Tube Matching

[Super_Reverb]

As I understand, asymmetry in output swing results in increased 2nd harmonic "distortion", so even harmonics are said to be pleasing to the ear as they are octaves of the notes that we are playing. They allegedly add richness and fullness, while depending on the relative proportions, odd harmonics add more crunch and grind - certainly an essential tool in most guitarists' toolbox.

If you analyze a typical preamp load line, it is obvious that with a symmetric input signal, the output swings asymmetrically due to inherent nonlinearity in triode and this asymmetry increases with output swing.

So... It seems as if a little mismatch in phase inverters and output tubes in PP would simply generate additional asymmetry, which CAN be pleasing within limits.

My question is why all the hype about ultra matched output tubes, PIs and the like?

Am I missing something? Is this heresy? Am I endanger of being exiled from the island? Is the emperor really wearing new clothes?


cheers,

rob

[David Root]

Matched output tubes and PI came from audio in the old days, when live distortion was the social equivalent of farting on stage.

Then along came Les Paul and Elmore James, who gave us the electric guitar and the purpose-built overdriven amp.

The reason Dumble type builders value matched PI and output tubes is because in HAD's design all overdrive is done before the PI and the power section is intended to be a clean audio type not adding any PI/power tube overdrive.

Tweaking the PI for maximum harmonics and offsetting power tubes from a match is doable, in fact HAD only used a single bias pot so if you pick the tubes off-balance you'll get some result there too.

If you really want to get into it the above is not all-inclusive. When tubes go into overdrive all the math goes oout the window.

[iknowjohnny]

I just mentioned this in another post, which may or may not have been what made you think to post this. But for me, technical jargon aside, it acts like a sort of high end and harshness roll-off Not sure about the harmonic content, but once i do this i have to turn the high end up more which i do with the treble and presence controls. The way i always look at things like high end attenuation, which is a big issue with most people, is that things like this can be done any number of ways and in any number of places, and there is usually one place and way that seems to work best.

I think a good example would be the rat pedal. I never really liked it that much because i felt it was mushy and the highs were harsh. I then found that if i turned the highs all the way down with the filter knob and then turned the highs up a lot on the amp, i got a fantastic sound. And thats the approach i take with building an amp. I'm not tech at all like many or most of you here. So i can only use common sense approaches that it doesn't take a tech to understand. So i can't talk about the theory you described, but when it comes right down to it the theory doesn't matter as long as the results are what you were after. And i find that unbalancing outputs lirterally enabled me to completely forget about high end issue in amps, which by the way before that was my #1 biggest issue....getting enough high end w/o being tinny or harsh. that problem no longer exists for me. I would never build another amp w/o dual bias.

[Cliff Schecht]

I have a lot to say about this and unfortunately not a lot of time today. I don't find power tube matching necessary or desirable in my amps for a lot of reasons. But as David points out it really depends on what you are trying to do.

I will post more on this later, gotta go for now! Get ready for a super music theory filled post .

[iknowjohnny]

I don't find power tube matching necessary or desirable in my amps for a lot of reasons. But as David points out it really depends on what you are trying to do.

Like was said, and it makes sense, it began in audiophile circles. That makes sense because when matched they are much truer to the input, and unmatched they would do what might be called ruining a signal such as recorded music. But what it does for a guitar signal is desirable, especially if cranked where matching could get much too bright and harsh.

[Phil_S]

I've seen it said, in the bad old days, the tube manufacturers did great quality control. Outliers were destroyed. Tubes of a given production run were probably fairly close in performance to one another. The need to consider matching was not as important. Today's production isn't what it used to be. Lot's of guitar amp people seem to think matching isn't all that important, but there is always an exception, as someone else pointed out.

[iknowjohnny]

It's actually quite ironic when you think about this....i installed a dual bias setup in my amp so i can mis-match my tubes !!! LOL! Actually thats not quite true, as i originally did it to match them. Or not so much match them as be able to use tubes that are very far apart in current draw.But It just happened tho that i ended up using it for the purpose of mis-matching them, albeit to just the right amount so that just using unmatched tubes wouldn't have been ideal. In any case i'm convinced a dual bias setup is essential. At least it is for me.

[Cliff Schecht]

There is a line between output tubes being "close enough" for 20% devices (1-5mA never bothers me with EL34s) but too much past this and the tendency of a push-pull stage to cancel hum goes away. In that respect, some ultra-high gain amps NEED matched tubes to operate properly as they are relying on the power stage for CLEAN amplification.

I tend to not have matched pairs lying around but almost all of what I have is close enough. Tubes tend to drift during their break-in period unless they are "pre-aged". If what I put in an amp sounds good, I use it and this should be the final criteria anyways IMO.

Most people use 100k/82k to get the best symmetry in the PI, this again is about as close to matching as is usually needed for clean amplification. Plus almost all Bassman derivative amps use a negative feedback loop between the secondary and the PI, thus canceling even harmonics this way. I like to use unmatched tubes in my Bassman build, they sound better when pushed (with a 12AY7 in V1 of course ).

As far as how the harmonic content of mismatching sounds, this is where a bit of "mojo" tone lies. Kinda like adding CC resistors for that extra bit of flavor. Some like what CC's add to their amp and others don't. Same with mismatching. There are a few amps that even let you dial in this mojo (sorry no examples come to mind ).


On an amp that's too clean mismatching will sweeten things up a bit and on an amp design that produces a lot of distortion on its own (or a lot of even in the preamps for Dumbles) it'll make the output sound buzzy a thin. If your second harmonic starts becoming a large portion of the spectrum then things really start to get "buzzy" like a sawtooth wave is known for. OTOH too much odd harmonics sound harsh and unpleasant, especially with anything other than a 1-3-5 chord (more complex chords sound unpleasant with too much odd). When you slam a triode with too much signal you get a hard squarish wave with abrupt transitions that sound as ugly as they look!

[iknowjohnny]

I may have misunderstood Cliff, but it sounds like you said a high gain pre with mis-matched outputs would sound buzzy. If you did mean that, then i'm surprised because mine is a cascaded pre, not real high gain but about the degree of drive in a peavey classic 30's drive channel. And mis-matching the outputs is if anything the opposite. It makes the amp sound much smoother in the top end and eliminates any harshness. Maybe i misunderstood, and if so nevermind.

[Cliff Schecht]

I was saying amps that produce a decent amount of even in the preamp will sound buzzy if you try to add too much more even. Dumble's come to mind with the nice combination of smooth high end and a sweet and strong even harmonic content (sounds that way but I haven't seen oscope screenshots) and I suspect they sound pretty bad with unmatched tubes. In something like the Express with a crap ton of gain you can get a lot of buzz (rectified 60 Hz hum) with too much mismatch in the output stage.

[iknowjohnny]

I see. Well, mine pulls thru somehow and sounds so much better unmatched with no buzz to speak of, at least nothing more than when they are matched. Maybe it's that the preamp isn't not a gain monster, tho I always thought the wrecks weren't either. I also find that tubes i wasn't as fond of before i now like. Winged C's for example. I liked them except that they seemed a bit too thin and bright like there were no mids to speak of. Mis matched i actually have been using them in place of my fav tube and now THEY may be my fav.

Maybe if i was into uber gain preamps like a 5150 or such i'd see what you're saying. But mine gets by i guess due to the amount of gain not being crazy at all. I'd call it classic rock gain. Sounds great with a clean boost driving it to crazy degrees tho.