Hi George -
I'm thinking this topology should work well with some of the (relatively) modern transmitting tubes. 6146 and 4D32 come to mind. They aren't great in most conventional circuits I've tried. And like sweep tubes both have screen voltage ratings around half the plate voltage rating.
6146 can do 35 watts plate dissipation (ICAS). 4D32 is listed as 49 watts in AB1, but looking at the massive plate structure I have to believe it can do more.
I have both tubes sitting around, and will do some experiments. Maybe 500V/250V to start.
Pete
I'm thinking this topology should work well with some of the (relatively) modern transmitting tubes. 6146 and 4D32 come to mind. They aren't great in most conventional circuits I've tried. And like sweep tubes both have screen voltage ratings around half the plate voltage rating.
6146 can do 35 watts plate dissipation (ICAS). 4D32 is listed as 49 watts in AB1, but looking at the massive plate structure I have to believe it can do more.
I have both tubes sitting around, and will do some experiments. Maybe 500V/250V to start.
Pete
I haven't tried the 4D32 in UNSET yet, and tried a 6146 in anything recently.
I did stuff a pair of 4D32's in my Universal Driver push pull test amp. They made 200 watts from a pair in AB2 at the onset of clipping from 650 volts into 3300 ohms. Screen voltage was 250 volts. I tried them into 2500 ohms at the same voltage where they made 250 watts. Nothing was glowing, but the efficiency had dropped from the 3300 ohm load, so I'm guessing 200 to 250 watts per pair is the limit. I can get the same power from the 36LW6's which I have more of, so that's where I'm experimenting now.
Attachments
I finally got the oscillation on the negative peak eliminated.
I managed to get 15W at 10% THD out of a 6P41S.
15.125 W at 10% THD 3.04Vrms drive
4W at 1.8 % THD 1.36Vrms drive 3rd 10dB below 2nd
1W at 0.88% THD 0.64V drive
The schematic is not 100% accurate, I am using pots to adjust the bias of the two tubes.
Listening tests will be next.
I managed to get 15W at 10% THD out of a 6P41S.
15.125 W at 10% THD 3.04Vrms drive
4W at 1.8 % THD 1.36Vrms drive 3rd 10dB below 2nd
1W at 0.88% THD 0.64V drive
The schematic is not 100% accurate, I am using pots to adjust the bias of the two tubes.
Listening tests will be next.
Attachments
I put a very basic PP together. ECC81 loaded with gyrator (no unset here, just a led in the cathode), mosfet based concertina and the 250V P fets to feed the cathodes of 25DQ6Bs. For the moment the grids are at a fixed potential, but I want to use these for local feedback later on.
I used the zener in the P fet, just like shown here by George https://www.diyaudio.com/forums/tub...sing-mosfet-ultralinear-feed.html#post6267790 , but I could not get any current going through the circuit. I drew the P fets, and then it became clear that the gate is at a lower potential than the source, so with the zeners as drawn there can never be more than 0.7V between source and gate (and more is needed to turn them on). It saved the Fets, but I had to reverse the zeners to get current flowing.
Now going for the fine tuning!
I used the zener in the P fet, just like shown here by George https://www.diyaudio.com/forums/tub...sing-mosfet-ultralinear-feed.html#post6267790 , but I could not get any current going through the circuit. I drew the P fets, and then it became clear that the gate is at a lower potential than the source, so with the zeners as drawn there can never be more than 0.7V between source and gate (and more is needed to turn them on). It saved the Fets, but I had to reverse the zeners to get current flowing.
Now going for the fine tuning!
Attachments
Last edited:
Could someone who is LTSpice-savvy please post the schematic for George’s simulation in his post #48. I assume that is his latest UNSET schematic that lead to Zintolo’s GU50 variation and TheGimp’s 6P41S.
I’m very eager to build a UNSET 36LW6; hopefully George will be able to offer a Tubelab PCB soon.
I’m very eager to build a UNSET 36LW6; hopefully George will be able to offer a Tubelab PCB soon.
Look familiar. I believe you have the edcore outputs I had for this project. Hopefully get to see them I use.
Attachments
Beta Board (for "hobby planning purposes")
(a) this is a great phrase,and I have the same question.
(b) Whilst I recognize the artistic merit of Pushing the Limits and Blowing Things Up*, I would be ecstatic with a UNSET that made an honest 10W at low distortion; i.e. a replacement for the TSE 300B (even if it is loaded with SAND. The PowerDrive has made me a Believer.
(c) A Reference Point: if y'all really are getting clean outputs at ~50% of dissipation, this circuit could comfortably get 10W out from select Sylvania 6AV5s (REM those were rebadged as 6B4GAs, and sound pretty good); I am running those right now in uLinear at 400vdc B+ and >22W dissipation with no Embarrassing Redness. (I do know that 6AV5s are difficult to get "overseas", but we Yanks are still up to our gills in them, and it's my Horizontal Sweep Tube of Reference.)
(d) I think this circuit is so promising that TubeLab George could get enough solid commitments for a Beta version of the board, with the advantage that lots of folks would then be Making Mistakes Quickly.
I would be in for AT LEAST a couple. Heck, I still have unbuilt SSE boards with full parts complements, so no worries...
*Blowing Things Up is of course why we all gather 'round TubeLab, like teenagers staring into an engine compartment...
(a) this is a great phrase,and I have the same question.
(b) Whilst I recognize the artistic merit of Pushing the Limits and Blowing Things Up*, I would be ecstatic with a UNSET that made an honest 10W at low distortion; i.e. a replacement for the TSE 300B (even if it is loaded with SAND. The PowerDrive has made me a Believer.
(c) A Reference Point: if y'all really are getting clean outputs at ~50% of dissipation, this circuit could comfortably get 10W out from select Sylvania 6AV5s (REM those were rebadged as 6B4GAs, and sound pretty good); I am running those right now in uLinear at 400vdc B+ and >22W dissipation with no Embarrassing Redness. (I do know that 6AV5s are difficult to get "overseas", but we Yanks are still up to our gills in them, and it's my Horizontal Sweep Tube of Reference.)
(d) I think this circuit is so promising that TubeLab George could get enough solid commitments for a Beta version of the board, with the advantage that lots of folks would then be Making Mistakes Quickly.
I would be in for AT LEAST a couple. Heck, I still have unbuilt SSE boards with full parts complements, so no worries...
*Blowing Things Up is of course why we all gather 'round TubeLab, like teenagers staring into an engine compartment...
In Europe we have plenty of topcap sweeps: EL36/PL36, PL500/504 (*GB5, not unknown to George) and others.
I also got a bunch of 6CU6 (topcap 6AV5).
And don't forget the Soviet sweeps.
Yes, I'm interested in UNSET
Francios,
LTSpice is easy to use, and there is lots of information (and models) in the spice model stickie.
Vacuum Tube SPICE Models
LTSpice is easy to use, and there is lots of information (and models) in the spice model stickie.
Vacuum Tube SPICE Models
Hi, I've one question: I would like to adapt a cheap chinese SE kit to be used in PSE with EL34 and this kind of design (it is in configuration 12ax7 + octal). Can I reference the grid leak resistor of the output tube to a negative rail instead of ground, and reference the driver Mosfet to ground targeting the output tube cathode to go as close to gnd as possible during peaks? This also would help to keep the grid dc reference stable in case the lnfb plate to grid will be increased to more than 10%.
Thanks
Thanks