An All-Tube 500V 100mA Regulated Power Supply

Foreword

This page is the outcome of pure nostalgia. If you are awkward enough to construct an all-tube power supply in 2018 or 2019, read on. All others better leave this page because it does not make any engineering sense.

This page is a stub that reflects the status of an ongoing project. Updates will be published as the project proceeds and results are encouraging.

Principal Design Decisions

The power supply I have in mind runs along the following mental framework of ideas:

  • No exotic parts, of possible. Common off the shelf tubes, caps, transformers, resistors, … I admit that in the tube world this goal is not easy to accomplish
  • No overloaded parts. The design should support a 24/7 operation at full power with shorted output for an indefinite time. Safety margins are a must.
  • Regulation from 0 to 500V. Maximum current is 100mA.
  • Only a modest output capacitor charge so a DUT with a fault is not automically electrocuted.
  • No silicon, glass only. This makes design a whole lot harder than in the semiconductor world.
  • Space consumption, weight, efficiency, … are secondary. Safety, easy maintenance, thermal stability is primary.
  • Pentodes as regulator tubes. We need some gain here, and the effect of anode voltage variations (ripple) should be made small. Triodes are not good at that.
  • Analog Meters.

Another design decision is to create a floating regulator. If we make a ground-based design, reaching 0V at the output asks for a quite complex circuitry, also involving auxiliary power supplies. For a floating design, 0V is not problem at all.

Raw Supplies, Choice of Tubes

If we stick to our target of using only common components we need the following voltages:

  • a raw HV input voltage around 700V minimum. We need ca. 200V of headroom to ensure that the tubes dont draw screen grid currents and stay in the classic linear range.
  • an auxiliary HV voltage to supply current to the error amp, the voltage reference and other circuits.
  • a lot of heater supplies for the tubes used. In a floating design, it is easy to set the level of heater voltage in a range where the maximum cathode-heater voltage is not exceeded.

Tube Candidates can now be selected.

Circuit Sections and Tube Candidates

For the raw HV rectifier tube, we should have at least 1kV of maximum rectified voltage in a mid-tap configuration. Candidates here are the PY500 series, with anode currents of 400mA and several kV of blocking voltage. Voltage drop would be only 20V for 400mA. We need to compute the size of the rectifier reservoir capacitor by maximum pulse current, and an inrush-limiting resistor would be needed. A RC-filter with some 20V full load drop would be used to further reduce ripple before the pass tubes. An advantage of the mid-tap configuration is that the two tube heaters sit at the same potential and can be fed from a common heater transformer.

The auxiliary power supply should be around 150V. A small double rectifier like the EZ80. The auxiliary power supply should provide power to the error amp, the voltage reference, a constant current source and maybe some auxiliary circuits. My guess is a current consumption of about 30mA maximum.

For the error amp, a constant current source and a voltage reference we could use small-signal pentodes, a triode based constant current source and a neon gas stabilizer tube. I am still searching for the best candidates, but I have not made a final decision yet. ECC83 (12AX7), another ECC83 as the constant current source plus an OA2 as a stabilizer are my first choices.

Finally, we need to choose the pass tubes. The maximum power dissipation must exceed the raw HV voltage multiplied by the short circuit current, so this means at least 70W here (700V * 100mA). In the interest of cheap and easily obtainable parts, I chose sweep tubes over more exotic (and expensive) tubes with higher power ratings. We could dissipate this marginally by two PL519 sweep tubes, but in the interest of sustainability I chose 3 x PL519 tubes in parallel, with equalizing resistors.

The tube candidates chosen so far:

Main HV rectifier – 2xPY500

PY500TubePhoto.png

Click here for the Datasheet of the PY500A

Auxiliary Rectifier – EZ40

EZ80TubePhoto.png

Click here for the Datasheet of the EZ80

Voltage Reference – OA2

OA2TubePhoto.png

Click here for a Datasheet of the 0A2

Error Amplifier – ECC83, same for constant current source.

ECC83TubePhoto-2.png

Click here for a Datasheet of the ECC83S

 

 

 

 

 

 

 

One thought on “An All-Tube 500V 100mA Regulated Power Supply

Leave a comment