Primary impedance for 4x6L6?

[bob p]

just to clarify things, the generic "output characteristic" that i posted is the characteristic of the tube itself; its not necessarily the same characteristic that the tube would display when operating in any particular topology.

How much 2nd order is there going to be in a push-pull amp, regardless of Z?

PP << SE

by using PP instead of SE topology, the 2nd and 4th harmonics are eliminated. i think that this is what you were implying, right?

the whole point of PP amplification is to permit a shift of the location of the operating point from the point of minimum distortion (a) to the point of maximum power output (b). in doing this, there is still some compromise with respect to increasing 3rd harmonic and THD. that's significant in a HiFi application.

when the 2nd and 4th are eliminated by using a PP topology the amp sounds quite different. this explains why SE and PP amps sound different when they distort. i think it also explains why some HiFi types like the sound of SE Triodes. imho its because they've developed a liking for the warmth of 2nd order harmonic distortion, whether they realize it or not. <ducking for cover in case @Rob reads this!>

[Satamax]

Well 2nd and 4th order are canceled provioding both tubes produce the same amount on each side!

[R.G.]

That gets back to one of my personal contentions about tubes, matching and biasing. Tubes can be matched for several characteristics. DC static current is mandatory; but this can be done with independent DC bias controls. AC gain per side will cause gross mismatch per side and a form of even harmonic distortion. This can be tuned out with a per-tube volume control if you want to take the trouble. The individual tube quirks, nonlinearities in the middle of the AC signal wiggles may or may not be matched in otherwise matched tubes, and there is no good way to tune them out.

Second harmonic distortion is only cancelled in the Class A region. When one tube is turned off, it can't very well cancel whatever is going on in the active tube, can it? There can't be an equal-and-opposite cancellation if one tube is on and one off.

[Bruce / Mission Amps]

PP << SE

by using PP instead of SE topology, the 2nd and 4th harmonics are eliminated. i think that this is what you were implying, right?
>

BINGO!

I actually enjoy the different tone higher powered SE outputs have, but the power consumption and lower power levels make it a more expensive ride.

I really wanted to build a higher powered SE class A amp using a single 4X150-A power tube for a long time and even started collecting tubes, all the parts, PT and sockets for one.

http://www.mif.pg.gda.pl/homepages/f...049/7/7034.pdf

The OT was the sticker....
The SE Class A rating of one of thise tiny littlle 4X150-A tubes (half the size of a big bottle 6L6GC) is an amazing 250 watts each.
With only 1500vdc on the plate (from a bigass ass, high dollar, class A OT of 75-100 watts), ... an amp like this could easily make 75 watts in SE class A with forced air around the tube.
And it would only take a couple 12AX7s with good tone controls in circuit to drive it to full power.

[MWJB]

I'd be just as concerned about the 23-24mA/tube at 510v. Say you lose 40-50v at the cathode that's 460-470v effective plate volts x 0.024 = 11W plate dissipation?

[R.G.]

I actually enjoy the different tone higher powered SE outputs have, but the power consumption and lower power levels make it a more expensive ride.

Oh, yeah. My favorite amp of all time is a 2W SE with a 6AQ5 power tube. I pulled two of the OTs out of a Magnavox console stereo setup I got for $10. They refunded $5 when I just pulled out the electronics chassis and left them the console.

I set one OT and the PT up with two 12AX7s in a new chassis. This thing turns out everything from subtle crunch to mindbending grind. The output is used into a speaker for practice and into a speaker emulator circuit and a PA if I want more volume.

The sound of an SE is quite a different animal. It's the solution you pick to the power and practicality problems that make the difference.

I really wanted to build a higher powered SE class A amp using a single 4X150-A power tube for a long time and even started collecting tubes, all the parts, PT and sockets for one.
...
the OT was the sticker....

A SE output transformer is probably ten times the mass of an equivalent non-SE. You'd need a 2.5kVA core for that OT, and at that size the leakage inductance would be massive, so your highs would be shot before you started. Not to mention it weighing maybe 200 pounds.

[Hallingamp]

Thank you for all the replies!

I'm afraid a lot of this is way over my head, but here goes:

My thoughts were exactly like Rob Mercure's:
6,6k : 2 = 3,3k. 4k = Close enough.

But I see there are other thoughts and considerations here. Maybe the smartest thing I could do is to just drop two of the tubes? But then maybe I should buy the 6,6k OT?


And yes, the biasing may sound a bit cold. I just looked at some bias chart suggestions for 6L6's at 510V, forgetting the plate to cathode voltage would drop a fair bit with cathode bias.

Jan

[bob p]

Oh, yeah. My favorite amp of all time is a 2W SE with a 6AQ5 power tube.

interesting. i've just inherited a bunch of boxed 6AQ5 (among other things) in an old TV serviceman's tube caddy. i've been thinking about using them to build some low powered guitar amps, though i have to admit i hadn't been thinking SE.

by any chance would you happen to have a sketch of that circuit?

[R.G.]

The same thing happens with beam power tubes - like the 6L6. It also happens for triodes and triode-mode pentodes, but the power max and distortion min are in different places.

For music production, a little more second harmonic is great.

I designed a circuit with JFETs and MOSFETs that did pure second harmonic. I took a FET diffamp and fed it out-of-phase inputs, then tied the drains together. Matched FETs will generate equal and opposite signals at the drains, but since both JFETs and MOSFETs are square-law devices, tying the drains together cancels the amplified signal but reinforces the second-order distortion products.

When fed a sine wave, it produced a sine of twice the frequency, substantially pure.

It's interesting that the distortion products we worry about in power tube stages are second and third, given that tubes are a 3/2 power law device.

[Steve Conner]

So, R.G., did your second harmonic generator do anything musically interesting? Also, could you generate the third, fourth, etc, in whatever proportions you want by chaining a whole bunch of them? That would be a neat project, a Fourier distortion pedal with a drive knob for each harmonic order.

BTW, I don't believe it's true that even harmonics only cancel when the tubes in a PP output stage are in their Class-A region. My argument is that if the bottom half-cycle of a wave is a mirror image of the top one, then you can say by Fourier analysis that the waveform contains no even harmonics. Therefore, a PP output could cancel even harmonics perfectly, even if it were biased at zero standing current. Just as long as the second tube does the same thing as the first one did a half-cycle before.

[R.G.]

So, R.G., did your second harmonic generator do anything musically interesting? Also, could you generate the third, fourth, etc, in whatever proportions you want by chaining a whole bunch of them? That would be a neat project, a Fourier distortion pedal with a drive knob for each harmonic order.

It did. It made the lowest distortion octave up I've ever found. The full wave rectifier octave up devices produce too much distortion for me. This was a much subtler device. Unfortunately, since it relies on the square-law distortion of a FET, it can only generate powers of two - 2nd, 4th, 8th, etc. if chained. We'd need cube-law, fifth-law and so on devices to do the fourier distortion this way. Also, on a complex waveform it produces the square of the waveform, which includes second order intermod products. These are much less than in a rectifier based octave, but they're still there.

BTW, I don't believe it's true that even harmonics only cancel when the tubes in a PP output stage are in their Class-A region. My argument is that if the bottom half-cycle of a wave is a mirror image of the top one, then you can say by Fourier analysis that the waveform contains no even harmonics. Therefore, a PP output could cancel even harmonics perfectly, even if it were biased at zero standing current. Just as long as the second tube does the same thing as the first one did a half-cycle before.

That would be correct if that last statement applied. That is, if both upper and lower-half tubes were perfectly matched, then the waveform would be symmetrical, and there would be no even-order products in the result.

But that's not a result of the PP setup then, it's a product of the two tubes being matched. With tubes which do not have the same AC characteristics -that is, they have non-matching wiggles in their linear range - there is no addition mechanism to match them, even partially.