More SEPs

[Cressy Snr]

This one is different from the recent ones, so I thought I’d better start a new thread. There’s nothing new about it as such but I’ve moved back into Gary Pimm ‘Tabor’ territory, although it is not a push-pull design. Also it is very much a work in progress, so suggestions are welcomed.

Anyway this is what I have so far.

A veritable, Manhattan skyline of tubes.

872C3C4F-568E-4CC1-8278-AADB54390576.jpeg (96.2 KiB) Viewed 2623 times

[Cressy Snr]

This amp uses D3a as the input/driver:

684648A5-7764-40B4-9E40-D734F3336285.jpeg (55.3 KiB) Viewed 2622 times

[Cressy Snr]

As the audio circuit has become simpler, the power supply has become more complex, but that’s fine as the circuit sounds great. The D3a plate resistor is also the feedback resistor and the whole thing is a bit of a back-of-a-fag-packet job.

Ideally I would like to use 150V on the screens, but at the moment, there is not really enough HT off the plate of the output tube to enable that option, as it puts the driver plate volts below the screen volts

Also the gain is a bit low, as it would be by using the D3a plate resistor as the RFb. To get the gain up, I might have to go to separate plate and RFb, or use an active preamp, but the first priority I think, is to get the driver screen volts to 150V with the plate volts higher or the same.

[Nick]

Would it not be better to use a lower GM pentode as the driver? Say e180f. Would give you more options with anode resistor.

[Cressy Snr]

Thanks Nick,
I’ll look into that option.

[Paul Barker]

Great to see another project.

A lot easier when heaters seperate from cathode.

I’m thinking of a cooking version all heater valves.

[izzy wizzy]

Have you tried a low impedance device in the d3a cathode to get the gain higher? Diode, led, RC combo would all sound different.

[Cressy Snr]

Thanks for the input guys.

This isn’t about gain per se, but I wonder what would happen if I did this:

Screenshot 2023-03-12 at 13.00.33.png (33.31 KiB) Viewed 2503 times

Taking the feedback from the UL tap but keeping the fixed screen volts on the output stage. I haven't a clue without trying it.

[Nick]

IMHO, like the other circuit, all you are doing is applying feedback to the output valve in just the same way as UL, but unlike real UL, you are also making the driving valve have to work harder. What would be interesting is doing something like I did on the ref phono, but take the feedback from the secondary side of the transformer level shift it and use that to drive the g2 feedback. You may need to add some extra gain to compensate for the transformer step down.

[Cressy Snr]

Just tried it and it was a mess of oscillation, dependent on the position of the volume control. It was interesting as the further around the track of the pot I got, the higher went the frequency. Apparently I built myself a signal generator, er... not quite what I want . Fun nonetheless.
I'm ordering some E180Fs tomorrow. Cheap as chips.

[Nick]

Not sure what you tried, maybe my explanation was at fault. This is the sort of thing I had in mind. Works in spice.

[Cressy Snr]

Ah!

See what you mean there. I’d never have thought that. Thanks Nick.

At the moment I’m in the process of setting the amp up with a separate feedback resistor. With the first try, I have too much gain, which in addition to being bloody loud, also causes the D3a to become microphonic. However, on a positive note, I now have sufficient HT to get the driver screen to 150V whilst keeping the driver plate 27V higher. Just need to increase the feedback level to reduce the gain a tad. I’m therefore in a better position than I was yesterday.

[Cressy Snr]

Not sure what you tried…

I used the UL tap to feed back to the driver. Didn’t work.

[Nick]

It probably would, but you would need to invert the feedback.

What you could do was connect the b+ to the ul tap, then take the unused tap that used to be b+ and use that to feed g2. Would invert the feedback.

[Cressy Snr]

It probably would, but you would need to invert the feedback.

What you could do was connect the b+ to the ul tap, then take the unused tap that used to be b+ and use that to feed g2. Would invert the feedback.

Thick as pigsh*t that's me. Never mind.