What do you think about this tube regulator to be used into a preamp? How about hum?
An externally hosted image should be here but it was not working when we last tested it.
Piero7,
While there may be other issues, and I'm not sure about hum, the 0B2 will be much quieter if fed with a current source, instead of the 30k resistor. Ditto for the plate load of the right error amp element.
Why not use a ECC83 for more gain instead of the pitiful gain of the ECC82? Much better feedback. If I'm correct in my thinking (and it is often faulty) the error amp setup will net you mu/2 gain. So, if a tube with a gain of 17 is used, your error amp will only have a gain of 8.5. The pass element is effectively a cathode follower so its gain will be less than one. The voltage divider used to establish DC on the error grid will further reduce gain in the error amp to the point where I'd wonder about DC stability.
For the all important AC feedback, you'll need a bigger capacitor than 4700pF for decent AC performance at low frequencies. With 150k-200k from the cap to ground, you'd need about a 1uF cap for -3dB below audio frequencies, no less than 0.1uF.
The regulator will be quieter still if you decouple the 0B2 from the left error amp tube with an RC filter stage, say 1 Meg and another 0.1uF film cap.
I'd probably ditch the series 47k to the right error amp tube, but maybe that won't have any deleterious effect. It'll take bigger brains than mine to figure that one out.
It would be a good idea to put a zener from the plate to cathode of the pass element to limit voltage across the tube before the circuit is warm and ready to conduct. Raw, no load B+ can really soar. A 200V zener can save a lot of headache and also permit a larger cap on the load side by virtue of starting the cap charging and minimizing inrush through the EL84.
I don't think the series 100R cathode resistor of the EL84 contributes anything positive apart from limiting current through the tube. Ditch it.
Separate the "sense" lead from the B+ current lead and attach to the circuit at the point where current is actually used, i.e., the plate load of a gain stage.
Noise control and regulator performance will very much depend on your circuit layout; where you connect grounds and so forth.
If this is to be used in a power amp, for input stages, it may work well enough. For lower level work, you might choose otherwise.
First tried this sort of circuit with an EL84 and single 12AX7 section (zener in the tail) in the late '80's. It still works perfectly today. Very reliable.
Shunt regulators sound better. You may consider one of those many designs. Look at Broskie's Tube Cad site for lots of great ideas.
Stuart
While there may be other issues, and I'm not sure about hum, the 0B2 will be much quieter if fed with a current source, instead of the 30k resistor. Ditto for the plate load of the right error amp element.
Why not use a ECC83 for more gain instead of the pitiful gain of the ECC82? Much better feedback. If I'm correct in my thinking (and it is often faulty) the error amp setup will net you mu/2 gain. So, if a tube with a gain of 17 is used, your error amp will only have a gain of 8.5. The pass element is effectively a cathode follower so its gain will be less than one. The voltage divider used to establish DC on the error grid will further reduce gain in the error amp to the point where I'd wonder about DC stability.
For the all important AC feedback, you'll need a bigger capacitor than 4700pF for decent AC performance at low frequencies. With 150k-200k from the cap to ground, you'd need about a 1uF cap for -3dB below audio frequencies, no less than 0.1uF.
The regulator will be quieter still if you decouple the 0B2 from the left error amp tube with an RC filter stage, say 1 Meg and another 0.1uF film cap.
I'd probably ditch the series 47k to the right error amp tube, but maybe that won't have any deleterious effect. It'll take bigger brains than mine to figure that one out.
It would be a good idea to put a zener from the plate to cathode of the pass element to limit voltage across the tube before the circuit is warm and ready to conduct. Raw, no load B+ can really soar. A 200V zener can save a lot of headache and also permit a larger cap on the load side by virtue of starting the cap charging and minimizing inrush through the EL84.
I don't think the series 100R cathode resistor of the EL84 contributes anything positive apart from limiting current through the tube. Ditch it.
Separate the "sense" lead from the B+ current lead and attach to the circuit at the point where current is actually used, i.e., the plate load of a gain stage.
Noise control and regulator performance will very much depend on your circuit layout; where you connect grounds and so forth.
If this is to be used in a power amp, for input stages, it may work well enough. For lower level work, you might choose otherwise.
First tried this sort of circuit with an EL84 and single 12AX7 section (zener in the tail) in the late '80's. It still works perfectly today. Very reliable.
Shunt regulators sound better. You may consider one of those many designs. Look at Broskie's Tube Cad site for lots of great ideas.
Stuart
I think it's definitely on the right track, but this one will work much better.
An externally hosted image should be here but it was not working when we last tested it.
The pentode half of the 6KE8 has a rather large g(m), and gives the error amp a good deal of open loop gain. The voltage coming off the VR75 is easily filtered, since there is no current draw from it, and you can get that noise down (even though the noise from a gas discharge tube is a helluvalot less annoying than that from Zeners). There is also no possibility that you've made an oscillator, which is what you get when paralleling any gas discharge tube with a capacitor. It also holds its output voltage to about 0.8V with a 40V variation on the main supply rail. Since putting this into service as a screen regulator, the output voltage has not changed in the past two years.
If you need to beef it up, use something other than a trioded 6AQ5 as the series pass element.
I've also redesigned the thing to make use of those oddball 6X8's, which will serve just perfectly here as well. And they're cheaper than 6KE8s.
Thanks pals! 
And about this? DUe I have some ECC81 and 6080 could be a better idea than the one above?
And about this? DUe I have some ECC81 and 6080 could be a better idea than the one above?
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.