Bench Power Supply for Tubes

[David, K3TUE]

I'm in the process of sketching out a design and collecting parts for a bench supply for tube projects.  I'd like to use this thread as a place to ask the willing questions.

Filament Supply Section Questions:

1) For the filament portion of the PS, I have seen schematics where some supplies use AC and some use DC for filament power.  Is there a good reason why I should make the filament supplies DC over AC?  Or should I make it switchable?

2) If I were going to go the fixed filament voltage route, should I need any more filament voltages than 2.5V, 6.3V, and 12.6V?  I know there are other filament voltages used out there, but just thinking of experimenting, I would like to cover the most common for most projects ranging from the range of tubes I am most likely to use.  Tubes not often used would warrant the building of their own power supplies when a project come up (voltages I excluded from common use include 5V (for HV PS rectifier filaments), 10V (for larger tubes which are less likely to be a simply project), that is just as best I know).

[Rick K5IAR]

Hi David,

This will be a very interesting project and thread.  I have wanted and threatened to do the same for years, but haven't.  Maybe your project and the comments you receive here will help insipire my giddy-up-go to do get with it.  I had always thought I'd use a variable filament supply with a large rheostat and volt meter to control it.  I had never really considered a DC filament supply, but it would sure be something to think about, especially if it were switchable between AC and DC as you suggested.  What do you have in mind for high voltage?  How about bias voltages?  My original thoughts were to build an all tube supply with a choke input, but I have just about decided when and if I ever build one it will be solid state to cut down on heat and maintenance.

I'm sure you're going to get a lot of great comments and advice.

73,
Rick/K5IAR

[David, K3TUE]

What do you have in mind for high voltage?  How about bias voltages?

I have ideas for them, but those thoughts and questions will be next.  But I'm trying to keep this as orderly as I can for now, so I'll wait to expand on them.

[David, K3TUE]

I had always thought I'd use a variable filament supply with a large rheostat and volt meter to control it.

I have considered this, but since I have seen enough circuits that require up to 2 different filament voltages.  It seems like there are a small number of common filament voltages used I was hoping to do my best to determine which were the most popular, and perhaps set up an additional single one on a variac for the odd voltage not covered by the common ones.  Still thinking about it.

[KD6VXI]

I'm in the process of sketching out a design and collecting parts for a bench supply for tube projects.  I'd like to use this thread as a place to ask the willing questions.

Filament Supply Section Questions:

1) For the filament portion of the PS, I have seen schematics where some supplies use AC and some use DC for filament power.  Is there a good reason why I should make the filament supplies DC over AC?  Or should I make it switchable?

2) If I were going to go the fixed filament voltage route, should I need any more filament voltages than 2.5V, 6.3V, and 12.6V?  I know there are other filament voltages used out there, but just thinking of experimenting, I would like to cover the most common for most projects ranging from the range of tubes I am most likely to use.  Tubes not often used would warrant the building of their own power supplies when a project come up (voltages I excluded from common use include 5V (for HV PS rectifier filaments), 10V (for larger tubes which are less likely to be a simply project), that is just as best I know).

AC on the fils can impress hum.  For testing purposes, this shouldn't make much of a difference.

DC on the fils can cause one end of the fil to be <<cold>> and the other end hot:  Some people say this is bad.  If you think about it, on a 3-500 style tube, it makes sense.  One side grounded and one riding above ground doesn't sound like good mojo.  I would think for a quick bench supply, this wouldn't be a big deal.

Hope it helps.

KD6VXI

[W1UJR]

How about just robbing the power supply section from an old tube tester?
You've got everything that you need right there, including metering!

Easy to in bad cosmetic condition, just save the guts and build your own enclosure.
I favor the early Weston and Simpson models.

Last one I picked up was a nice Supreme unit for $1 at the Union, Maine hamfest.

[David, K3TUE]

What kind of tube projects do you plan to use this general purpose power supply for?

I was expecting to "limit" the use of these power supplies (filament, LV, HV, bias) to:
o  AF amps
o  Small Novice/PW CW/AM transmitters
o  Receivers
o  Perhaps sub-100W transmitters

Were I to start looking at playing with transmitters over 100W, I think that a dedicated  power supply would seem more warranted.

I was figuring making each filament supply capable of 4-8A should be sufficient.

[AB2EZ]

If the tube has a filament that is separate from the cathode, and not connected to the cathode... then using DC rather than AC will be fine. There are some subtleties, but to a very good approximation, the applied DC voltage across the filament (grounding one side of the filament, and one side of the filament supply, is ok) should be the same as the rated rms voltage.

[Example of a subtlety: this assumes that the thermal time constant of the filament is long enough so that its temperature and, therefore, its resistance does not vary much during a half cycle of 60 Hz AC, when an AC filament supply is used. Thus the fixed (not varying with time) value of the resistance of the hot filament when DC is applied is the same as the approximately fixed (not varying significantly with time) value of resistance it has when an AC voltage having the same rms value is applied]

If the tube is of a type where the filament is the cathode, or one side of the filament is connected to the cathode... then things get more complicated. In that case, if you want to use a DC filament supply, it will have to be floating (i.e., neither side connected to ground). The floating DC filament supply would be connected across the filament, and the cathode connection would be made to either side of the filament (but, of course, not both).

Best regards
Stu
AB2EZ

[WU2D]

Hi David,

High quality military surplus transformers are great but I also have a lot of respect for TV transformers. I have no comment on AC or DC on fils- I have used both.

I have a couple of bench supplies which are based on old TV transformers and I use them constantly as I am experimenting. These are capable of handing small transmitters up to the 75 Watt input level.

Try to find an old TV transformer from the 50's though early 60's. This should have at least a 5V and a 6.3V or 12.6 VCT winding and a HV center tapped winding. This would give you a good foundation for a supply.

This is a rough example:  EBAY 250144068782  This transformer is not beyond repairing and cleaning up, as  long as there are no shorts.

A transformer like this would give you 6.3VAC, 5VAC, and almost 12VAC when the two windings are series aiding. If a conventional full wave center tapped system were used on the 700VCT, with a cap input filter, you would get at least 450 V. With a tapped bleeder, you could also obtain 200 - 250V for a low voltage and even provide a 150V VR tube output. Bias can be obtained by hooking a backwards 12V or 6V transformer to the 5V winding and rectifying it.   

Mike WU2D
 

[Rick K5IAR]

Try to find an old TV transformer from the 50's though early 60's. This should have at least a 5V and a 6.3V or 12.6 VCT winding and a HV center tapped winding. This would give you a good foundation for a supply.

I can vouch for this idea.  Those old TV transformers were designed to carry a considerable load, both high voltage and filament.  I've used them for replacements many times in small transmitters.

Rick/K5IAR

[K3ZS]

I was fortunate to pick up an old Lambda lab supply at an estate sale.    It has regulated high voltage adjustable from about 50 to 500 VDC at 200 mA max. It uses about 8 or 10 6V6 tubes as the series regulator.    It also has a variable bias supply to about 200 volts, with a higher output impedance that limits its current capabilty.   Both supplys are floating so that they can be used individually or in a series combination at any polarity.   The filament is a dual 6.3 VAC supply that can be used in parallel to combine the currents or in series for 12.6 VAC.    Thought you would be interested in what a vintage lab supply could provide.