Thanks -
What ecc81 tubes have you listened to that you would recommend?

[SY]

I've run a bunch of them in my RLD amp and consistently prefer the JJ. If you want to experiment, I have quite a few different types on hand (I may even have some of the Mullards left) and I'm local to you.

I'll edit your post title.

[TexAus]

That's the best news I've had in awhile SY.
I am really exited about getting back into playing around with this stuff. The amp kit is scheduled for deliver on Tuesday and it'll take probably 3-5 days to assemble and test it. I'll see how industrious I feel once I get started on it.
I don't know if your interested but if you haven't heard listened to one of these revisions of the old Dynaco stuff it could be an opportunity.
I got into the better audio stuff fairly young but even then it was at the end of the tube era. I only have owned two tube amp in my life and it they were an old ST 70, and a Audio Research D52. I was probably to in experienced form a decent opinion on the ST 70 but the D52 I tried didn't really sound all that great. It probably needed some TLC and possibly a fresh set of tubes. I did listen to a high end VAC I liked a lot and a Mc 275 that I also liked but both only briefly.
I am very interested in the tubes you may have.
This tube adventure is going to be fun!

[SY]

Haul it over (I'm in Northeast, the Techridge area), we'll hook it up and try a few different candidates. Drop me a pm for contact info.

[TheGimp]

Since you specified the ST-120 Sy's answer is sufficient. In more general terms Matching for a Class A amp is not as important since (1) Class a is normally fixed bias and (2) class A never goes into cutoff.

For class A there is not a great deal of difference in distortion with tubes which are off by 10% in their bias point. It is best to stay within 5% or less if possible to minimize dc bias of the OPT. 5% differnce in tubes biased at 60mA (6P1P-EV, 6BQ5, etc) amounts to 3ma bias of the transformer.

Mismatches in Gm can be even wider and will effect 2nd Harmonic slightly. The biggest effect is increased hum as Gm differntial increases as matching is responsible for cancelation.

In class AB1 things get a bit more complicated. one advantage of AB1 is most amps in this class are fixed bias and you can adjust for minimum bias difference to minimize transformer dc offset. At low signal levels the amp is in class A and the above apply. However when larger transients come along the amp switches to AB1 with one tube eventually cutting off (class B). If Gm is not matched closely, the amplitude of the two halfs of a sine wave will not be equal. Second harmonic distoriton will increase along with hum at a much more rapid rate than with class A.

[TexAus]

Thanks TheGimp -
That makes sense, things are beginning to come together for me - in terms of tube matching at least. I'm looking forward to learning more about tubes, tube amps, and designs.
The %'s are also nice to know -
I wondered if tubes within a single amp aged about equally or if the 'matching' was a moving target so to speak. I 'think' the bias setting is supposed to assist in this to a degree but many amps don't have a separate bias for each tube. Of course I'm basically clueless right now so none of the above may make any sense whatsoever. (grin)

[SY]

Just to make things more frustrating, matching is usually done at one point. If the transconductances aren't also matched, the tubes will show higher distortion and add a DC component to the primary.

None of this may be serious except in the case of a large mismatch, but it will help to use a distortion analyzer when swapping tubes around.

[TexAus]

Nice SY: challenging - yes. But an excuse to justify more toys (test eq) - this just keeps getting better!
The DC possibility sounds a little worrisome. Is this something that may reach the speakers at any significant level? (my tweeters are in fear as I type)
I have looked at some signal generators but many of them that are reasonably priced seem to have too much distortion to be all that useful for that purpose. I saw some very old HP's but haven't chased down the specs on those yet.
I've also seen some Rigol DSO's around that seem decent for the $ but I have to re-learn about all that stuff as well.
I will also need to chase down some sort of tube tester (more to learn).