Adjustable Power Supply Kit Board, or a Proper Power Supply Should Be Heavy

Many of you already know that I have a soft spot for all kinds of power supplies, so this is a two-in-one review. This time, I’ll be looking at a DIY kit for building the core of a bench power supply, along with an example of a finished build.
A heads-up: there’ll be lots of photos and text, so grab yourself some coffee :)

First, let me explain a little about what this is and what it’s for.
Pretty much every electronics hobbyist uses a bench power supply in their work. Whether it’s a complex model with software control or a very simple one based on an LM317, it does more or less the same job: powering different loads while you work on them.
Bench power supplies fall into three main types.
Switch-mode regulated.
Linear regulated.
Hybrid.

The first type uses a controlled switch-mode power supply, or simply a switch-mode power supply with a step-down PWM converter. I’ve already reviewed several versions of these power supplies. 1, 2, 3
Advantages: high power in a compact size and excellent efficiency.
Disadvantages: high-frequency ripple and large capacitors at the output.

The second type has no PWM converters at all; all regulation is done linearly, with excess energy simply dissipated by the regulating element.
Advantages: virtually no ripple and (almost) no need for output capacitors.
Disadvantages: efficiency, weight, and size.

The third type combines the first and second types: the linear regulator is powered by a step-down PWM converter. The PWM converter’s output voltage is always kept slightly above the final output voltage, and the rest is regulated by a transistor operating in linear mode.
Alternatively, it can be a linear power supply with a transformer that has several windings, which are switched as needed to reduce losses in the regulating element.
This design has just one drawback: it’s more complex than either of the first two options.

Today we’ll talk about the second type of power supply, with a regulating element operating in linear mode. But we’ll look at it through the example of a DIY kit; I think that should make things even more interesting. In my opinion, it’s a great starting point for a beginner electronics hobbyist: build one of the essential pieces of test equipment yourself.
Or, as they say, a proper power supply should be heavy :)

This review is aimed mostly at beginners; experienced folks probably won’t find anything useful in it.

I ordered a kit for this review that lets you build the main part of a bench power supply.
The main specifications are as follows (as listed by the seller):
Input voltage: 24 V AC
Adjustable output voltage: 0–30 V DC.
Adjustable output current: 2 mA–3 A
Output voltage ripple: 0.01%
Circuit board dimensions: 80 × 80 mm.

A little about the packaging.
The kit arrived in a regular plastic bag, wrapped in soft material.
All the necessary components, including the circuit board, were inside a resealable antistatic bag.


Everything was loose inside, but nothing was damaged; the circuit board provided some protection for the components.


I won’t list everything included in the kit; it’ll be easier to do that as we go through the review. I’ll just say that there was enough of everything, with a few parts left over.


A little about the circuit board.
The quality is excellent. There’s no schematic included, but all the component values are marked on the board.
The board is double-sided and covered with a solder mask.


The board finish, solder plating, and the quality of the PCB material itself are all excellent.
I only managed to lift a pad in one place, and that was after I tried to solder in a non-original part (more on why later).
In my opinion, this is just right for a beginner radio enthusiast; it would be hard to damage.


Before assembly, I drew the schematic for this power supply.


The circuit is fairly well thought out, though not without its flaws. I’ll discuss those as I go.
You can make out several main sections in the schematic; I’ve marked them in different colors.
Green — voltage adjustment and regulation section
Red — current adjustment and regulation section
Purple — indicator section for switching to constant-current mode
Blue — reference voltage source.
There are also:
1. Input bridge rectifier and filter capacitor
2. Power control section using transistors VT1 and VT2.
3. Protection circuit using transistor VT3, which disconnects the output until the op-amp supply voltage is within the normal range
4. Fan power regulator, built around a 7824 IC.
5. R16, R19, C6, C7, VD3, VD4, VD5: the circuit that generates the negative supply rail for the op-amps. Because this section is present, the power supply won’t work from DC alone; it requires AC input from a transformer.
6. C9 output capacitor and VD9 output protection diode.


First, I’ll go over the pros and cons of the circuit design.
Pros —
It’s nice to have a regulator for the fan, but the fan needs to be rated for 24 V.
The negative supply rail is a big plus; it greatly improves the power supply’s performance at currents and voltages close to zero.
Since the circuit has a negative supply rail, it also includes protection: until this voltage is present, the power supply output remains disconnected.
The power supply includes a 5.1 V reference voltage source. This not only allows accurate adjustment of the output voltage and current (with this design, both are adjusted linearly from zero to maximum, without “humps” or “dips” at the extremes), but also lets you control the power supply externally simply by changing the control voltage.
The output capacitor has a very small capacitance, which makes it safe to test LEDs: there’s no current surge while the output capacitor discharges and the power supply enters constant-current mode.
The output diode protects the power supply if reverse-polarity voltage is applied to its output. The diode is rather undersized, though, so it’s better to replace it.

Cons.
The current-sense shunt has too much resistance, so at a load current of 3 A it dissipates about 4.5 W as heat. The resistor is rated for 5 W, but it gets very hot.
The input bridge rectifier is made up of 3 A diodes. Ideally, it should use diodes rated for at least 5 A, since the current through the diodes in this circuit is 1.4 times the output current. That means they can carry 4.2 A in operation, even though they’re rated for only 3 A. The only mitigating factor is that the diode pairs in the bridge conduct alternately, but it’s still not quite right.
A major drawback is that when selecting the op-amps, the Chinese engineers chose ones with a maximum voltage rating of 36 V, but didn’t consider that the circuit has a negative voltage source. This limits the input voltage to 31 V (36-5=31). With an AC input of 24 V, the DC voltage will be around 32–33 V.
In other words, the op-amps will be operating beyond their rated limits (36 V is the absolute maximum; 30 V is the normal rating).

I’ll cover more pros and cons, as well as modifications, later. For now, let’s get on with the assembly.

First, lay out everything included in the kit. This makes assembly easier and gives you a clearer view of what’s already installed and what’s still left to do.


I recommend starting assembly with the lowest-profile components. If you install the tall ones first, it will be awkward to fit the shorter ones afterward.
It’s also best to start with the components that have the most identical parts.
I’ll start with the resistors, specifically the 10 kΩ ones.
These are good-quality resistors with 1% tolerance.
A few words about resistors. They use color bands to indicate their values, which may seem inconvenient to some. In fact, this is better than alphanumeric markings, because you can read the bands no matter how the resistor is positioned.
Don’t be put off by the color bands. At first, you can use online calculators, and with time you’ll be able to work them out without one.
To understand and work comfortably with these components, a beginner radio hobbyist only needs to remember two things that will come in handy.
1. The ten basic colors used in the markings
2. The values in the E24 series. They’re not very useful when working with precision E48 and E96 resistors, but those are much less common.
Any experienced radio hobbyist can list them from memory.
1, 1.1, 1.2, 1.3, 1.5, 1.6, 1.8, 2, 2.2, 2.4, 2.7, 3, 3.3, 3.6, 3.9, 4.3, 4.7, 5.1, 5.6, 6.2, 6.8, 7.5, 8.2, 9.1.
All the other values are these multiplied by 10, 100, and so on. For example, 22 kΩ, 360 kΩ, 39 Ω.
What’s the point of knowing this?
It means that if a resistor is from the E24 series, for example, this color combination—
blue + green + yellow—is impossible.
Blue — 6
Green — 5
Yellow — ×10000
So the calculation gives 650 kΩ, but that value isn’t in the E24 series; the nearest values are 620 or 680. That means either the color was misidentified, the color has faded, or the resistor isn’t from the E24 series. The last case is rare, though.

All right, enough theory. Let’s move on.
Before installation, I bend the resistor leads into shape, usually with tweezers, though some people use a small homemade jig for this.
Don’t throw away the clipped leads just yet; sometimes they come in handy for jumpers.


After installing most of them, I’ve reached the individual resistors.
This part may be trickier, since you’ll have to identify the values more often.


I don’t solder the components in place right away. I just clip the leads and bend them over—and I always clip them first, then bend them.
It’s very easy: hold the board in your left hand (if you’re right-handed), while pressing the component you’re installing in place.
Hold the side cutters in your right hand, clip the leads (sometimes several components at once), then use the side of the cutters to bend the leads over.
It’s all very quick, and after a while it becomes automatic.


And now we’ve reached the last small resistor. The value we need matches the one we have left—so far, so good :)


With the resistors installed, we move on to the diodes and Zener diodes.
There are four small diodes here, the popular 1N4148s, and two Zener diodes rated at 5.1 V each, so it’s hard to mix them up.
Bend their leads into shape as well.


The cathode is marked on the board with a stripe, just like on diodes and Zener diodes.


Although the board has a solder mask, I still recommend bending the leads so they don't cross any nearby traces. In the photo, the diode lead is bent away from the trace.


The Zener diodes are marked on the board just as they are on the components themselves: 5V1.


There aren't many ceramic capacitors in the circuit, but their markings can confuse a beginner. By the way, they also follow the E24 series.
The first two digits give the value in picofarads.
The third digit is the number of zeros to add to the value.
For example, 331 = 330 pF
101 = 100 pF
104 = 100000 pF, or 100 nF, or 0.1 µF
224 = 220000 pF, or 220 nF, or 0.22 µF


Most of the passive components are now installed.


Next, we install the op-amps.
I'd probably recommend getting sockets for them, but I soldered them directly in place.
The first pin is marked on the board, just as it is on the IC itself.
The remaining pins are numbered counterclockwise.
The photo shows the spot for the op-amp and how it should be installed.


I bend only a couple of the IC leads, usually the two diagonally opposite corner pins.
It's also best to trim them so they stick out about 1 mm above the board.


That's it. Now we can move on to soldering.
I use a basic temperature-controlled soldering iron, but an ordinary 25–30 W iron will do just fine.
1 mm solder wire with flux. I'm deliberately not naming the brand, since the spool isn't the original one (the original spools weigh 1 kg), and the name won't mean much to most people.


As I mentioned above, the board is well made and very easy to solder. I didn't use any flux; the flux in the solder is enough. You just need to remember to occasionally shake the excess flux off the tip.




Here I've included a photo showing an example of good soldering and not-so-good soldering.
A good solder joint should look like a small bead of solder enveloping the lead.
But there are a couple of spots in the photo where there's clearly too little solder. That might be fine on a plated-through, double-sided board (where solder also flows into the hole), but you shouldn't do this on a single-sided board; over time, the joint may come loose.


Transistor leads should also be pre-formed. Do this in a way that doesn't put stress on the lead near the base of the package (old-timers will remember the legendary KT315 transistors, whose leads were notorious for breaking off).
I form power components a little differently. Bend the leads so the component sits above the board. That way, less heat is transferred to the board and it won't damage it.


This is what the formed power resistors look like on the board.
All the components were soldered from the underside only. The solder you can see on the top of the board flowed through the holes by capillary action. Ideally, solder a little should flow through to the top as well; this makes the joints more reliable and helps heavy components stay in place.


Up to this point, I formed component leads with tweezers. For the diodes, you'll need small needle-nose pliers.
The leads are formed in much the same way as resistor leads.


But installation is different.
With components that have thin leads, you install them first and trim the leads afterward. With diodes, it's the other way around. You won't be able to bend the lead after trimming it, so bend it first, then cut off the excess.


The power stage is built around two transistors connected in a Darlington configuration.
One of the transistors is mounted on a small heatsink, preferably with thermal paste.
The kit included four M3 screws; one goes here.


A couple of photos of the almost-assembled board. I won't go into how to install the terminal blocks and the other components; it's pretty self-explanatory, and you can see it in the photos.
Speaking of terminal blocks, the board has them for the input, output, and fan power connections.




I haven't cleaned the board yet, although I often do that at this stage.
That's because there's still a little more work to do.


After the main assembly stage, we have the following components left.
A power transistor
Two potentiometers
Two board-mount connectors
Two connectors with wires. The wires are very flexible, by the way, but have a small cross-section.
Three screws.


The manufacturer originally intended the potentiometers to be mounted directly on the board, but they're positioned so awkwardly that I didn't even solder them in; I just showed them as an example.
They're very close together, so adjusting them would be extremely awkward, though not impossible.


But at least they included wires with connectors, which is much more convenient.
This way, the potentiometers can be mounted on the instrument's front panel, with the board installed wherever it's convenient.
I also soldered in the power transistor. It's a standard bipolar transistor, but with a maximum power dissipation of up to 100 W (provided it's mounted on a heatsink, of course).
There are three screws left. I couldn't figure out what they're for: if they're meant for the corners of the board, you need four; if they're for mounting the power transistor, they're too short. A mystery, all in all.


The board can be powered by any transformer with an output voltage of up to 22 V (the specs say 24 V, but I explained above why you shouldn't use that voltage).
I decided to use a transformer I'd had lying around for ages, intended for a Romantika amplifier. Why “intended for” and not “from”? Because it hasn't been used in anything yet :)
This transformer has two 21 V power secondary windings, two 16 V auxiliary windings, and a shielding winding.
The voltage is specified for a 220 V input, but since our standard is now 230 V, the output voltages will be a little higher too.
The transformer's rated power is about 100 W.
I connected the power secondary windings in parallel to get more current. I could have used a two-diode rectifier circuit, of course, but it wouldn't be any better, so I left it as is.


First test power-up. I fitted a small heatsink to the transistor, but it still got quite hot, since this is a linear power supply.
Adjusting the current and voltage was no problem; everything worked right away, so I can already recommend this kit.
The first photo shows voltage regulation, the second shows current regulation.


First, I checked the transformer's output after rectification, since this determines the maximum output voltage.
I got about 25 V—not much. The filter capacitor is 3300 µF. I'd recommend increasing its capacitance, but the unit works perfectly well as it is.


Since I needed a proper heatsink for further testing, I moved on to assembling the whole unit, as the heatsink installation depended on the planned layout.
I decided to use an Igloo7200 heatsink I had lying around. According to the manufacturer, it can dissipate up to 90 W of heat.


The unit will use a Z2A enclosure, apparently made in Poland, which costs about $3.


At first, I wanted to move away from the enclosure my readers have grown tired of—the one I use to build all sorts of electronic gadgets.
So I chose a slightly smaller enclosure and bought a fan with a grille for it, but I couldn't fit all the components inside. I ended up buying a second enclosure and, accordingly, a second fan.
In both cases, I bought Sunon fans. I really like this company's products, and both fans were rated for 24 V.


This is how I planned to mount the heatsink, board, and transformer. There's even a little room left for adding more components.
There was no way to fit the fan inside, so I decided to mount it externally.


Mark the mounting holes, tap the threads, and screw everything in place for a test fit.


The chosen enclosure has an internal height of 80 mm, and the board is the same size, so I mounted the heatsink to center the board relative to it.


The leads of the power transistor also need a little bending so they don't deform when the transistor is pressed against the heatsink.


A brief digression.
For some reason, the manufacturer designed the mounting area for a rather small heatsink. As a result, when you install a decent-sized one, the fan power regulator and its connector get in the way.
I had to desolder them and cover the area where they were with tape to prevent contact with the heatsink, since it carries voltage.


I trimmed off the excess tape on the back; otherwise it looked rather untidy. We'll do it properly, Feng Shui style :)


This is what the circuit board looks like with the heatsink fully installed. The transistor is mounted with thermal paste, and it's best to use a good-quality paste, since the transistor dissipates about as much power as a high-performance CPU—around 90 W.
While I was at it, I drilled a hole for mounting the fan speed controller board, though I ended up having to drill it again anyway :)


Some of you probably noticed the trimmer on the board and wondered what it's for.
It's there because the circuit design of this power supply is done properly.
Since the op-amps have a dual supply, the output needs to be zeroed with the voltage control turned fully counterclockwise.
At its end positions, the trimmer sets the output voltage to between -0.138 and 0.15 V.
The negative voltage sneaks through parasitically and has a very low current.


To zero the output, I turned both controls fully counterclockwise, disconnected the load, and set the output to zero. The output voltage can now be adjusted from zero.


Next, a few tests.
I checked how accurately it maintains the output voltage.
No load, voltage 10.00 V
1. Load current 1 A, voltage 10.00 V
2. Load current 2 A, voltage 9.99 V
3. Load current 3 A, voltage 9.98 V.
4. Load current 3.97 A, voltage 9.97 V.
These are pretty good results. If you want, you can improve them a little more by changing where the voltage feedback resistors connect, but I think this is good enough.


I also checked the ripple level. The test was performed at a current of 3 A and an output voltage of 10 V


The ripple level was about 15 mV, which is very good. Though I thought that the ripple shown in the screenshot was actually more likely coming from the electronic load than from the power supply itself.


After that, I started assembling the device as a whole.
I began by installing the heatsink with the power supply board.
To do this, I marked out the positions for the fan and the power input connector.
The opening wasn't marked as a perfect circle; it had small “flats” at the top and bottom to make the rear panel stronger after cutting the hole.
Holes with complex shapes are usually the most difficult, for example, one for a power connector.


A large hole is cut from a big pile of little ones :)
A drill and a 1 mm bit can sometimes work wonders.
Drill holes. Lots of holes. It might seem slow and tedious, but it's actually very quick: drilling the entire panel takes about 3 minutes.


After that, I usually switch to a slightly larger bit, say 1.2–1.3 mm, and use it like a milling bit. This is the slot you get:


Then we take a small knife and clean up the holes, trimming the plastic a little if the opening turned out slightly too small. The plastic is fairly soft, so it's easy to work with.


As the final preparation step, we drill the mounting holes. The main work on the rear panel is now done.


Install the heatsink with the board and the fan, check how everything fits, and if necessary, “finish it off with a file.”


Near the very beginning, I mentioned making some modifications.
I won't be doing much.
First, I decided to replace the original diodes in the input bridge rectifier with Schottky diodes. I bought four 31DQ06 diodes for this. And here I repeated the board designers' mistake: out of habit, I bought diodes rated for the same current, when I should have gone with higher-rated ones. Still, the diodes will run cooler because Schottky diodes have a lower voltage drop than regular ones.
Next, I decided to replace the shunt. I wasn't happy not only with how hot it got, but also with the fact that about 1.5 V was dropped across it—voltage that could be put to use (that is, delivered to the load). For this, I used two locally made 0.27 Ω, 1% resistors (which should also improve stability). Why the designers didn't do this is unclear; it would cost exactly the same as the original 0.47 Ω resistor.
And, more as an extra, I decided to replace the original 3300 µF filter capacitor with a higher-quality, higher-capacitance 10000 µF Capxon…


This is what the finished assembly looks like, with the components replaced and the fan temperature-control board installed.
It turned out a bit makeshift, and I also accidentally lifted a pad on the board while installing the power resistors. I could easily have used lower-power resistors—for example, a single 2 W resistor—but I didn't have one on hand.


A few more components were added underneath.
A 3.9 kΩ resistor connected in parallel with the outer contacts of the connector for the current-adjustment resistor. It's needed to reduce the adjustment voltage, since the voltage across the shunt is different now.
Two 0.22 µF capacitors: one in parallel with the output of the current-adjustment resistor to reduce interference, and the other simply across the power supply output. The second one isn't really needed; I just happened to take out a pair and decided to use both.


The entire power section is wired up. I also mounted a board with a bridge rectifier and capacitor on the transformer to power the voltage display.
Strictly speaking, this board isn't necessary in the current setup, but I couldn't bring myself to power the display at its 30 V limit, so I decided to use an additional 16 V winding.


The following components were used for the front panel:
Binding posts for connecting the load
A pair of metal knobs
Power switch
Red filter, listed as a light filter for KM35 enclosures
For the current and voltage displays, I decided to use a board left over from one of my reviews. But the displays were too small, so I bought larger ones with 14 mm-high digits and made a PCB for them.

This is only a temporary solution, but I wanted it to look neat even temporarily.


Several steps in preparing the front panel.
1. Draw a full-size layout of the front panel (I use the regular Sprint Layout). The advantage of using identical enclosures is that preparing a new panel is very easy, since the required dimensions are already known.
Place the printout on the front panel and drill 1 mm pilot holes at the corners of the square/rectangular cutouts. Use the same drill bit to mark the centers of the other holes.
2. Use the holes as a guide to mark the cut lines. Switch to a thin cutting disc.
3. Cut the straight lines, following the dimensions precisely from the front and cutting a little farther from the back so the cuts go all the way through.
4. Break out the cut pieces of plastic. I usually keep them, since they may still come in handy.


As with the rear panel, clean up the resulting holes with a knife.
For large-diameter holes, I recommend using a step drill bit; it won't grab the plastic.


Test-fit everything and, if necessary, make adjustments with a needle file.
I had to widen the hole for the switch a little.


As I mentioned above, I decided to use a board left over from one of my previous reviews for the display. This is actually a very poor solution, but it's more than adequate as a temporary setup. I'll explain why later.
Desolder the displays and connectors from the board, then check the pinouts of the old and new displays with a multimeter.
I wrote down the pinouts of both displays so I wouldn't get them mixed up.
The original design used four-digit displays; I used three-digit ones because I couldn't fit anything larger in the window. But since the fourth digit is only needed to display the letter A or U, losing it isn't critical.
I placed the current-limit mode indicator LED between the displays.


I get everything ready and desolder the 50 mΩ resistor from the old board. As before, it will serve as the current-sense shunt.
The problem is with this shunt. In this setup, the output voltage will drop by 50 mV for every 1 A of load current.
There are two ways to solve this: use separate meters for current and voltage, and power the voltmeter from a separate power supply.
The second option is to install the shunt in the positive rail of the power supply. Neither option suited my temporary setup, so I decided to swallow my perfectionism and go with a simplified, though far from ideal, solution.


For the assembly, I used mounting standoffs left over from the DC-DC converter board.
They made for a very convenient assembly: the display board attaches to the ammeter-voltmeter board, which in turn attaches to the power terminal board.
It turned out even better than I expected :)
I also mounted the current-sense shunt on the power terminal board.


The finished front panel assembly.


Then I remembered that I’d forgotten to install a more powerful protection diode, so I had to solder it in afterward. I used a diode left over from replacing the diodes in the board’s input bridge.
Of course, ideally I should add a fuse too, but that will have to wait for another version.


I decided to use better current and voltage adjustment potentiometers than the ones supplied by the manufacturer.
The originals are quite good quality and turn smoothly, but they’re ordinary potentiometers, and in my opinion a bench power supply should allow finer adjustment of the output voltage and current.
When I was considering ordering the power supply board, I spotted multi-turn potentiometers in the store and ordered some for review as well, especially since they had the same resistance value.


I usually use different potentiometers for this kind of thing. They combine two resistors in one, for coarse and fine adjustment, but lately I haven’t been able to find them for sale.
Does anyone know of any imported equivalents?


The potentiometers are quite good quality, with a rotation of 3600 degrees—or, more simply, 10 full turns—which gives an adjustment range of 3 V or 0.3 A per turn.
With these potentiometers, adjustment is about 11 times finer than with ordinary ones.


The new potentiometers next to the originals—their size is certainly impressive.
While I was at it, I shortened the wires to the potentiometers a little, which should improve noise immunity.


I fitted everything into the case. There’s even a little room left, so there’s room to grow :)


I connected the shielding winding to the connector’s ground wire. The auxiliary power board sits right on the transformer terminals. It’s not very neat, of course, but I haven’t come up with another solution yet.


Testing after assembly. Everything worked almost on the first try. I accidentally swapped two digits on the display and spent a long time trying to figure out what was wrong with the adjustment. After switching them, everything worked as it should.


The final step was gluing in the filter, installing the knobs, and assembling the case.
The filter has a thinner edge around its perimeter. The main part sits recessed in the case opening, while the thinner part is attached with double-sided tape.
The knobs were originally designed for a 6.3 mm shaft diameter, if I remember correctly. The shafts on the new potentiometers are thinner, so I had to put a couple of layers of heat-shrink tubing on them.
For now, I decided not to label the front panel, for two reasons:
1. The controls are so intuitive that labels don't seem particularly necessary yet.
2. I plan to make further modifications to this power supply, so the front-panel design may change.


A couple of photos of the finished build.
Front view:


Rear view.
Attentive readers have probably noticed that the fan is installed to blow hot air out of the case rather than push cool air between the heatsink fins.
I decided to do it this way because the heatsink is slightly shorter than the case. To keep hot air from getting inside, I installed the fan the other way around. This does reduce cooling efficiency, of course, but it also provides some ventilation inside the power supply.
I would also recommend making a few holes in the bottom half of the case, but that's more of an optional extra.


After all the modifications, the current came out a little lower than in the original version, at about 3.35 A.


So, I'll try to list the pros and cons of this board.
Pros
Excellent build quality.
The circuit design is almost right.
A complete set of parts for assembling the power supply regulator board.
Well suited to beginner electronics enthusiasts.
In its basic configuration, it only needs a transformer and heatsink; the more complete version also needs a voltmeter/ammeter.
Works right after assembly, albeit with a few caveats.
No large capacitors at the power supply output, making it safe for testing LEDs and the like.

Cons
The wrong type of op-amps was chosen, so the input voltage range has to be limited to 22 V.
The current-sense resistor value isn't quite right. It operates within its normal thermal range, but it's better to replace it because it gets very hot and could damage nearby components.
The input bridge rectifier is operating at its limit; it's better to replace the diodes with higher-power ones.

My take: while assembling it, I got the impression that two different people had designed the circuit. One used the right approach to regulation, a reference voltage source, a negative voltage supply, and protection. The other chose the wrong shunt, op-amps, and bridge rectifier for the job.
I really liked the circuit design. In the modifications section, I initially planned to replace the op-amps and even bought ICs rated for a maximum operating voltage of 40 V, but then decided not to modify it. Otherwise, though, it's a pretty good design, with smooth, linear adjustment. Of course it gets hot—that's unavoidable. Overall, I think this is a very good and useful kit for a beginner electronics enthusiast.
I'm sure some people will say it's easier to buy a ready-made unit, but I think building one yourself is more interesting (probably the most important thing) and more useful. Besides, many people probably have a transformer, a heatsink from an old CPU, and some kind of box lying around at home.

As I was writing this review, I became even more convinced that it would be the first in a series of reviews about linear power supplies. I have a few ideas for modifications:
1. Convert the display and control circuitry to digital, possibly with a computer connection.
2. Replace the op-amps with high-voltage ones (I haven't decided which yet).
3. After replacing the op-amps, I want to add two automatically switched ranges and extend the output voltage range.
4. Change how the display unit measures current so that the voltage doesn't sag under load.
5. Add a button to turn off the output voltage.

That's probably all for now. I may think of something else and add it later, but for now I'm looking forward to your questions in the comments.
I also plan to do a few more reviews of kits for beginner electronics enthusiasts. Maybe someone has suggestions for particular kits.


The product was provided by the store for review.