Bench supply ideas

[mashaffer]

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?

[aardvarkash10]

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...

[boywonder]

High-voltage bench supply

[tomchr]

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

[LinuksGuru]

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.

[Printer2]

Lot of circuits out there to regulate power amp supplies using MOSFETS. Google VVR.


http://www.rh-tech.org/index.php?topic=41.0

[jackinnj]

Originally Posted by boywonder

High-voltage bench supply

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.

[tubelab.com]

My favorite bench supply for B+ is a 350 V tranny, a pair of diodes, reservoir cap, and a variac.

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!

[mashaffer]

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.

[astouffer]

Originally Posted by mashaffer

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.