Which 12AX7 for my 18 watt plexi build ?

[soundmasterg]

I wonder what would happen if Stan met Eric Johnson. They'd probably annihilate each other like matter and antimatter.

As I said in another thread: Tubes don't have filters inside, and so can't shape the frequency content of a signal in the way that tone controls do. But when overdriven, different tubes make different distortion harmonics, and they can shape the frequency content that way. A tube that clips hard, generating a lot of high order harmonics, will sound brighter than one that clips softly.

Stan is probably not too far off in his assessment of how much tubes really matter. But some of us like to tweak! Right now I'm rocking some Mazda 6L13s salvaged from a tube regulated power supply.

I would think that in a way individual tubes will have a quasi-filter inside because each tube has unique Miller capacitance effects due to the spacing of the elements....but I would expect that any tonal contribution from that would be small...certainly much smaller than the circuitry surrounding the tube stage. Still, it probably makes a subtle difference......

Greg

[Steve Conner]

Yes, the Miller capacitance can roll the high end off a bit. A 12AX7 can have something like 100pF, and in a typical scenario, the grid is driven from the wiper of a 1M pot, giving a worst-case source impedance of 250k. I don't expect it would vary a great deal between 12AX7s though.

This may explain why so many amps use a "bright cap" on the volume control.

[km6xz]

Yes, Miller effect is there but remember, the area and distance between the grid and anode that determines the capacitance also impacts the whole dynamic transfer function of the tube so if the Miller Capacitance is much different, the tube also has very different gain and curve family than a 12AX7. If there is enough difference in capacitance, even that impact on roll off( depending on circuit topography, it might be in the Mhz) will be swamped by the larger gain difference between that tube and a 12AX7. Having a lower bandwidth also makes it something else other than a 12AX7. The capacitance would have to be pretty large to impact 10khz, the upper reaches of the meaningful harmonics of a guitar. The differences in compression onset level between two random tubes due to gain differences sure has more impact on harmonic content than clean within-spec operation of a tube. So any tests need to be gain normalized so see if there is any audible harmonic differences. There are so many reasons subjective tests with uncontrolled variables are less than useful. So every time I see a post by some gold-ear who claims a particular tone characteristic based on plugging in a tube and listening, and comparing that to one pulled out 10 minutes before, I stop reading and assume he is trying to sell something.

[JoeM]

Yes, Miller effect is there but remember, the area and distance between the grid and anode that determines the capacitance also impacts the whole dynamic transfer function of the tube so if the Miller Capacitance is much different, the tube also has very different gain and curve family than a 12AX7. If there is enough difference in capacitance, even that impact on roll off( depending on circuit topography, it might be in the Mhz) will be swamped by the larger gain difference between that tube and a 12AX7. Having a lower bandwidth also makes it something else other than a 12AX7...

So how close to true to spec 12AX7s are new production tubes from China or Russia?

[km6xz]

Based on tests on tubes on hand here in my workshop( a bedroom in my city center apartment) and test instruments present, including a curve tracer, of working tubes about +20/-40% in gm, some falling into the 12AT7 and 12AU7 range of 60 and 20 gain respectively, when using typical plate loads and anode voltage found in guitar amps. The curve tracer I have here is on the left side of the bench. It is a modified Fairchild 6200B that can sweep anode voltages out to 1000volts and loads to 150 watts for a single tube.
This gear was mostly in my storage unit in California and slowly over a number of years, brought it into Russia in my suitcases on each trip back home. I have two similar benches set up in my apartment for work from DC to 2.0Ghz

When someone brings me tubes or I find some to test, I log the results to get a survey of what is being sold as what. With such a wide range of parameters, designing to the tube manual standard will be a frustrating experience. The tubes intended to be used should be tested to figure out what you really have and forget the tube manual or the spec sheets published by the re-branders. It is very obvious they do not test the tubes in a relevant way. They might be testing for emissions only for all I can tell, they sure have not done characteristic curves on them and made them available.
Many are not lousy tubes, they just are not what is claimed. Any can be made into amplifiers that do a good job but the designer has to know what they actually spec as. Essentially the same sonic traits can be had from any of them if they are mated to a circuit that take their true parameters into account.