I think that I really need to knock up a simple bench supply if I am going to experiment effectively. I have on hand a few 6LQ6 and a hand full of other sweep tubes (onesies) such as 6JS6, 6JN6 etc., a handful of RF pentodes along with a bunch of 12AX7 and 12AU7. I also have a Power transformer from an organ that has 6V and 12V filament windings and a HT winding capable of something in the 300 to 400V range.
What I had in mind was possibly using a 6LQ6 for the main HT pass tube (series regulator) and one of the lesser tubes to provide similar for screen supplies. I thought that a single 12A_7 tube could be used to provide the control signals by using one triode for the B+ and one for the screen.
I suppose that one of the various trannies that I have salvaged from SS equipment could be used to provide a variable bias supply.
Any thoughts on this idea? Circuit suggestions?
What I had in mind was possibly using a 6LQ6 for the main HT pass tube (series regulator) and one of the lesser tubes to provide similar for screen supplies. I thought that a single 12A_7 tube could be used to provide the control signals by using one triode for the B+ and one for the screen.
I suppose that one of the various trannies that I have salvaged from SS equipment could be used to provide a variable bias supply.
Any thoughts on this idea? Circuit suggestions?
a tube pass element is going to necessarily drop a number of volts - why not use a MOSFET instead?
But if you MUST go glass, steve bench is a tough guy to dispute...
But if you MUST go glass, steve bench is a tough guy to dispute...
Last edited:
My favorite bench supply for B+ is a 350 V tranny, a pair of diodes, reservoir cap, and a variac. That way I can get a wide array of voltages and avoid the loss in a pass device.
Calculate the plate dissipation in your pass device at the highest output current, lowest output voltage. It'll be substantial. That said, a lab supply could be a fun project...
~Tom
Calculate the plate dissipation in your pass device at the highest output current, lowest output voltage. It'll be substantial. That said, a lab supply could be a fun project...
~Tom
I have built one on the basis of old industrial design. It uses SG3S, 2x6N9S and 2x6080 (6N13S) tubes. Output voltage is 490 - 360V, 0.5A. It shown on top of the photo attached, bottom you can see amplifier in testing.
This bench power supply need some extra debugging (time delay circuit is not working as expected), and slight adjustments/fixes for PCB design.
Once I complete whole design, I will post schematic.
This bench power supply need some extra debugging (time delay circuit is not working as expected), and slight adjustments/fixes for PCB design.
Once I complete whole design, I will post schematic.
Lot of circuits out there to regulate power amp supplies using MOSFETS. Google VVR.
http://www.rh-tech.org/index.php?topic=41.0
http://www.rh-tech.org/index.php?topic=41.0
Last edited:
What Pete did with a bunch of 807's -- looks like a space heater for the ham shack. (Must be cold down there in TX).
FWIW, I don't like having the current meter in the ouput line per se -- but using a 0.1 Ohm shunt and a low bias current opamp to read the voltage drop. You can power the opamp with a zener across pins 7 and 4 and a bias resistor. Too many burned out Weston and Keithley meter movements around here!
For bias voltages -- just run a 6.3VAC transformer in reverse.
Helpful to be able to lift the filament voltage above ground for some of those 6922 circuits SY has developed.
I think that some of the HP/Harrison HV PS schematics are on BAMA.
Last edited:
I used a 480 volt industrial control transformer. They are designed to generate 240 or 120 volts from 240 or 480 volts. Turned around they make 480 VCT from 120 or 240 volts. New they are expensive. They can be found surplus on Ebay for cheap. They come in sizes from 50VA to several KVA. Resist the temptation to buy the 1KVA size since it cost the same as a 200VA unit. I built a monster power supply using 2 20 amp Variacs and 2 1KVA transformers. I couldn't lift it when I was done and it took up too much bench space.
Of course a fully regulated bench supply would be ideal, but as previously mentioned the heat dissipation becomes a problem. Lets say you want a power supply that goes from 0 to 450 volts at 1/2 an amp. The raw B+ must be about 550 volts if a pass tube is used and 500 volts if a mosfet pass device is used. Set the output to 450 volts and draw 1/2 am amp from the supply. It will dissipate 50 watts + heater heat with a tube and 25 watts with a mosfet. Set the output to 50 volts and draw half an amp. THe dissipation will be 250 watts + heater heat!
Mosfets appear to be the ideal pass device, but they tend do blow up if the power supply is accidentally (or intentionally) shorted. Mosfets tend to fail to a short which will subject your circuit under test to the full raw B+ voltage.
The venerable Fluke 407D solves the dissipation issue with a multi tapped power transformer so that the raw B+ voltage is only 100 to 150 volts above the output voltage. Yes, I have sparked the leads together on mine several times without hurting it although the current meter really doesn't like it. It uses 807's for the pass device.
It is possible to use a variac on the input of the raw B+ transformer to solve the dissipation issue, but this requires some thinking when actually using the power supply. I was planning to use several 120 to 120 volt toroid isolation transformers with their secondaries in series to create a multi tapped secondary (0-120-240-360-480 volts) feeding sweep a tube pass device when I lucked into a surplus HP6448B. This guy goes from 0 to 650 volts at up to 1.7 amps. It is capable of cranking out enough power to run the red board to 500 watts of audio output!
That was my fear with the MOSFETs. Not specifically but in general they seem pretty fragile. I really wouldn't need much over a couple hundred mA for the most part and not likely to want to go much over 400V due to cap costs.
Are the tubes I mentioned like 6LQ6 and other Horizontal Sweep tubes reasonable performers as pass tubes? A pair of 6LQ6 in parallel ought to be able to pass quite a bit of current.
Are the tubes I mentioned like 6LQ6 and other Horizontal Sweep tubes reasonable performers as pass tubes? A pair of 6LQ6 in parallel ought to be able to pass quite a bit of current.
The 6LQ6 is getting rare these days, I would not use them as pass devices. Sell them on ebay and start looking at electrophoresis power supplies.
I haven't tried the 6LQ6 specifically because it is not cheap and I like to use cheap sweep tubes. They should do nicely and require less voltage across them than the typical 6L6GC or 807 pass tube. I was planning to use 6LW6's since I have a fair collection of them. I since scored the motherload of 35LR6's at a hamfest so I have been using them for experiments. I have no plans to build a power supply since I got the HP.
Your power requirements fit with the typical Kinght - Eico - Heathkit - whatever type bench power supply that uses 4 X 6L6GC. I have an old Knight KG-664 that goes from 0 to 400 volts and is rated for 200 mA. It will crank out 250 mA for an hour or so before the transformer gets hot and 300 mA for short periods of time. It has been shorted out dozens of times and I have not had the cover off of it yet. It was my main bench supply before I got the Fluke 407D. I got it at a hamfest for $25. They can still be found cheap at hamfests but go for too much money on Ebay. There are also versions that use 2 X 6L6GC and are rated for 100 mA. I have one of them too, same schematic.
It should be possible to build one, but you would need to kludge something together for the power transformer. It uses one raw B+ winding (about 500 VAC), a seperate floating screen supply winding (about 250 VAC), and yet another winding for bias (about 250 VAC). There is a filament winding for the tubes inside the supply and two more for external use. The circuit is similar to all of the bench supplies from that era (1960's). It uses an 0A2 and an 0B2 for reference, a 6DK6 control tube, and 4 x 6L6GC's for pass tubes. I have seen zeners for reference and 6CB6's for the control tube. Sweep tubes can be used for the pass tubes but the floating screen supply will need to be about 150 volts instead of 320 volts.
I have the manual with schematic scanned to PDF. It is too big to attach here (700K). I can email it to you if you want.
Your power requirements fit with the typical Kinght - Eico - Heathkit - whatever type bench power supply that uses 4 X 6L6GC. I have an old Knight KG-664 that goes from 0 to 400 volts and is rated for 200 mA. It will crank out 250 mA for an hour or so before the transformer gets hot and 300 mA for short periods of time. It has been shorted out dozens of times and I have not had the cover off of it yet. It was my main bench supply before I got the Fluke 407D. I got it at a hamfest for $25. They can still be found cheap at hamfests but go for too much money on Ebay. There are also versions that use 2 X 6L6GC and are rated for 100 mA. I have one of them too, same schematic.
It should be possible to build one, but you would need to kludge something together for the power transformer. It uses one raw B+ winding (about 500 VAC), a seperate floating screen supply winding (about 250 VAC), and yet another winding for bias (about 250 VAC). There is a filament winding for the tubes inside the supply and two more for external use. The circuit is similar to all of the bench supplies from that era (1960's). It uses an 0A2 and an 0B2 for reference, a 6DK6 control tube, and 4 x 6L6GC's for pass tubes. I have seen zeners for reference and 6CB6's for the control tube. Sweep tubes can be used for the pass tubes but the floating screen supply will need to be about 150 volts instead of 320 volts.
I have the manual with schematic scanned to PDF. It is too big to attach here (700K). I can email it to you if you want.
The TV HD finals will work just great as series pass regulators. Although, if you have 6LQ6s, I'd save those for audio final duty. It would be nice if these were more common and cheaper, but just esssss-loads of 6LQ6s got poofed in illegal CB equipment, and there was also a commercial amp that used them as well.
You can find lots more common TV HD types that'll work, aren't excessively pricey, and available.
Hmm, just thinking... I have a couple dozen 500V/26A FETs (IXFH26N50) in my junk box. I wonder if they'd work for pass duty if I found an appropriate supply circuit. http://ixdev.ixys.com/DataSheet/91525.pdf
I can send someone a few if they want to play with designing a circuit.
..Todd
I can send someone a few if they want to play with designing a circuit.
..Todd
Hello ,
more power supply circuits , but in german language :
Verschiedene Roehren-Netzteile und ihre Schaltungen
Regards and a happy new year ,
Alexander .
more power supply circuits , but in german language :
Verschiedene Roehren-Netzteile und ihre Schaltungen
Regards and a happy new year ,
Alexander .
Thanks George. You have mail.
I probably should sell the 6LQ6s and use the 6SJ6 instead. I guess I should knock up a test jig of some sort and do a quick manual gm test to make sure that they are at least functional. I have a tube tester but it has not been restored to usable condition yet. I suspect that plotting some data points by hand would be a more accurate measure of condition anyway.
I probably should sell the 6LQ6s and use the 6SJ6 instead. I guess I should knock up a test jig of some sort and do a quick manual gm test to make sure that they are at least functional. I have a tube tester but it has not been restored to usable condition yet. I suspect that plotting some data points by hand would be a more accurate measure of condition anyway.
I have two tube testers. I quit using them because I kept finding tubes that tested good but didn't work in a real circuit. One is a popular Gm tester (I don't remember which one since it has been in the warehouse for years). The other is a programmable (12 slide switches) creation that probably seemed like a good idea at the time but isn't too repeatable today.
I have found that even small signal tubes can have good transconductance but show high distortion in an audio amp. I sorted a big box full of 6U8's and 6BL8's for distortion and was going to investigate this effect but never found the time. Big output tubes and sweep tubes can show normal transconductuctance at the power levels seen in a tube tester, but fail misaeably when asked to produce big power. On the other hand if the tube tester says it's dead, it usually is.
I emailed the manual a few minutes ago.
i thought i'll give some inspirations to frontpanel, this is my HV PSU, use analog meters if you can, much easier to see when things goes wrong..
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.