DIY Adjustable Bench Power Supply Build





Introduction: DIY Adjustable Bench Power Supply Build

About: Electronics, metalwork, machining and tinkering

I've been using an old power supply based off of a linear regulator for many years now, but the 15V-3A maximum output, coupled with the inaccurate analogue displays pushed me to make my own power supply which addresses these issues.

I looked at other power supplies that people have made for inspiration and decided upon some basic requirements:

-More Power than the old analogue one could deliver

-Cooling Fan (if necessary)

-Digital Display

-Sleek Looking and Safe (not that the analogue one wasn't either of these things....)

For the electronics, all items were sourced from eBay or from a skip outside of my college (seriously) so the bill of materials is rather difficult to determine. I estimate that I spent less than €12 in parts, but this will be higher if you cannot obtain some parts (power source) for free, where the price is very much dependant on the power output you desire.

Please note that this 'ible focuses on my power supply build and so not all steps are in a how-to style, but more-so a synopsis of the steps taken. If more detail is required I am more than happy to help of course, just leave a comment here or on the demonstration video on youtube and I will reply ASAP :)

Step 1: The Power Electronics

The power source used was a high current (8A) SMPS (Switch-Mode-Power-Supply) that outputs 19V, which I fortunately obtained for free. Similar power sources that could be used include a laptop charger or even a transformer with a full bridge rectifier circuit.

To stop power being drawn when not used, the Live connection was extended to a switch at the front panel of the case, and back to the SMPS. Since the case is metal, I connected the Earth pin to the base plate with a screw.

The D.C. output of the SMPS was connected to a step down DCDC Buck converter, the output of which went to the positive and negative connections on the front panel of the case (via the shunt resistor on the digital display).

The digital display, along with a 5V buck converter (for the USB ports) was powered by the 19V SMPS, as this would stay constant no matter what the output voltage was set to.

A 24V computer fan was also connected to the SMPS via a MOSFET circuit, which limits the current (and thus the speed) of the fan.
NOTE: The current limiting circuit is not necessary and the MOSFET is just acting as a resistor. It was added to reduce the speed of the fan and many other circuits (even an LM317 based circuit) would probably work better than my implementation, but I can include it if someone does want it.

Step 2: Control Electronics and Display Wiring

The digital display meter needs to be wired in series with the negative output terminal to sense the current and another wire goes to the positive output terminal to measure the output voltage, as shown in the above picture.

To adjust the output voltage, a 50kOhm trimmer pot on the 15A buck converter is replaced by a similar rated single turn potentiometer that's extended to the front case by a ribbon cable. One side of the potentiometer is connected to a 2kOhm potentiometer in an attempt to have a "fine tune" voltage knob but as discussed later, this is rarely used.

An inherent flaw with using a buck converter is that the output voltage is limited to roughly 1V less than the input voltage, but the potentiometer resistance is matched to the maximum input voltage (in this case max. input voltage = 30V). This means that if you supply the buck converter with a voltage well below the maximum input voltage, the potentiometer will have a dead zone - where turning the knob doesn't change the voltage. To overcome this, there are two options:

1) Use a combined Buck/Boost Converter which either steps up or steps down the input voltage to whatever is desired - this option would be best for having a large output voltage range that is independent from (not limited by) the input voltage.

2) Choose a potentiometer with a resistance that reduces the dead zone to an acceptable level - this is the cheapest option but only reduces the dead zone (which increases the resolution as a result) so the output voltage is still capped to a certain amount under the input voltage.

I went with option 2 as I already had a 15A buck converter and didn't want to wait for more parts to arrive from China. As the required potentiometer resistance wasn't near a standard value, I put a resistor across the outer terminals of the potentiometer, effectively reducing the resistance to the desired value.

Step 3: The Case

Now for the fun and tedious part - making the case. You could use anything you want for this; wood, MDF, plastic, metal, or completely 3D printed if you really wanted. I went with metal and plastic as I'm most comfortable with these materials and they look nice together (sorry wood enthusiasts).

I had a good amount of stainless steel sheet material so the main cover was made with this. The front and back panels were made out of plastic (acrylic at the front, unknown chewy plastic at the back) and the base plate was made from a sheet of steel from a TV stand.

The base was cut to be slightly wider and much longer than the SMPS and holes were drilled in the 4 corners where the SMPS case fasteners used to be located (as the top half of the case was removed for wires and better heat dissipation).

These holes were tapped with an M4 tap so machine screws could be used to secure the SMPS to the base, along with stainless steel right angle plates which are used to connect the base to the stainless steel cover and the rear panel. Two similar holes were drilled and tapped to hold the front panel in place with a plastic right angle piece being used this time (due to the proximity of power connections).

The front and back panels were marked out and drilled where required, then the pieces were cut and hand filed to dimension, including the rectangular holes for the display, USB ports and the mains power connection at the back.

The main cover was marked out on 0.8mm SS sheeting and cut to size with an angle grinder, including a port on the side for an air intake. Holes for the side and the top were marked and drilled before bending, but since I don't have a sheet metal brake (yet) the bends I managed to get had a severe radius to them. As I calculated for a smaller radius for the holes, I hammered the edges against some angle iron in a vice to get everything to line up properly - this introduces some "character" into the piece and makes sure everyone knows it's bespoke...

Everything is assembled with M4 machine screws, or glue for parts that wont need to be replaced. I think it's important to build things with serviceability in mind.

Step 4: Review

After assembling, testing and using for several months, I discovered the 2K potentiometer for the "fine tune" function was noisy (goes open circuit occasionally when turning). This was unacceptable as it made the output voltage jump unexpectedly, and so I simply turned the 2k pot to its minimum position so it doesn't interfere with the main adjustment pot. High quality potentiometers are a must for projects like these.

I hope this helps some of you out there as other 'ibles helped me. This is just one approach of many and I encourage questions if any additional information is needed, either here or on my youtube video. Thank you so much and well done if you have made it this far, happy making!



    • Creative Misuse Contest

      Creative Misuse Contest
    • Clocks Contest

      Clocks Contest
    • Water Contest

      Water Contest

    3 Discussions

    Hello and thanks for that very neat instructables. I have a question, as I want to build one of these variable power supply myself : what would you say are the pros of the linear type over the smps + buck/boost type ?

    I am asking this because it seems to me that 90% of the diy power supply I see on internet are linear ones and I can't really figure out why.

    1 reply

    Hi Salvagelt, thank you for your comment :)

    There are a few advantages to linear regulators which are:

    -Low to no noise, buck converters are noisy as they switch current through an inductor but linear regulators are basically a resistor so its very low noise.

    -Low cost and low complexity, linear regulators are notably cheaper and are very simple compared to the buck converter. You could make an entire power supply from scratch with the linear regulator and support components but making your own buck converter is far more complex (still doable but not worth the effort compared to buying a module)

    The lower noise of the linear regulators is the most advantageous, but there are multiple disadvantages to linear regulators also:

    -Low efficiency, high heat loss, most notably where the input voltage is significantly higher than the output voltage (efficiency = Vout/Vin. Of course, if the input voltage was always slightly higher than the output then it would have high efficiency, but this isn't the case fora variable power supply)
    Whereas buck converters can be >90% efficient for a wide voltage range.

    -Input voltage must be greater than output voltage, you cannot step-up voltage with these regulators, only step-down. (buck-boost converters can do both)

    -Usually limited to lower power due to the low efficiency and the large heatsink associated with dissipating that wasted power.

    I hope this helps you in some way, feel free to ask for any additional information :)