Looking at making an 811A SET, first off, then maybe a P-P 811A, Got some tubes coming, and just got to try them out!
I dont mind winding the O/P Tx, but interstage, forget it--Just too many variables, and the cost of a good interstage isnt really worth the trouble just for experimenting with!
Ive seen quite a few pentode-drive schemes, but not a MOSFET drive anywhere!--Would have thought a MOSFET a good alternative to a great big pentode drive!
Was thinking of a 6SL7 Mu stage followed by a driver stage then the 811A--Any ideas?
I dont mind winding the O/P Tx, but interstage, forget it--Just too many variables, and the cost of a good interstage isnt really worth the trouble just for experimenting with!
Ive seen quite a few pentode-drive schemes, but not a MOSFET drive anywhere!--Would have thought a MOSFET a good alternative to a great big pentode drive!
Was thinking of a 6SL7 Mu stage followed by a driver stage then the 811A--Any ideas?
I think member tubelab.com uses one for driving 833's. Check his site under PowerDrive: http://www.tubelab.com
PowerDrive
I have used the same basic PowerDrive circuit to drive many different transmitting triodes including the 811A. For tubes like the 811A and the 833A that require positive grid bias, the 22K resistor that is connected to the bias pot needs to be connected to a positive voltage source instead of ground. The 1K resistor, the pot, and the 22K resistor form a variable voltage divider that allows the grid voltage to be adjusted over a range from tube cutoff to heavy conduction. This voltage range (and the resistor values) need to be changed. The values will depend on the power supply voltages that you have available. The values shown on the schematic were for an 845. Since an 811A has a very high Mu you won't need a wide voltage range. You can reduce the -400 volt supply to -100 volts or less depending on the plate voltage of the 811A. I would use a + and - 150 volt supply for an 811A. This supply must provide about 40 mA per channel, and since it is connected to the grid of a high Mu tube it should be well filtered.
You can use the fet portion of the circuit with a Mu follower if you want, but I would use the CCS circuit with 1/2 of a dual triode like the 6SL7. A 6SN7 may have enough gain to drive an 811A, depending on your source. I have been using 5842's since I had a bunch of them, and they are the most linear tube that I have found. My supply is getting thin and the price of them has quadrupled in the past few years, so I must find a new "favorite tube".
http://www.tubelab.com/powerdrive.htm
I have used the same basic PowerDrive circuit to drive many different transmitting triodes including the 811A. For tubes like the 811A and the 833A that require positive grid bias, the 22K resistor that is connected to the bias pot needs to be connected to a positive voltage source instead of ground. The 1K resistor, the pot, and the 22K resistor form a variable voltage divider that allows the grid voltage to be adjusted over a range from tube cutoff to heavy conduction. This voltage range (and the resistor values) need to be changed. The values will depend on the power supply voltages that you have available. The values shown on the schematic were for an 845. Since an 811A has a very high Mu you won't need a wide voltage range. You can reduce the -400 volt supply to -100 volts or less depending on the plate voltage of the 811A. I would use a + and - 150 volt supply for an 811A. This supply must provide about 40 mA per channel, and since it is connected to the grid of a high Mu tube it should be well filtered.
You can use the fet portion of the circuit with a Mu follower if you want, but I would use the CCS circuit with 1/2 of a dual triode like the 6SL7. A 6SN7 may have enough gain to drive an 811A, depending on your source. I have been using 5842's since I had a bunch of them, and they are the most linear tube that I have found. My supply is getting thin and the price of them has quadrupled in the past few years, so I must find a new "favorite tube".
http://www.tubelab.com/powerdrive.htm
ray_moth said:
As a source follower, they work very, very well.
I would normally be hesitant to use them as a common source stage, but maybe as the top device of a cascode they could do OK. I've prototyped drivers like that and they measured well, but I've never had one in an amp running music through it.
Mosfet
I agree, at first I had reservations about putting a solid state device in my tube amplifiers. I would not use a solid state device (bipolar, FET, or IGBT) in anything other than common drain (or collector) they are just too non linear. In common drain (collector) they are very linear current amplifiers. Tubes are very linear voltage amplifiers, why not combine the two.
I agree, at first I had reservations about putting a solid state device in my tube amplifiers. I would not use a solid state device (bipolar, FET, or IGBT) in anything other than common drain (or collector) they are just too non linear. In common drain (collector) they are very linear current amplifiers. Tubes are very linear voltage amplifiers, why not combine the two.
Re: PowerDrive
Yes, I checked out that great Website, Tubelabs, and had great fun reading the 833 SE article!--You must live well away from anyone, and have a very forgiving wife!--Excellent article on meter-safety too!
I was thinking of a simplified version of your 'super-Drive' idea, as the 811A according to all the schemes Ive seen so far needs somewhere around +20 odd volts at the grid at somewhere around 10-20 mA, at idle to give 100 odd mA Ia/Ip with Va at 400 giving around 40W Anode Diss. ish... ( havent sourced a mains Tx so far..........)
RCA suggest, at 1000v Vp, with grid at 0v, theres no Ia, unless I mis-read it!
I was thinking of a MOSFET with the source connected directly to the 811A grid, but grounded as in the Super-Drive, with a 10 K. I was thinking of a voltage divider on its gate (with gate-stopper say 200 ohm), with a pot in the lower half to set quiescent current in the 811A. The lower portion of the divider grounded, as the 811A doesnt need any neg volts on the grid, (from what I can gather) thus no need to go to a neg rail with the MOSFET gate voltage stage. The drain of the MOSFET supplied by say a 100V filtered regulated supply. There is no need to have a huge supply voltage for the MOSFET stage, as the drive voltage for the 811A is quite low, but it must Source quite some current! The drive for the MOSFET stage capacitively coupled to the preceeding stage
Am I missing something really obvious and stupid!?? (Probably knowing me!)
tubelab.com said:
Yes, I checked out that great Website, Tubelabs, and had great fun reading the 833 SE article!--You must live well away from anyone, and have a very forgiving wife!--Excellent article on meter-safety too!
I was thinking of a simplified version of your 'super-Drive' idea, as the 811A according to all the schemes Ive seen so far needs somewhere around +20 odd volts at the grid at somewhere around 10-20 mA, at idle to give 100 odd mA Ia/Ip with Va at 400 giving around 40W Anode Diss. ish... ( havent sourced a mains Tx so far..........)
RCA suggest, at 1000v Vp, with grid at 0v, theres no Ia, unless I mis-read it!
I was thinking of a MOSFET with the source connected directly to the 811A grid, but grounded as in the Super-Drive, with a 10 K. I was thinking of a voltage divider on its gate (with gate-stopper say 200 ohm), with a pot in the lower half to set quiescent current in the 811A. The lower portion of the divider grounded, as the 811A doesnt need any neg volts on the grid, (from what I can gather) thus no need to go to a neg rail with the MOSFET gate voltage stage. The drain of the MOSFET supplied by say a 100V filtered regulated supply. There is no need to have a huge supply voltage for the MOSFET stage, as the drive voltage for the 811A is quite low, but it must Source quite some current! The drive for the MOSFET stage capacitively coupled to the preceeding stage
Am I missing something really obvious and stupid!?? (Probably knowing me!)
PowerDrive
The voltage divider for the pot sets the static bias value fot the 811A and for applications low plate voltage does not need to go negative.
The negative voltage supply that is connected to the 10 K resistor is the return for the current through the mosfet. As the grid swings in the negative direction it draws less and less current. In order to keep sufficient current flowing through the mosfet there must be a supply voltage that is considerably more negative than the most negative excursion expected at the output tubes grid. For an 811A operated at 400 volts you might get away with grounding this resistor. I have not tried this. I would experiment with reducing the value of the 10 K resistor at the grid of the tube. If you have a scope, look at the output tubes grid and drive the amplifier to clipping with a sine wave. If there is any distortion on the bottom of the sine wave, you need more negative voltage.
When I did my testing with an 811A, I was using a 1000 volt power supply and negative voltage is required. I used a small isolation transformer that has two 117 volt windings and a silicon diode bridge to generate + and - 150 volts DC to operate the PowerDrive circuit. Total power supply cost is about 15 USD.
The component values shown on the schematic were optimized for an 845 tube. I have used them with a 211. Operation with an 833A required +30 volts DC at the 22 K resistor to get enough tube current.
An 811A needs less drive voltage than any of these tubes. As I get time I will try some of these experiments and update the PowerDrive page.
We are currently under a hurricane warning, and the storm is coming (again). This usually causes my internet service to go away for a day or two. I may not reply to this for a while.
The voltage divider for the pot sets the static bias value fot the 811A and for applications low plate voltage does not need to go negative.
The negative voltage supply that is connected to the 10 K resistor is the return for the current through the mosfet. As the grid swings in the negative direction it draws less and less current. In order to keep sufficient current flowing through the mosfet there must be a supply voltage that is considerably more negative than the most negative excursion expected at the output tubes grid. For an 811A operated at 400 volts you might get away with grounding this resistor. I have not tried this. I would experiment with reducing the value of the 10 K resistor at the grid of the tube. If you have a scope, look at the output tubes grid and drive the amplifier to clipping with a sine wave. If there is any distortion on the bottom of the sine wave, you need more negative voltage.
When I did my testing with an 811A, I was using a 1000 volt power supply and negative voltage is required. I used a small isolation transformer that has two 117 volt windings and a silicon diode bridge to generate + and - 150 volts DC to operate the PowerDrive circuit. Total power supply cost is about 15 USD.
The component values shown on the schematic were optimized for an 845 tube. I have used them with a 211. Operation with an 833A required +30 volts DC at the 22 K resistor to get enough tube current.
An 811A needs less drive voltage than any of these tubes. As I get time I will try some of these experiments and update the PowerDrive page.
We are currently under a hurricane warning, and the storm is coming (again). This usually causes my internet service to go away for a day or two. I may not reply to this for a while.
minus drive
Yes, I see now the error of my ways. As the MOSFET tries to drive the grid near zero, it would start to become more non-linear. At what point this starts probably depends on the actual current through it, which as the voltage is falling, would also be falling--Its bound to cause audible problems at even moderate sound levels, especially as I, like yourself like a little loud Rock music!.
Hmm--Looks like an extra say -50V from the PSU is needed, just to be on the safe-side, Keep that MOSFET happy.--Better still, maybe a +/- 75V supplies to the MOSFET stage, and follow your Super-Drive Technology, and maybe 'tune' the 10 K resistor while observing on the 'Scope--Why change something thats been tried and tested!
Just thought I could make the whole thing a little simpler, but just like everything, there's compromises and shortcomings! Thanks, I guess its time to start winding the O/P Tx!
1-2----1725, Next Tx!--And thats just the Primary!
Yes, I see now the error of my ways. As the MOSFET tries to drive the grid near zero, it would start to become more non-linear. At what point this starts probably depends on the actual current through it, which as the voltage is falling, would also be falling--Its bound to cause audible problems at even moderate sound levels, especially as I, like yourself like a little loud Rock music!.
Hmm--Looks like an extra say -50V from the PSU is needed, just to be on the safe-side, Keep that MOSFET happy.--Better still, maybe a +/- 75V supplies to the MOSFET stage, and follow your Super-Drive Technology, and maybe 'tune' the 10 K resistor while observing on the 'Scope--Why change something thats been tried and tested!
Just thought I could make the whole thing a little simpler, but just like everything, there's compromises and shortcomings! Thanks, I guess its time to start winding the O/P Tx!
1-2----1725, Next Tx!--And thats just the Primary!- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.