The 832A and 829B are pentodes which can be used individually and are usually run as a single tube push pull amp in RF (radio transmitter) applications. The screen grid is shared between the two halves, so triode wiring the individual sections is not possible.
Going further out in left field are the 6DY7 and 6DZ7. Here a pair of audio amplifier pentodes are stuffed into a single octal bottle. The 6DY7 is somewhat like a pair of 50C5 / 25L6 / 6W6 tubes while the 6DZ7 contains a pair of 6BQ5 / EL84's. Both carry reduced specs due to their rather cramped living quarters. Somewhere in a box I may still have a pair of each, but I haven't seen them in years. Again the screen grids are shared so triode strapping the individual sections is impossible.
I'm looking forward to seeing why Jan needs a dual triode for an SE output! What's he up to....
I was thinking that you could build a stereo SE 5W/ch with a single tube. Probably not; it's borderline I think..
Or one of those low voltage reg tubes, but you'd probably need a custom OT.
I have an idea to wrap feedback around the output stage with very little impact from xformer phase shift.
A single ended amp with extremely low distortion.
Jan
Or one of those low voltage reg tubes, but you'd probably need a custom OT.
I have an idea to wrap feedback around the output stage with very little impact from xformer phase shift.
A single ended amp with extremely low distortion.
Jan
Last edited:
Single output tube, maybe. You would still need something to generate enough drive voltage to get 5 watts out of a low Mu triode. Just from the data sheets I see that a 6DY7 dual pentode could generate 5 watts per channel at 9% THD from 25 V p-p of drive. A 6BN11 tube contains a pair of 6EW6 pentodes that have ample Gm for generating enough gain to drive those 6DY7's and apply enough plate to plate feedback to make the combo sound like a triode amp with two tubes on 250 volts. 6DY7's were not too common and may be rare in Europe. You might get similar performance from an 829B but more drive and B+ voltage would be needed.
Why does it have to be triode? You could do one 6DZ7 and a 12ax7 and have @5W and some GNFB stereo SE. The 6DZ7 isn't rare, just have to look for them. 5U4, 6DZ7, 12AX7... nicely compact, excellent amp.
There's the DET19 2C34, though that has less power than you specified. Two in parallel driven into A2 might work. I built a little 10w AB2 PP amp with them, it's myfavorite amp, sounds sweet. They look groovy too. http://www.r-type.org/exhib/aaa1369.htm
For a dual triode with cojones there's the 6528 - https://tube-data.com/sheets/115/6/6528.pdf each triode is rated at 30w. It too looks far out. I have three NOS itching to be used.
Andy.
For a dual triode with cojones there's the 6528 - https://tube-data.com/sheets/115/6/6528.pdf each triode is rated at 30w. It too looks far out. I have three NOS itching to be used.
Andy.
Last edited:
It doesn't have to be a triode. And drive level is no issue with my idea.
That 6DY7 has a common cathode and I need two completely separate tubes (common heater is OK).
Right now I have it working in sim with an EL34 per channel but would like to have the two tubes in a single envelope, just for kicks.
Jan
That 6DY7 has a common cathode and I need two completely separate tubes (common heater is OK).
Right now I have it working in sim with an EL34 per channel but would like to have the two tubes in a single envelope, just for kicks.
Jan
Just for kicks is easy with a common cathode. One resistor and one bypass cap. The bypass cap gives the tubes max gain, too. Some amps tie 4 cathodes together.
Maybe he is using my UNSET design or something similar. Here the drive signal is fed to the cathode via a P-ch fet follower and feedback goes to the control grid via a resistive divider from plate to ground. Common cathodes don't work here.
How about a Hall of Mirrors amplifier? Put right angle mirror sets all over the top of the amp, so that when you plug in one tube, tubes appear everywhere. Then you can put in -two- tube/ sockets, and no one can tell how many tubes are actually there. And then design it with two more tubes paralleled below deck, so that pulling out either of the top tubes does not stop the amplifier working.
Last edited:
6336 and 6528 make great single-ended amps. About 6 watts into a 2.5K load. Run them at higher plate voltage and lower current, like 350VDC and 60mA. Seems counterintuitive looking at the curves but it works very well. The 6528 is easy to drive. The 6336 needs as much voltage drive as an 845. ;-)
I did not know about the old Music Man amps until after I came up with UNSET and I still have never seen one in the flesh. I did repair a lot of guitar amps in the late 70's and through most of the 80's.
There are some differences between UNSET and Music Man. UNSET uses a P-ch mosfet as a follower under the cathode, so the cathode sees a near perfect copy of the applied drive voltage. Music Man uses an NPN BJT to in common emitter configuration to drive the cathode, so the cathode sees a drive current dependent on the Beta of the transistor and the current feeding its base. Music Man has the grid bypassed to ground for AC. The control gird of the tube in UNSET sees a scaled down copy of the plate voltage on its control grid. As Schade revealed a long time ago this provides a triode like response from the output pentode. This configuration eliminates the capacitor in the feedback path seen in the original Schade scheme. The ratio of the two resistors in the feedback path controls the DC bias point and the slope of the triode like curves generated by the mosfet / pentode pair.
In a SE power amp with a big sweep tube on 500+ volts clipping is seen in the 20 to 30 watts of audio power area depending on the size of the tube and how far you want to bend the specs. In this configuration THD is usually in the 0.3 to 0.5% range at 1 watt and under 1% at 10 watts. The THD is also dependant on the feedback ratio. These numbers were with a 10% feedback ratio with a 30 K grid to ground resistor and a 300 K grid to plate resistor. The screen grid gets between 150 and 175 volts bypassed to ground. Since the cathode is driven there is some drive applied between G2 and the cathode.
I have seen some outstanding THD levels in a driver application while also using a CCS load or bootstrapped plate load resistor. Several different tubes were capable of under 0.2% THD at 50 volts output.
There are some differences between UNSET and Music Man. UNSET uses a P-ch mosfet as a follower under the cathode, so the cathode sees a near perfect copy of the applied drive voltage. Music Man uses an NPN BJT to in common emitter configuration to drive the cathode, so the cathode sees a drive current dependent on the Beta of the transistor and the current feeding its base. Music Man has the grid bypassed to ground for AC. The control gird of the tube in UNSET sees a scaled down copy of the plate voltage on its control grid. As Schade revealed a long time ago this provides a triode like response from the output pentode. This configuration eliminates the capacitor in the feedback path seen in the original Schade scheme. The ratio of the two resistors in the feedback path controls the DC bias point and the slope of the triode like curves generated by the mosfet / pentode pair.
In a SE power amp with a big sweep tube on 500+ volts clipping is seen in the 20 to 30 watts of audio power area depending on the size of the tube and how far you want to bend the specs. In this configuration THD is usually in the 0.3 to 0.5% range at 1 watt and under 1% at 10 watts. The THD is also dependant on the feedback ratio. These numbers were with a 10% feedback ratio with a 30 K grid to ground resistor and a 300 K grid to plate resistor. The screen grid gets between 150 and 175 volts bypassed to ground. Since the cathode is driven there is some drive applied between G2 and the cathode.
I have seen some outstanding THD levels in a driver application while also using a CCS load or bootstrapped plate load resistor. Several different tubes were capable of under 0.2% THD at 50 volts output.
If often goes that way, doesn't it. You have an idea and then you find that others before you have thought about it too.
I used the principle in my direct drive high voltage amp for the QUAD ESLs getting distortion at 1kHz below -100dB. That's two zeros after the comma. I used a MOSFET for the cathode drive.
Interesting you want a triode like transfer, I was aiming for pentode-like operation with a triode to get very high output impedance and an active CCS load I learned from Nelson Pass.
I am now looking at keeping the feedback as tight as possible around the feedback loop to improve stability in the face of transformer phase shift.
Try to get to the situation that the only phase shift at HF is from the transformer and no further impact from other circuit elements.
That would help in increasing loop gain and lowering distortion. In this time and age there's no excuse for a tube amp with >1% distortion 😎
https://audioxpress.com/article/you-can-diy-edd-a-direct-drive-amplifier-for-electrostatic-speakers
Jan
I used the principle in my direct drive high voltage amp for the QUAD ESLs getting distortion at 1kHz below -100dB. That's two zeros after the comma. I used a MOSFET for the cathode drive.
Interesting you want a triode like transfer, I was aiming for pentode-like operation with a triode to get very high output impedance and an active CCS load I learned from Nelson Pass.
I am now looking at keeping the feedback as tight as possible around the feedback loop to improve stability in the face of transformer phase shift.
Try to get to the situation that the only phase shift at HF is from the transformer and no further impact from other circuit elements.
That would help in increasing loop gain and lowering distortion. In this time and age there's no excuse for a tube amp with >1% distortion 😎
https://audioxpress.com/article/you-can-diy-edd-a-direct-drive-amplifier-for-electrostatic-speakers
Jan
The difference between driving the tube's cathode with a transistor's Collector/Drain versus as a follower with the Emitter/Source is an important one. The C/D version gives you the gm of the SS device determining the current in the tube. While the E/S follower gives you the gm of the tube determining the current in the tube. (the lowest gm device rules ) The C/D version gives you the sound of SS ( especially for NPN bipolar Collector drive ), while the E/S version gives you the sound of a Tube. Although the difference between Mosfet gm and Tube gm is more of a semantic call, since both Mosfet and (power) Tube have near square law gm transfer curves.
High dollar tube Amps are typically using Old triodes or Xmitter triodes with fairly (but not 0.001 % ) low distortion. The UnSet scheme allows one to get the same high quality triode curves using a cheap TV Sweep tube or power pentode, but without all the HV expense and high Z OT issues. It also eliminates the issue with lower screen V (for making a psuedo triode) with TV Sweeps.
It's all a matter of taste I think.
High dollar tube Amps are typically using Old triodes or Xmitter triodes with fairly (but not 0.001 % ) low distortion. The UnSet scheme allows one to get the same high quality triode curves using a cheap TV Sweep tube or power pentode, but without all the HV expense and high Z OT issues. It also eliminates the issue with lower screen V (for making a psuedo triode) with TV Sweeps.
It's all a matter of taste I think.
Last edited: