Tubes used in SE output circuits

[Paul1453]

I just noticed something I wasn't expecting. 

I set up a simple 12AX7 6V6 SE output circuit and it worked fine.  I have 5 6V6 tubes and my tester tells me 2 are bad.  It gives me values of 66, 52, 50, 30, and 23.  I tried all of these tubes in my circuit and I'll be damned if I could hear much if any difference between any of the tubes, even the 23 and 30 functioned and sounded fine to me.  The only difference I noted was that some of the tubes drew 1-2 mA more current from my PS, other than that 23 seemed to work just as well as 66.

Now on a PP circuit I wouldn't want to pair my 23 with my 66, and maybe some of you are pulling those 23's from PP circuits and chucking those tubes.  From my small sampling, maybe those low testing tubes can still be used in SE circuits?   

[shooter]

drew 1-2 mA more current

That is the # I shoot for when *matching* tubes.  I don't think I've put a tube in my tester since I got to this site.  most of my builds are PSE so I just *match* up currents and call it good - given they sound good, and scope good. 

[HotBluePlates]

... I have 5 6V6 tubes and my tester tells me 2 are bad.  It gives me values of 66, 52, 50, 30, and 23.  I tried all of these tubes in my circuit and I'll be damned if I could hear much if any difference between any of the tubes, even the 23 and 30 functioned and sounded fine to me. 

... maybe those low testing tubes can still be used in SE circuits?   

Transconductance (Gm) tests aren't diagnostic for output tubes; if you look in the Radiotron Designer's Handbook (available in the Library of Information, linked at the bottom of this screen), a power output test is recommended.  In this test, you use the tube as you have, apply a signal and measure output power. You could either measure output power for a fixed applied signal, or apply a signal which drives the tube to the onset of clipping and note the drive signal needed.

Idle Gm really only matches tubes at idle. It doesn't guarantee the tubes are matched at all times. And as you found out, it doesn't really show whether the tube will perform in every amp circuit.

So what about low-Gm tubes? I would expect a low-Gm tube to need more bias voltage to keep plate current at a specified level. That's because Gm is a measure of how effective the grid voltage is at controlling plate current. Say the tube is capable of passing just as much current as any point in its life; since the plate load impedance and power supply capabilities dictate the output power under this circumstance, the tube is capable of performing in the amp and making just as-loud a speaker output.

But the bigger bias voltage implies a bigger drive signal is required to push that low-Gm tube to the same plate current swings. So Same-Power-Output / More-Drive-Signal = Lower Power Sensitivity. In other words, an extreme case will result in your amp seeming weak ("I swear it used to get louder at '7' on the Volume control..."). However, as you're seeing, you may be perfectly happy with how the tube performs.

I didn't even get into the issues of whether the tube tester provides a fair test of the tube, has correct setup conditions vs specified reading (even in the old days, tester manufacturers sometimes got this wrong), how much your tester varies from every other tester out there, etc. You really have to get a feel for your individual tester, the readings it provides, and how that translates to reality in your experience.

[Paul1453]

an extreme case will result in your amp seeming weak ("I swear it used to get louder at '7' on the Volume control..."). However, as you're seeing, you may be perfectly happy with how the tube performs.

This is what I was expecting, but I really couldn't notice any difference in this low power output circuit.  Thanks for clarifying this issue and letting me know my thought process was correct.

[HotBluePlates]

I remember reading Aspen Pittman's Tube Amp Book and thinking I had to replace output tubes every 6 months. In the couple-decades since, I have some output tubes I've used in the same amp for 15 years or more...

[Fresh_Start]

If the single-ended amp uses cathode bias for the power tube, won't that be more forgiving of differences between 6V6s? 

Since the tube with "23" from the tester measures 2 ma lower in current, the voltage drop acros the cathode resistor must be lower also. IOW the cathode biased output stage is "self biasing" to a point.  Now 2 ma doesn't seem like much of a difference to me, but I'm guessing that the difference in perceived output might be greater in a push pull amp. However, only if both tubes read "23".

HBP's amps are biased correctly. My Blues Jr. OTOH ate EL84s until I added bias adjustment and raised the bias voltage to cool 'em down.

Chip

[HotBluePlates]

If the single-ended amp uses cathode bias for the power tube, won't that be more forgiving of differences between 6V6s?

What do you get for perfectly matched output devices? Maximum clean power before the onset of distortion. So mismatched output tubes will cause the amp to distort a little sooner.

Again referring to RDH4, there is a discussion of the importance of matching power output triodes. An experiment was run in which two dissimilar output tubes were used in a push-pull output stage. One of the tube had a Gm half that of the other tube, intended to simulate a push-pull pair where one tube was new and the other near the end of its life. When operated in class A, the two dissimilar tubes resulting in just 5% THD at some power output level. The editor noted the effect is a bit worse for class AB, and that pentodes fair a bit worse than triodes for both cases. Still that amount to matching yielding maximum clean output power.

Now grossly-imbalanced idle or operating currents may exceed the amount of unbalanced d.c. the OT can handle, resulting in OT saturation. That can restrict bandwidth through the OT; it may also disrupt the ability of the OT to reject power supply hum, which then seems to increase at the speaker. Allowable unbalanced d.c. (before saturation) for typical push-pull OT's seems to be around 5-10mA, so you'll want to keep that in mind