Pentode for a cathodyne? Working start

[Michael Allen]

I put together an amp with spare parts. All octal, parallel 6SC7 into 6SH7 driving a 6SH7 as a cathodyne into 2x6L6s.

It works but not as well as I think it can. Right now the 6l6 plates are around 450V, screens at 399, preamp nodes at around 320V. Fixed bias with the idle at 25ma apiece.

I put a 1khz signal into, i'm only getting 14W clean into an 8ohm load. Actually I'm not sure how many clean watts I'm getting. If i adjust the volume the signal coming out of the second stage as pictured below is as big as it can be before distorting. So it's not really clean at the OT since the cathodyne is buggin, but as clean as the amp should be before the preamp distorts. I measure 7.6VAC at the output jack. So (7.6*1.4)^2 / 8 = about 14W. The signal coming out of my second gain stage (6SH7 pentode, 130k Rp, 220k screen, 0.1u screen bypass to cathode, 390 ohm cathode bypassed with 2.2uF) is looking normal and fine (the signal gen has an asymmetrical shape.)


The cathodyne has 100k load resistors, 3.5k cathode resistor, 1M grid stopper, 470k screen unbypassed, 470k grid leak. Cathode signal comes at the junction of the cathode resistor and the load resistor/grid leak resistor. On the oscilloscope, here I'm getting what Merlin called "frequency doubling" on the anode from grid current messing up the cathode voltage. This is happening even though i used the suggested 1M grid stopper.

All scales are 5V/div.

Cathode output:


Anode output:


Combined:


The signal coming out of the cathodyne is much smaller than the signal going in. It should be 0.9 to 1 gain right? Why is the signal so much smaller?

This small signal then doesn't really get amplified by the power tubes since I think it's not enough voltage from the concertina to drive the power tubes. Any ideas of how to get the output from the cathodyne? Has anyone done it with a pentode before?

OT:

[tubeswell]

Have you tried it with the pentode in triode mode (to see how much the screen could be affecting the symmetry and/or output)?

[Merlinb]

I don't think this problem can be solved by description alone. A schematic is needed.
Nevertheless, as a quick suggestion, try adding a 100nF (or thereabouts) capacitor between screen grid and cathode of the cathodyne.

[Michael Allen]

Merlin, I'll try and add that cap and see what happens. I did think about running it in triode mode but forgot to try.

Thanks guys,

Here's the schematic

[soundmasterg]

You can get more signal swing from the cathodyne if you increase the supply voltage for it too....

Greg

[loudthud]

I looked up the 6SH7 in an old RCA tube manual and it says it's not recommended for high gain audio because excessive hum may be encountered. Besides that, it would appear that it wants about 1K for a cathode resistor although screen voltage can change that. What are you seeing as supply, plate and cathode voltages in the concertina? Instead of fooling with self bias, you might want to consider a voltage divider to set the grid voltage like used in the famous Paul C mod. Set the grid voltage at about 20% of B+.

[Michael Allen]

I think that may be due to the construction of the 6SH7 tube. I think the heater is too close to the grid or something so a lot of noise gets introduced. I'm using DC on the preamp and noise is not really bad at all. Plus this is a low gain amp so I don't think it'll be an issue.

I'll look into that paul C mod. I've been too busy to try the screen cap on the concertina, but i measured some voltages for reference. Here's the updated schem with voltages.

[Michael Allen]

Okay, that cap seems to have helped immensely. Now looking at the concertina output section before it distorts i measure about 8.8V or 18.6W at the dummy load. This is the output of the concertina, all scale are now 2V scale.



This is the input and one phase of output at the cathode. Does this look about right for a gain loss?




This is the concertina output and the output at the dummy load. Looks like I'm losing a lot of signal somewhere in the output section. All my voltages look right?

[octal]

Re: "losing a lot of signal in the output section." That's normal! An output transformer is a step down transformer. You're losing voltage and gaining current, with a net increase in power.

Nathan

EDIT: That was poorly phrased and makes it sound like there's power gain in the OT. I mean that the complete power amp section results in a voltage loss but current and power gains. (Comparing the input to the output.)

[Michael Allen]

oh geeze, of course. thank you!

[jrfrond]

I don't think you can really optimize a pentode as a cathodyne PI. Good symmetry might be impossible with the control and suppressor grids active, because they are unidirectional control for the plate, not the cathode. This is why tubes like the 7199 and other triode/pentode tubes were used as a 2-stage cathodyne with the pentode gain-only stage preceding the unity-gain phase inversion.

[Michael Allen]

Hmm, so if it's in triode mode with the supressor grid connected to the cathode are you saying that the supressor and screen will still have an effect on the symmetry? My signal generator is not symmetrical so it's hard to tell but it seems to look fine as of right now. I'll try it in triode mode and see if it gets any better, along with increasing the supply voltage.

I'm biased at about -45v or so, which means I need 90V swing to overdrive the 6L6 right? I read that the swing supply should be around 75% of the voltage drops on the concertina, which means voltage drops of around 120V. This would mean 120V across the anode resistor, 240V across the tube, and another120V across the cathode resistor totaling 480V. Well I can only get around 400V supply for the PI so I guess I will get some PI distortion before the output section breaks up, but it shouldn't be too horrible I'm guessing.

[jrfrond]

Hmm, so if it's in triode mode with the supressor grid connected to the cathode are you saying that the supressor and screen will still have an effect on the symmetry?

I have to say that I don't know. My guess is that it SHOULD work in triode mode because the suppressor becomes electronically transparent, but then you can just use a triode.

[Merlinb]

Pentodes work fine as cathodynes (in fact, they will give you more output voltage than a triode), after all, they're not much different from transistors. The only thing you have to watch for is that the screen impedance is effectively in parallel with the cathode load, since the two are coupled via the screen bypass cap, but in this case it doesn't seem to cause any serious problems.

[Steve Conner]

the screen impedance is effectively in parallel with the cathode load, since the two are coupled via the screen bypass cap

That's probably your problem: You don't have the screen bypassed to the cathode.