Hackaday Prize Entry: Scorpion DC-DC Voltage Converter

Finding the right wall wart or charger to go with an appliance might be a matter of convenience to you and I, but there are some people who really, really need the right charger, because not having it could mean a fire.

[marius] is a Romanian hardware engineer who moved to Papua New Guinea, where he had the opportunity to travel in the remote jungle of that country. There, he saw many people who used solar panels to charge car batteries for a 3W light bulb at night, their phones, or other conveniences that only need a few Watt-hours a day. Connecting car batteries directly to solar panels isn’t a smart idea, so [marius] set out to create a simple, very low-cost DC-DC voltage converter. He’s calling it the Scorpion 3.0, and it looks like a fantastic tool for low-income areas that are far off the grid.

The design of the Scorpion consists of a 3D printed enclosure, with one forked end containing some alligator clip leads, and a standard barrel jack on the other. In the middle is a character display showing the input and output voltage, and a simple rotary encoder for user interaction. The circuit for the Scorpion 3.0 consists mostly of a cheap, low-power MSP430 microcontroller managing the display, encoder, and a buck converter.

Designing something for off-the-grid usage means a few engineering challenges, and being in Paupa New Guinea, there are a few environmental considerations as well. [marius] is varnishing his 3D prints. No, it’s not going to be IP68 rated, but it helps. Making the Scorpion cheap is also a big consideration, most probably resulting in the choice of the MSP430.

It’s a great project, and an excellent entry to the Hackaday Prize. You can check out the demo video of the Scorpion below.

32 thoughts on “Hackaday Prize Entry: Scorpion DC-DC Voltage Converter

    1. Tip 1:
      Replace the ‘nl’ with ‘www’ and you’ll get an understandable language (note i’m not from an english speaking country, so don’t start with the USA snarks)

      Tip2:
      remove all the crap after and including the ‘?’

    1. The voltage is probably all over the place. I’d imagine on a cloudy or foggy day (and with 80% humidity they’re probably common,) you’d have to lay several cells out in series to get a decent 9-12V. The cells would be rated to 12V, but solar panels’ power output changes by like an order of magnitude between bright sunlight and overcast weather.

      I was recently designing a boost converter with mini-cells, and it was annoying because in practice, even with an ultra-low-dropout converter, reliably hitting a range of 0.8 – 5.5V was genuinely difficult without involving something else, like something to toggle a transistor with 1ohm to ground when the input goes overvoltage. If you use a 5V cell, you really only get 0.25-0.5V indoors, which means no power generated. If you string some in series, you might fry the boost chip with like 15V when you step outside. If you hooked something like that right up to a battery, I could see it causing strain and damage the cell at the very least.

      Or I dunno, maybe the electron pixies get angry when there aren’t enough capacitors nearby. I’m just spitballing.

      1. The load is attached to the battery, which is an approximate constant voltage source. Any fluctuations in solar output are reflected in the charge current, not voltage. The behaviour of the solar cell differs open versus closed circuit.

    2. It strains the PVs. The real trouble comes when you link multiple panels or battery banks together. It protects the PVs at night and keeps the batteries safe during the day. For the short term it’s not a problem but it can lessen the life span of the system over the long haul.

      1. ‘Strains’ how? Pseudoscience nonsense. The panels will, provided their output is greater than vbatt, match their output to vbatt without any issue. No life shortened anywhere, just works. Some loss of overall efficiency due to panel not operating at maximum power point.
        If you have no money, you can probably live with that minor compromise.

        1. You should tell the manufacturers it’s pseudoscience. They’ll stop putting in bypass diodes in their big panels and selling charge controllers.
          Like I said above, a single panel or low voltage system isn’t at risk, but once you start stringing lots of panels together or using higher voltage battery banks you need to provide some protection to both the battery and the panels to prevent current from damaging them.

          1. You’re a troll. The scorpion isn’t running multi kW installs, it is being used in situations where there are one or two old car batteries – not even leisure batteries – and a couple of panels. The currents involved are low, and even if they weren’t it is hard to see a situation where the bypass diodes are going to be “strained”.
            Fact is, the charge controller part of the scorpion, whilst nice to have, is far from a necessity in the kind of environments where it is going to be deployed.
            Source: lived off grid for years, qualified EE.

    3. If the panel isn’t too big, and the battery is of a decent size and the usual flooded-cell lead-acid type, it’s not that bad to connect the solar panel directly.

      Probably the biggest risk is overcharging. yes, that itty-bitty solar panel can overcharge that big old battery. In the case of an overcharged flooded-cell battery, the electrolyte will boil off, but a slight loss can be restored by topping up the battery with distilled water.

      Most cheap small solar panels that I’ve come across have a built-in diode, so they won’t discharge the battery at night. If not, spend a dime and add a diode.

      I’ve used a small portable solar panel direct-connected to the battery for over 4 years now on our small boat, with no problems. If I see the voltage is over 15v, I disconnect the panel.

      A charge controller, or even a voltage monitor that kicks out at around 15v will prevent overcharging.

      1. This. There’s a whole market of 5 – 7W cells + diode that you can just attach to your storage cell and it’s just fine. I’ve run my sailboat battery like this for more than a decade – with the same farm-store marine battery. With the combination of solar charging and LED running lights (the only real use other than the stereo) the only maintenance I have to do is pull the battery in winter to prevent freezing and top up the cells’ water mid-summer.

  1. You’re making the thing almost as wide as a typical car battery so why not go one step further and design the board with two holes on each end the size of battery terminals and a piece of bent metal soldered to the board for each hole, and then they could get one of these devices and push it straight down over the battery terminals and be done…?

    1. There’s no such thing as a typical car battery for dimensions. There’s dozens of sizes that exist and also they like to swap polarity sometimes. clips of some kind are the best way to deal with it.

          1. I would agree except that he’s already in the ballpark in terms of size. Why not go the extra step and make it even easier for those who can use it. You can still provide the wires if you want, or else provide them as an add-on, or up to the user to provide. Or even just leave some exposed copper on either end and people can use ready-made jumper cables which they might already have. Any or all of which would probably add in functionality / ease of use plenty to offset any slight difference in pcb. If you rethought the shape a little more, made it wider and thinner, you might even reduce the overall square size of the pcb and reduce parts count a little to boot.

    2. This device aims to be a Swiss Army Knife of voltage conversion: small, rugged, flexible. It can be used to directly power stuff, or to be a battery charger. So, I think the author has made reasonable choices here. About the only thing I’d want to add is a bank of standard USB jacks, since USB is used so often for power.

      Also, let’s not underrate the importance of good physical design. I kind of like the ‘scorpion’ concept – it’s a practical size and shape, and if this idea takes off, it will be a convenient identifier…”Hey, my phone’s dead; you gotta scorpion I could use?”

  2. The basic idea seems reasonably useful, but the actual implementation leaves a lot to be desired.

    30V on the input will wreck the MOSFET because it violates the maximum gate-source voltage.

    There is no hardware current limit for the inductor current, only the output voltage can be measured. A hard short circuit on the output could easily destroy the MOSFET, especially if a battery is used as the supply, instead of a power supply that will at least limit the input current (the inductor current will still be excessive).

    Without a way to measure input voltage and current, I don’t see any options to implement proper maximum power point tracking; I thought this was meant to charge a battery from a solar panel, how can you not have MPPT?

    I think requiring the user to set the proper voltage is a no-go. A USB socket should always be limited to 5V, and with the current design, it’s very easy to accidentally connect the USB socket while the other leads are still connected to the battery. The battery should have a dedicated connection, and I think I would have used a separate cable for each device that can be connected, with a way to detect which cable was connected and automatically select the proper voltage.

    Right now, the user must disconnect a device, change the voltage setting to the correct value, and only then connect another device. This is too easy to screw up, possibly leading to the destruction of the device being used or charged.

    1. These are very reasonable suggestions and not that hard to implement. A low side shunt resistor for current measurement wouldn’t hurt and the rest is software stuff that is easy to implement. Let’s give the po’ man a break and let him take these suggestions into account for his v4. I havent paid too much attention but the board layout didnt look terrible at first glance.

      1. It’s an overegineered case with underegineered electronics. Besides, there are cheap, adjustable buck converters with both current limiting and current sense, some even include cases. And they are cheaper than CD-printed case for this fancy toy…

  3. Quite a few design issues there… A quick glance… 10k pull up on a mosfet gate on a smpsu, no. Gate capacitance needs a bit of current to charge and discharge if you want reasonable switching speed.. An LM317, while available, can be a bit of a current hog. Not really suitable for low power design. A tranzorb and a small capacitor at the input is a good idea. The input series diode is a large source of power loss, there are other ways of doing it, like this one… http://opend.co.za/hardware/200ds232/index.html
    (full disclosure: one of my designs)

Leave a Reply to LeithoaCancel 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.