Hi-frequency AC filament supply
In the old Hell & Bowel 16mm filmsound projectors that I used to run in high school there were three 6V6's on the amplifier chassis. Two perform the P-P audio output function while the third tube is used as a HF power oscillator to provide energizing current for the optical sound systems' exciter lamp. It was never clear to me why they didn't just use pure filtered DC for this light bulb. The reason for this care and trouble is that any intensity modulation would be amplified as an audio signal. This power oscillator ran at a pretty high frequency (several hundred KHz?) as I recall the transformer was an air cored coil reminiscent of an oscillator coil in an old console type tube radio.
This idea would not be hard to implement in a DIY amp employing DHTubes. One could wind the filament xfmer easily DIY and one power oscillator could handle all the filaments, each getting it's own isolated winding as needed.
BTW, these amps make wonderful practice/jamming amplifiers for guitar and harps. Just go in from the Mic input on the front control panel. I had one complete working projector some twenty years ago and used it as an electric guitar amp. It got left behind during a major long distance move back to civilization from the wilderness of the Western Arctic with appreciative musician friends where there aren't many things that will work as a guitar amp. I found a complete 16mm projector at a thrift store in Buffalo about 4 years ago and scopped it for five bucks remembering how sweet that little amp in there is.
tubelab.com said:
In the old Hell & Bowel 16mm filmsound projectors that I used to run in high school there were three 6V6's on the amplifier chassis. Two perform the P-P audio output function while the third tube is used as a HF power oscillator to provide energizing current for the optical sound systems' exciter lamp. It was never clear to me why they didn't just use pure filtered DC for this light bulb. The reason for this care and trouble is that any intensity modulation would be amplified as an audio signal. This power oscillator ran at a pretty high frequency (several hundred KHz?) as I recall the transformer was an air cored coil reminiscent of an oscillator coil in an old console type tube radio.
This idea would not be hard to implement in a DIY amp employing DHTubes. One could wind the filament xfmer easily DIY and one power oscillator could handle all the filaments, each getting it's own isolated winding as needed.
BTW, these amps make wonderful practice/jamming amplifiers for guitar and harps. Just go in from the Mic input on the front control panel. I had one complete working projector some twenty years ago and used it as an electric guitar amp. It got left behind during a major long distance move back to civilization from the wilderness of the Western Arctic with appreciative musician friends where there aren't many things that will work as a guitar amp. I found a complete 16mm projector at a thrift store in Buffalo about 4 years ago and scopped it for five bucks remembering how sweet that little amp in there is.
I used a toroid core for the transformer, and a solid state push pull amplifier to drive it. That was driven with an HP audio oscillator. The problem comes from regulating the gain to keep the filament voltage constant, without letting the amplifier clip. Each 811A requires about 25 watts of filament power. The resistance of the filament changes drastically as the tube warms up. Most bipolar transistors tend to melt at this power and frequency. Mosfets work fine.
tubelab.com said:
I would be taking the vacuum tube power oscillator route I think if I were to experiment in this direction. It would be an excellent opportunity to add more impressive looking tubes to a chassis and would keep the purity issue of sand-free better served for anyone to whom that might be a philosophical or religious issue. The low cold resistance of the driven filaments in the load would not be a problem for a tube based power supply. True, a mosfet based HF filament supply would result in a higher efficiency amplifier, but efficiency is apparently not a concern in tube audio hi-fi. If it were, it would almost not exist.
When I have the time to explore this further I will probably resume the sand state experiments, since I have that engineering mentality about choosing the best component for the job.
For those that want to follow the pure tube path, we need to find the schematic for the old Tektronix 504 scope. It used a power supply that derived all of the operating voltages off of a power oscillator operating at an ultrasonic frequency. If my fuzzy memory is correct it used a 6DQ6A. It's been 25 years since I worked on these.
Anyone out there got the diagram?
For those that want to follow the pure tube path, we need to find the schematic for the old Tektronix 504 scope. It used a power supply that derived all of the operating voltages off of a power oscillator operating at an ultrasonic frequency. If my fuzzy memory is correct it used a 6DQ6A. It's been 25 years since I worked on these.
Anyone out there got the diagram?
811 Amp
Hello,
I was looking for an 811 SET circuit to build and it looks like I've found one. Thank you Astouffer.
I also saw the second amp with the transistor (nasty things) in the circuit. But with my heavy towards thermionic emission (an age thing),that one is not for me.
Has anyone got an update or mods to the circuit since 2005?
I noticed some chat about pentodes and bias.... any comments on that??
I did see the "Ganti 811" amp but not keen on the transformer couplings between each stage.
I'm reading that the Svetlana 811 valve is the one to go
for. So I'm looking for a pair if anyone has a vendor.
Thank you for any help.
John
Aug-29-07
Hello,
I was looking for an 811 SET circuit to build and it looks like I've found one. Thank you Astouffer.
I also saw the second amp with the transistor (nasty things) in the circuit. But with my heavy towards thermionic emission (an age thing),that one is not for me.
Has anyone got an update or mods to the circuit since 2005?
I noticed some chat about pentodes and bias.... any comments on that??
I did see the "Ganti 811" amp but not keen on the transformer couplings between each stage.
I'm reading that the Svetlana 811 valve is the one to go
for. So I'm looking for a pair if anyone has a vendor.
Thank you for any help.
John
Aug-29-07
I have a couple of 811A's and have been anxious to build something with them. I have a copy of the schematic attached to post #5 of this thread but am having difficulty understanding several things.
If some kind soul would help me, what is the device labeled E202? Also, what voltage/wattage are the zener's?
I just noticed something else, it looks like there is a 100 ohm resistor attached to the paralleled 33 ohm resistors on the cathode of the 811A, but then there is an arrow with 200 ohm/3W in parentheses. Is this a 200 ohm potentiometer?
And finally, the 2SC3783 transistor is hard to source in the U.S.. Can anyone recommend a suitable substitute?
Thanks.
If some kind soul would help me, what is the device labeled E202? Also, what voltage/wattage are the zener's?
I just noticed something else, it looks like there is a 100 ohm resistor attached to the paralleled 33 ohm resistors on the cathode of the 811A, but then there is an arrow with 200 ohm/3W in parentheses. Is this a 200 ohm potentiometer?
And finally, the 2SC3783 transistor is hard to source in the U.S.. Can anyone recommend a suitable substitute?
Thanks.
E202 looks like a constant current source diode. The zeners are a good question. HZ36L is a valid part and the datasheet says 400mw at 36 and 24 volts. http://www.datasheetcatalog.com/datasheets_pdf/H/Z/3/6/HZ36L.shtml
Its probably two 100 ohm 3 watt resistors wired in series for 200 ohms at 6 watts.
That transistor regulator isn't really necessary. You could get by with a suitable dropping resistor and bypass cap. If you want to give it a try MCM electronics lists the 2SC3783 as in stock for $5.25
Thanks once again. From the data sheet you so kindly provided, it looks like a 1N5305 is a suitable substitute. Now I can start collecting all the necessary bits and have a go at building it once the weather turns. (With the price of fuel, I may be spending a lot of time huddled around the amplifier this winter.)
Have you seen this:
811A A2 simple amp
This is the same ckt topology as post #5 with substantially less
complexity and I will argue better performance.
My gripes with post #5 are as follows:
He is forced to use a semiconductor B+ regulator on his driver
stage because he uses a variable-resistor divider network to bias
the EL84 cathode follower at the grid! If, in fact, he used a choke
or resistor to bias the cathode of the EL84, he'd not need this complexity.
He's using feedback from the 811A plate to the 12AX7 cathode.
Gaaaah !!!!
His operating point is too low. At lower B+, the 811A grid is going to produce GOBS of grid current. He's biased the 811A hot at 33V.
The result is you'll get a lower Ra, but also less power and instead
have to dissipate alot of current from the 811A grid. Keep in mind
that grid current is not useable for audio power - it's simply a sunk
cost!
He's using DC filaments but no common mode chokes! I'd not
even bother and use straight AC with a humbucking pot in the
filament circuit to null out hum. If you're going to do DC filaments, do them right - use CMC's. Heck - you can get CMC's in the 6A range upwards of hundred of mH in the Digikey catalogue.
You're better off with slightly higher B+ (400V or alittle more). In
this regime, you'll get MORE plate current and LESS grid current.
You'll get more power. Your driver stage will be alot cleaner.
With my design, I run at 430V, which is a typical, attainable B+ for the kinds of off-the-shelf transformers for guitar amps or the Hammond catalog. The grid is happy at 22-24V (you can go higher if you want). You can use many different tubes on the cathode follower: 6{v,w,y,k}6GT, 6CL6, EL34, 6L6... Finally, since we need to swing 40V, you can use something from the 6SN7 family
as your gain tube. That includes stuff like the 5687, the 6CG7,
12BH7 and all that jazz. Since the gain stage only needs to
develop 40V, the amp is now 2Vp-p input sensitive. This means this
design DOESNT NEED a preamp - you can simply place a 100K stereo attenuator pot on the amp and drive this directly from a D/A converter.
-- Jim
811A A2 simple amp
This is the same ckt topology as post #5 with substantially less
complexity and I will argue better performance.
My gripes with post #5 are as follows:
He is forced to use a semiconductor B+ regulator on his driver
stage because he uses a variable-resistor divider network to bias
the EL84 cathode follower at the grid! If, in fact, he used a choke
or resistor to bias the cathode of the EL84, he'd not need this complexity.
He's using feedback from the 811A plate to the 12AX7 cathode.
Gaaaah !!!!
His operating point is too low. At lower B+, the 811A grid is going to produce GOBS of grid current. He's biased the 811A hot at 33V.
The result is you'll get a lower Ra, but also less power and instead
have to dissipate alot of current from the 811A grid. Keep in mind
that grid current is not useable for audio power - it's simply a sunk
cost!
He's using DC filaments but no common mode chokes! I'd not
even bother and use straight AC with a humbucking pot in the
filament circuit to null out hum. If you're going to do DC filaments, do them right - use CMC's. Heck - you can get CMC's in the 6A range upwards of hundred of mH in the Digikey catalogue.
You're better off with slightly higher B+ (400V or alittle more). In
this regime, you'll get MORE plate current and LESS grid current.
You'll get more power. Your driver stage will be alot cleaner.
With my design, I run at 430V, which is a typical, attainable B+ for the kinds of off-the-shelf transformers for guitar amps or the Hammond catalog. The grid is happy at 22-24V (you can go higher if you want). You can use many different tubes on the cathode follower: 6{v,w,y,k}6GT, 6CL6, EL34, 6L6... Finally, since we need to swing 40V, you can use something from the 6SN7 family
as your gain tube. That includes stuff like the 5687, the 6CG7,
12BH7 and all that jazz. Since the gain stage only needs to
develop 40V, the amp is now 2Vp-p input sensitive. This means this
design DOESNT NEED a preamp - you can simply place a 100K stereo attenuator pot on the amp and drive this directly from a D/A converter.
-- Jim
Jim,
Thanks very much for your post. I printed out the schematic and will give it a try during the cold months ahead.
I assume you have built this circuit. How would you characterize the sound? Also, the 811A's I have have a radioactivity warning label on the boxes. They were made in 1962. Do these things contain a radioactive isotope or do they generate x-rays in use?
Mike L.
Thanks very much for your post. I printed out the schematic and will give it a try during the cold months ahead.
I assume you have built this circuit. How would you characterize the sound? Also, the 811A's I have have a radioactivity warning label on the boxes. They were made in 1962. Do these things contain a radioactive isotope or do they generate x-rays in use?
Mike L.
Oh yes, I've built it and it sounds amazing. I was even surprised
at how little I can hear the background AC hum on 98dB eff
speakers. This amp has alot of power (10W) and can drive many
kinds of speakers - not just limited to horns or Klipschen.
As far as radioactivity, the filament is coated with trace amounts of Thorium which enhances the metallurgy and reliability of
the tungsten at hi temp. Natural thorium is weak alpha emitter,
similar to what's in your smoke detector. Thorium is also used
in Coleman lantern mantles, as a catalyst. All DHTs use thorium:
845s, 211s, 572 and the big RF tubes.
The radioactivity labelling on the boxes is more for the benefit
of military cargo handlers.... They like to know if cargo is supposed to be hot so that they dont have to worry if something
(1) got contaminated or (2) a stray nuclear weapons component
has somehow gotten into the regular parts depot shipment for the Signal Corps.
If you follow the news, last year, Minuteman ICBM nosecone parts were accidentally shipped to the Taiwanese army.. They
Taiwanese had ordered batteries for a helicopter, and due to mixups in warehouse, got something for more interesting to play with!
-- Jim
at how little I can hear the background AC hum on 98dB eff
speakers. This amp has alot of power (10W) and can drive many
kinds of speakers - not just limited to horns or Klipschen.
As far as radioactivity, the filament is coated with trace amounts of Thorium which enhances the metallurgy and reliability of
the tungsten at hi temp. Natural thorium is weak alpha emitter,
similar to what's in your smoke detector. Thorium is also used
in Coleman lantern mantles, as a catalyst. All DHTs use thorium:
845s, 211s, 572 and the big RF tubes.
The radioactivity labelling on the boxes is more for the benefit
of military cargo handlers.... They like to know if cargo is supposed to be hot so that they dont have to worry if something
(1) got contaminated or (2) a stray nuclear weapons component
has somehow gotten into the regular parts depot shipment for the Signal Corps.
If you follow the news, last year, Minuteman ICBM nosecone parts were accidentally shipped to the Taiwanese army.. They
Taiwanese had ordered batteries for a helicopter, and due to mixups in warehouse, got something for more interesting to play with!
-- Jim
Jim,
I follow your work and I like it. I am thinking of building your 811a circuit. I am unclear about PS though. At idle what load current I should consider while modeling the PS?
I undestand 811/811A ra is high, how 5K OPT behabes then! I do not want the input tube because I will use my AIKIDO pre so how I can go for some feedback If I want to....!!
Pls need your help.
Regards
I follow your work and I like it. I am thinking of building your 811a circuit. I am unclear about PS though. At idle what load current I should consider while modeling the PS?
I undestand 811/811A ra is high, how 5K OPT behabes then! I do not want the input tube because I will use my AIKIDO pre so how I can go for some feedback If I want to....!!
Pls need your help.
Regards
Miles Prower said:
I doubt that the radioactivity from such a tube is very high at all. It would only be an issue if the tube was broken and there was a possibility of ingesting / inhaling the stuff.
I’m a licensed handler and use quite potent radioactive sources (ones that would inflict amputation-requiring burns with a several hours of body contact) at work.
A guy where I work once drilled into a sealed source of several GBq in the lathe and sprayed caesium salts all over the place. I don’t have any 811’s at the moment, but I could measure an 813 or two and some other tubes with the calibrated equipment I have access to if anyone is interested.
In fact, I’d be interested my self just to see how easily (or otherwise) any radiation from such tubes can be detected.
I have a collection of old WWII military radios and I’ve been quite surprised at just how hot the glow in the dark radium paint on a few of the various dials was.
Hmmm.....
Something intersting to do on mondays lunch break.
Cheers,
Glen
I realized how old this thread is, but I found it searching 811 power supply info... I started an 811 SE amp last fall when I found some 811s at a meet... It's been thru 4 iterations so far, and the only thing left from the original schem is the 811!!
I'm using a 10A 6.3V FT and a 25A 50VDC bridge with 2x 22000uF caps, which gets it dead quiet, but the voltage is lower than I'd like (5.85VDC)... The amp sounds very good and is running on the breadboard with the last iteration for teh L channel, and the 3rd for the Right... Needed two channels running to load the PS to fine tune it ...
I am considering using a separate FT for each 811 heater, to bring the V up, but not sure if its really necessary, as the Vp is around 390-400V, which is pretty low for this tube... Grid is about 24-26VDC...
Anyone still working with 811s? What did you do to nail down the heater voltage? I'm not too keen on using a lot of SS devices, but open to any input...
Thanks in advance...
I'm using a 10A 6.3V FT and a 25A 50VDC bridge with 2x 22000uF caps, which gets it dead quiet, but the voltage is lower than I'd like (5.85VDC)... The amp sounds very good and is running on the breadboard with the last iteration for teh L channel, and the 3rd for the Right... Needed two channels running to load the PS to fine tune it ...
I am considering using a separate FT for each 811 heater, to bring the V up, but not sure if its really necessary, as the Vp is around 390-400V, which is pretty low for this tube... Grid is about 24-26VDC...
Anyone still working with 811s? What did you do to nail down the heater voltage? I'm not too keen on using a lot of SS devices, but open to any input...
Thanks in advance...
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.