Tube HV power supply

[WA4WAX]

Does anyone have a schematic for a full wave, tube DC PSU using a 5U4 or 5R4?

Voltage around 600 VDC under load, 225 to 250 ma rating.

Thanks!!

[W7TFO]

That current will be pretty tough on a 5R4.

A 5U4 is a better choice.

Healthy iron, and a L-C topology is rather simple.

Say, 10Hy or more, 40 to 60uF.

Bleed Ohms @ 10% of output.

73DG

[WA4WAX]

250 mA would be an overload situation.

It will operate around 165 mA, which is reasonable I think.

[w4bfs]

is that full wave center tap or full wave bridge ?

[w8khk]

Matt, are you looking at this theoretically, or from a practical standpoint?

If you are looking at sourcing the components and assembling it, you might consider looking at the high voltage power supply in a typical boat anchor AM transmitter.  The HVPS in the Collins 32V1, 2, or 3 seems to be just what you are looking for.  It even has taps on the HV transformer primary to set a low or high output voltage.  Similar components might be found in the Heath Dixie-100 or Apache TX1, or even the Johnson Viking I or II.

I am fairly certain that if a WTB were posted for the HVPS components, someone likely has already parted out the requisite iron.

[WA4WAX]

Good idea Rick.

Full wave center tap.

[WU2D]

I just doctored this up. It is a standard Pi Type Power Supply with the choke in the negative lead. This takes the pressure off the chokes windings. Up front I would suggest a circuit breaker or 3A fuse and a small variac. And you could put a meter on the output.

73's Mike WU2D

[WA4WAX]

I like it!

Thoughts:  How about a switch in the center tap lead for turning HV on and off?  Also, MOV's in the secondary  to limit surge into the bank when the switch is thrown?

I am going to bring the filament up on a variac, then throw the HV switch.

[w4bfs]

many hv xfmrs do not like to have the center tap raised much above ground.  it is all about how it is constructed.  for example: the EFJ Thunderbolt has a FWCT hv supply.  the construction of the xfmr is such that the center tap is only a few wraps of thin paper away from the xfmr core.  EFJ had the fuse and on/off switch in the center tap to ground in early versions of the T bolt.  later versions switched the primary on/off

if you insist on doing this try using a megger or hipot tester of the ct to core and see what you get   

[KC2ZFA]

is this the schematic but with SS diodes ?

Peter

[w8khk]

Is this supply intended for ham radio use, perhaps a class-C final or a modulator? Or might it be some newfangled medical torture device? (Tongue in cheek, no disrespect intended...  I understand your experiments are non-invasive! Pervasive, yes!)

It is kinda hard to make a valid recommendation without knowing a bit more.  Is the load static, or is it dynamically changing?  What level of regulation is required?  Is the duty cycle 10%, 50%, or maybe 24x7?  If the requirements are defined, perhaps a simulation might be helpful.  I would assume Phil, AC0OB might like to step up to the plate in this stadium.

Choke input would be gentler on the plate transformer, but may impact regulation.  A center-tapped full-wave HVPS is old school for this power level, whereas a full-wave voltage doubler might work out better with a lower voltage secondary with less heat dissipation due to the reduced resistance of the larger wire in the secondary turns. But then again, capacitor heating due to ESR is another consideration with doubler circuits.

More valuable suggestions might result from an understanding of your ultimate goals.  It is very difficult to shoot a bullseye without knowing where the target is located!  But it has happened.  Perhaps the project is so secret that you do not know anything more than we do about the load circuit......

[ka8gef]

Below is a link to PSUd2, a power supply design software....if you do not already have it....it allows you to insert/play with various forms of filtering, basic rectification/filtering (full, half, doubler, choke/cap input etc.) components and values...you can then view the resulting output, how values and changes impact the results as well. Simple to use, provides graphed voltage/current results over your defined period, design problem warnings (i.e rectifier tube limits exceeded) etc.....ka8gef

http://www.duncanamps.com/psud2/

[MikeKE0ZUinkcmo]

Hi Matt,

Here is the schematic for the old Collins 516f-2 power supply using the type tubes you're asking about.   Click the pic for a bigger picture of the schematic. 

[WA4WAX]

What would you recommend for surge protection to the filter bank?

How about a CL-130 in series with the hot lead of the line going to the primary?

[w4bfs]

I prefer step start to a power thermistor

[DMOD]

What would you recommend for surge protection to the filter bank?

How about a CL-130 in series with the hot lead of the line going to the primary?

If you are using a tube rectifier the Filament warm up serves as the voltage ramp up and the tube also has a series plate resistance Rp.

Phil - AC0OB

[Opcom]

The peak forward repetitive current of the 5R4 is 0.71A. It's important to choose the right capacitance to avoid exceeding that or the rectifier life will be shortened.

The effective series resistance of the supply before the input capacitor must be high enough to limit the current to below the PFRC value.

From the tube manual, a 4uF input cap requires a 125 Ohm series resistance. This assumes a nominal impedance of the plate transformer added in, but don't stop there as other factors are in play.

Higher input capacitance will require higher series resistance. It's important to get this figure right or the output voltage and regulation will differ from what is shown in the program. The program defaults to a generic value, but that won't be what your parts offer.

I also use Duncan Amps' 'PSUD' program to help choose components that will work together to get the result, and it's usually very accurate unless I'm trying to do something weird.


A choke input filter gives low ripple and also low peak rectifier current. No soft start is required if the choke is large enough.

The rectifier filament should be allowed to warm up before a situation of high DC output load or large repetitve peak currents, such as in charging a large input capacitor, is allowed.

[Opcom]

2 more

[KA3EKH]

Wow, never imagined that something as simple as building power supplies can be so complicated! I have built a couple for tube and solid state projects once and again and somehow it always turns out what can you do with what you have at hand, but seeing that PSUD program makes me think I must be some form of luddite or cave man by today's standards. But then again, they have software and programs for doing things like building a dipole.
Like the idea of putting the choke in the negative leg of the power supply to ground, tried and tried and just can’t see how that wont work. Something like in the old days of solid state electronics where it was common to regulate the negative side of the power supply to ground, sound idea but a bear to troubleshoot!

[WA4WAX]

Patrick:

I have a UTC S-31.  Nominally 100 ohms, and 6H.  Rated for 225 ma continuous, and in nice shape.

Would that be enough inductance, or would you need to revise a little?

Many thanks!

[DMOD]

Patrick:

I have a UTC S-31.  Nominally 100 ohms, and 6H.  Rated for 225 ma continuous, and in nice shape.

Would that be enough inductance, or would you need to revise a little?

Many thanks!

One could resonate the choke with a 0.5uF cap and use about a 22uF filter cap if using a FW tube rectifier circuit.

Phil

[Opcom]

as requested by OP, LTspice for resonant choke. The schematic contains pulsers to show what happens for step currents, but it also clearly shows the answers for full load and a small bleeder load. This is just a simulation based on assumptions, Z is assumed 67 ohms, but not known for power transformer. simulated with bridge on half-secondary winding voltage 704VAC (700-0-700 for FWCT). YMMV.

accorcing to https://goodcalculators.com/resonant-frequency-calculator/
 2*pi*F= 1/sqrt(L*C):
a 6H choke resonates @120Hz with 0.293uF capacitor.

A 50uF filter capacitor was used after the choke, however ripple is only 3V p-p, so a smaller one could be used.

[w8khk]

We still know absolutely NOTHING about the application for which this "electron inconveniencer" is intended.  We do not know whether the duty cycle is 24/7 or intermittent.  We do not know the required load regulation or allowable ripple.  And yet we all continue to make suggestions as to the design. 

Matt, you need to come clean and tell us if this is for an angel music generator, or another of your medical torture devices.  Inquiring minds WANT TO KNOW!

[Opcom]

Yes let's know more about the application which will be more interesting than the supply.

[DMOD]

Patrick:

I have a UTC S-31.  Nominally 100 ohms, and 6H.  Rated for 225 ma continuous, and in nice shape.

Would that be enough inductance, or would you need to revise a little?

Many thanks!

One could resonate the choke with a 0.5uF cap and use about a 22uF filter cap if using a FW tube rectifier circuit.

Phil

Most resonating choke power supply systems are designed to resonate at 95 Hz in order to provide some roll-off.

We still know absolutely NOTHING about the application for which this "electron inconveniencer" is intended.  We do not know whether the duty cycle is 24/7 or intermittent.  We do not know the required load regulation or allowable ripple.  And yet we all continue to make suggestions as to the design.

Matt, you need to come clean and tell us if this is for an angel music generator, or another of your medical torture devices.  Inquiring minds WANT TO KNOW!

And yes, we need to know the intended application and specs so we don't broadcast unneeded info.

Phil - AC0OB