A Tiny Bench Power Supply

One of the more popular projects for beginners in electronics is a power supply. Yes, you can always go to Amazon and buy a nice power supply, but unfortunately, we haven’t set up our Amazon affiliate links yet. Instead, we’ll have to go with the next best thing and check out [Tron900]’s mini bench power supply build. It’s extremely capable and cute as a button.

The design goals for this project were to build a small and compact unit using mostly salvaged and recycled components, with all through-hole circuitry. The best guide you’ll ever find for a DIY power supply is one of [Dave Jones]’ earlier video series going over the construction of an adjustable power supply based on an LT3080. [Tron] didn’t have this regulator on hand and wanted to base his design around an op-amp instead. After rummaging through his parts, he found what he was looking for: a TIP3055 power transistor, a neat enclosure that could double as a heatsink and an AD680 voltage reference.

The design of this power supply was simulated in SIMETRIX, and after a few revisions [Tron] had a circuit that worked reasonably well. The circuit was populated on a piece of perfboard, a fantastic front panel was constructed, and one of those ubiquitous volt/ammeter panels added.

This is just a one-off project, but the results are fantastic. This is a very small, very capable power supply that does everything [Tron] needs. It’s accurate enough, at least when measured with a fancy benchtop HP meter, and looks adorable. What more could you want in a benchtop power supply?

22 thoughts on “A Tiny Bench Power Supply

  1. Cute, but it has a couple of beginner errors in it. Since the LM324 input range does not cover the positive rail, he used a divider, which without matched resistors is quite bad for common mode rejection. Also learned this lesson the hard way. Extra: it becomes very slow because of the triple slow op-amps which control the transistor in case of current limitation. This is also why it required so much output capacitance, which is not desirable for such a small supply.
    I would recommend instead to adapt his parts to the electronics lab power supply design, even using the fancy reference instead of the one with zenner diode.
    Still, I admire the project, but has a few things to fix.

  2. Dave Jones’ series was pretty good until he decided to scrap the “down to 0 volts output” spec. I think he changed it (which is the one thing he told people NOT TO DO in the first or second video) to a minimum of 1.2 volts, which is terrible.

      1. Less than 2 min. into his first video he listed “0 volt output” as a requirement!


        @ 3:40 – “Those are our specs, remember them!”

        This may not be a problem for you, but many parts these days require a very low supply voltage. Some even less than 1v. So I was a bit upset that after watching several hours of his series he threw away that requirement.

    1. He did a lot few weird things in there, like using the expensive LT3080 almost as a simple tranzistor just to try and reach 0V. A negative supply would have fixed things easier and end up allowing real zero output (note his will still not output zero in case of an overload).
      [Says the guy who never managed to finish his own design….]

      1. My school had a design like this (and we where not allowed to change the design, flowers are red, etc.) it used a -3V reference. When switching off the supply, unplugging it, something else tripping a mains breaker or in case of power loss, the capacitors of this -3V reference would run out first, resulting in a 3V jump of the reference and also in the power supply output. This was 20 years ago when my ARM computers needed a 1.5V core supply and 3.3V IO supply. The rest wast not that bad, but this single feature made the whole supply too dangerous to use.

      2. @electrobob: I haven’t examined [Tron900]’s design in detail yet, but you’re right – a well designed small negative-rail (e.g. charge-pump) source solves a lot of issues with reaching 0V, especially with linear regulator and (relatively simple) switching regulator designs. As for the “expensive” LTC part, I agree. LTC parts are expensive and sometimes hard to obtain in smaller prototype-level quantities. But there is a BIG advantage to using an LTC part: (usually) decent and Free (albeit not Open-Source) LTspice simulation! [Tron900] seems to take advantage of this (in a simple way). Unfortunately, as of my post-time [Tron900] does NOT share the LTspice simulation files with us on his hackaday.io project page (put sadface here). I would ask [Tron900] for these files to be included, but alas it seems I have to sign-up for a separate hackaday.io account first to comment there (not gonna happen HaD!)

        1. I was talking about Dave doing weird things in his design, not Tron900.
          LT spice is a good tool. But don’t forget you can use it with other parts, the libs are plenty. Plus, many manufacturers offer spice models for their parts, even though they are not the easiest thing to find.
          I have used LT spice with external models to do the topology reviews from here http://www.electrobob.com/digital-power-supply-part-3-concept-comparison/

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.