Trying to make a bench power supply. What would be a cheap easy way to reduce voltage (to protect the series pass transistor)? 3 Amp transformer has these terminals 12-6-0-6-12. No load voltage is 33 VAC (47 Vpeak DC). Primary taps are 0 and 230V. I came up with these options, and can't decide between them.

  1. Couple of 5A rated household AC staircase switches between the 6V and 12V terminals. Cheap and easy, but manual control. Series transistor needs thermal protection against "accidents" (including output short circuits).

  2. Couple of relay tap changer between 6V and 12V terminals. Maybe cheap, not easy. Needs zero crossover, Schmitt trigger control circuit. Relays will probably need eventual replacement.

  3. Triac tap changer. Possibly more complex than above, and may need more parts list. Triac has voltage drop.

  4. Triac (or SCR) to chop AC waveform on secondary side. I have no clue where to begin. Have seen the old HP manuals. But if bad comes to worse, I am ready to pop in an Atmega88.

  5. 60V buck converter. Where to get a reliable one? Is switching noise an issue?

  6. Series pass regulator to reduce voltage to below 40V (max dissipation 2W @ 750mA), then a 40V buck converter, and then a series pass linear regulator? This is becoming confusing, and wasting a lot of volts. Will also have switching noise.

  7. Roll my own buck converter???

I'm leaning towards #1 right now (couple of $0.15 switches). Power supply will be used for analog work mostly (test op amps and audio amplifiers etc and other hobby electronics stuff). Is the inability to smoothly change voltage from zero to 30V or 40V, a likely hindrance? I don't think I've ever had to change voltage by more than ±3V for anything (usually characterizing LED, or optimizing current draw for small projects). Why would someone want to smoothly change voltage all the way from 0 to 30V? Only convenience, or something more?

I'm hoping this works out, as transformers seem to cost less than a third of a good 36V SMPS (same wattage). Only problem with transformers is the >40V at no load, and then it goes downhill.

So, any suggestions?

  • \$\begingroup\$ why are you connecting to the two 12V taps? \$\endgroup\$
    – jsotola
    Oct 21, 2018 at 1:41
  • \$\begingroup\$ To get more than 20V at full load. Connecting across 18V taps gives me only 19 VDC @ 2.5A, but it will give 28 VDC if connected across the two 12V taps. I guess all transformers do this, even if not as much as the one I have. Hench the tap changers. At no load it is 47V across the 12V taps and 35V across 6 and 12 (6-0-6-12). \$\endgroup\$
    – Indraneel
    Oct 21, 2018 at 1:47
  • \$\begingroup\$ You have a transformer with a center tap. Use a positive and a negative voltage regulator. Each one only sees about 18V. You want this kind of split power supply anyways for a bench supply. \$\endgroup\$
    – Janka
    Oct 21, 2018 at 1:51
  • \$\begingroup\$ I have a split supply ±20V 500mA for op amps already. I need something in the 20V 2A region. If I make this one a split supply, the voltage drop is too much at 2A. \$\endgroup\$
    – Indraneel
    Oct 21, 2018 at 1:56
  • \$\begingroup\$ You can set it to +10V, -10V and have your 20V 2A then. Noone can stop you. \$\endgroup\$
    – Janka
    Oct 21, 2018 at 1:59

1 Answer 1


After the thought provoking comments above, I decided to operate on the transformer for some flexibility. I carefully removed the 6V and Common connections from the binding posts, and isolated them. So now I have four 6V windings. The transformer capability is shown below (Amps DC on X axis, VDC (regulated) on Y axis, for the 4 windings if connected in series).

enter image description here

I can potentially get about 6V @ 1.5A from each winding, or 16V @ 1.5A from two windings in series. Two windings in parallel now give 6V @ 3A. And, two windings in series times 2 gives 16V @ 3A. With a XL6009 boost converter, I might be able to get 24V @ 2A from this. So, I guess that solves the problem. Multiple windings are better than a continuous one.

I'm planning to do something like this https://www.youtube.com/watch?v=-04CQ0Mug1o&t=865s (HP E3610A), but for 2 channels instead of one. Each channel with a switch to toggle between 6V @ 3A and 16V @ 1.5A. There will be 2 of those. So those can again be paralleled (6V @ 6A or 16V @ 3A) or put in series (12V @ 3A or 32V @ 1.5A). So that's a lot of possibilities. The two 6V windings of each channel can be connected like this, before it goes to the linear regulator for that channel:

toggle switch to connect/disconnect common terminals

There is a toggle switch to connect/disconnect common terminals (see long wire near wire marked "common"). There will be no short circuit due to the bridge. I will lose a few volts, but hopefully the amps and flexibility will make up for it.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.