An idea for a partial feedback pentode amp using 6AU6's and 807's

[Shoog]

Hi there,
I have an idea for a PP amp. I intend using microphone transformers as a front end step down and phase splitting stage. This would then feed a pentode configured 6AU6 (with regulated screen) LTP with CCS in the tail. This would then feed a pair of 807's in pentode mode with partial feedback to the grids.
This idea is based on the pentode amps which Gary Pimm developed.
Even with the front end step down of 9:1, the 6AU6 should have well enough gain to drive the 807's.
My question is what is a good operating point for the 6AU6, I have looked at a few datasheets and end up with a headache. I want a bias point of at least -3V , with a +B of between 250V and 350V.

Any experience greatly appreciated.

Shoog

[Tubes4e4]

Hi Shoog,

there are plenty of 6AU6 datasheets containing tons of tables of suggested operation points for audio duty. The manufacturers certainly knew why they suggested them. Just pick one that suits your gain, voltage swing and Zout demands.

Maybe you also could have a look at EF94 datasheets, which is the same tube, but European designation. When not developing for "special" situations, I just stick to the Lorenz EF94 table for RC-coupled gain stages.

As a more linear and capable alternative, EF80/6BX6 comes to mind, which is cheap and plenty (at least in Europe). I got stunning results and capabilities (extensively published elsewhere) using just a certain op point recommendation from the spec sheet.

Tom Schlangen

[Brett]

Pity Gary Pimm took his site down as there was a very similar schem based on his earlier PP47 (which I built and it was flippin luvverly). If I can remember what it was called and where I saved the schem, I'll post it.

Found it.

[Bandersnatch]

hey-Hey!!!,
If you go regulated voltage for the g2's of a LTP, make sure you ref it to the cathodes. LTP in both pentode and cascode with the upper g1, or g2 (in the case of the pentodes) didn't work anywhere near as well with the ref to ground.
cheers,
Douglas

[Shoog]

I had been looking for the Gary Pimm schematic and though I had it, but I didn't. Thanks for that.

Useful info on referencing to cathode rather than ground.

I will work up a rough of my idea in the next couple of days and leave it open to comment/flamming.

Thanks

Shoog

[AndreasS]

Originally posted by Shoog
Hi there,
I have an idea for a PP amp....feed a pentode configured 6AU6 (with regulated screen) LTP with CCS in the tail. ...Shoog

Hi Shoog,

why do you need a CCS in the LTP, for the phase inversion you have the input transformer?

Regards Andreas

[ilimzn]

Originally posted by Bandersnatch
hey-Hey!!!,
If you go regulated voltage for the g2's of a LTP, make sure you ref it to the cathodes. LTP in both pentode and cascode with the upper g1, or g2 (in the case of the pentodes) didn't work anywhere near as well with the ref to ground.
cheers,
Douglas

Very useful tip - but sometimes somewhat difficult to do, as the common cathodes go up and down in unison with the input signal. Since a regulator is a feedback device, you may end up injecting distortion through G2.

OTOH, I have seen differential pentode LTPs in tube scopes and normally they don't bother with referencing the G2 supply to cathode, because the input signal is actually very small, so the change in Vg2-k is also very small. It is a form of feedback but since mu g1-g2 is so much smaller than effective pentode mu wrt plate, it is acceptable. Although input voltage swing is higher in audio amps compared to scopes (by 1-2 orders of magnitude!), it may still be fairly small (considerably less than 1%) compared to Vg2. Assuming the pentodes are not working to the extremes of their voltage swing, we can approximate that Ig2 of both pentodes together is fairly constant and will not upset the differential action of the LTP, adjusting the tail CCS higher to compensate for Ig2 of both tubes. If botha assumptions are correct, then investing in G2 regulation referenced to the common LTP cathodes may not provide a signifficant benefit compared to the cost.

There is a third soultuin, which is, however, only usable if three conditions are satisfied:
1) The amplifier uses single ended input
2) Fairly high NFB factors are used
3) The feedback is arranged for an inverting configuration, i.e. the input of the amp is the (-) input of the LTP and the (+) input is grounded, so the (-) input of the LTP is a virtual ground. This often implies lower input impedances.
Under these conditions, because the input of the LTP behaves as a virtual ground, the voltage swing on it is small, and so is, by consequence, the swing on the common cathode connection of the LTP. By extension, the variation in Vg2-K is also small compared to Vg2, the latter remains referenced to ground.

[Bandersnatch]

Originally posted by ilimzn


Very useful tip - but sometimes somewhat difficult to do, as the common cathodes go up and down in unison with the input signal. Since a regulator is a feedback device, you may end up injecting distortion through G2.


There is a third soultuin, which is, however, only usable if three conditions are satisfied:
1) The amplifier uses single ended input
2) Fairly high NFB factors are used
3) The feedback is arranged for an inverting configuration, i.e. the input of the amp is the (-) input of the LTP and the (+) input is grounded, so the (-) input of the LTP is a virtual ground. This often implies lower input impedances.
Under these conditions, because the input of the LTP behaves as a virtual ground, the voltage swing on it is small, and so is, by consequence, the swing on the common cathode connection of the LTP. By extension, the variation in Vg2-K is also small compared to Vg2, the latter remains referenced to ground.

Difficulty has Zero influence on importance.

Further, the seemingly small change offered by ref'ing g2 to cathode v. the ground makes important changes to the sonic signature IMO.
cheers,
Douglas

[Shoog]

why do you need a CCS in the LTP, for the phase inversion you have the input transformer?

Good question.
Mainly because the presence of the Input transformer allows you easily to introduce the CCS without the need for a negative supply. The bottom of the input transformer can be referenced to the junction of the two 6AU6 cathodes and the top of the CCS. Each cathode has its own cathode bias resistor and the CCS can be a simple LM317. In this way the LTP with both sides driven by the input, should help to correct any slight imbalance introduced by the input transformer.
It really only constitutes a few extra components and its very unlikely to degrade the performance over them not been there. A sort of belt and braces approach.

Shoog

[AndreasS]

Originally posted by Shoog
... In this way the LTP with both sides driven by the input, should help to correct any slight imbalance introduced by the input transformer. ...Shoog

Hi Shoog,

a very good idea, because often the capacitances to center tap are different (nonsymmetrical) in that transformers.

Regards Andreas