35 V 5 A Linear Adjustable Power Supply Kit

To be honest, I ordered this kit mostly to fill out the package, but it turned out to be quite useful, especially for beginner electronics hobbyists. A while ago I reviewed a simple adjustable power supply, and as it turned out, it was useful. Now imagine roughly the same power supply, but:
For a higher voltage
For a higher current
With transformer winding switching
With fan control

Interested? Then I don't think you'll regret it.


I'll start today's review by telling you about the seller—or, more precisely, by mentioning that I accidentally realized this is already my fourth review of their products. The previous ones may ring a bell too; they covered:
1. LCR meter
2. Simple oscilloscope
3. Electronic load

So I can recommend ordering several items from this seller at once. The load + power supply combination is especially good value.


The intermediary ships everything together in one package. According to their information, the kit weighs 175 grams; weight matters when ordering from Taobao.


You should receive a circuit board and a large bag of components. A box isn't included; it's just there to give you an idea of the size :)


As with the electronic load, no schematic is included. All the information needed to assemble it is printed on the board as silkscreen. The values of every component are marked, so assembly shouldn't be a problem.


The board is entirely through-hole, with no SMD components, which I think may be important for a beginner electronics hobbyist.


The silkscreen quality is very good: the printing is sharp and everything is easy to read.


The layout, however, isn't very optimal. The footprints for the power transistors are at the edge of the board, and the transformer connector is there too, so one of them will have to be wired to the board. I'll come back to that later.


There are four kit options:
1. Complete kit: components plus board. This is the one I got; it costs about $8.64
2. Same as above, but without the pair of output transistors; about $7.76
3. All the components, but no circuit board; about $6.73
4. Board only, no components; about $1.9.

Since there are quite a lot of components, I'd recommend the first option. But some of the components aren't very good quality (the capacitors, for example), so option 4 might also make sense. In my opinion, options 2 and 3 aren't particularly worthwhile.


And here's one downside of Taobao: my kit was missing the potentiometer knobs. They cost next to nothing, but it's still a shame :(


The product page includes the power supply schematic, which can also help with assembly. I did have to refer to it a couple of times, but I'll cover the details in the assembly section. The schematic isn't very high resolution. The seller offers an “HD” version, but I couldn't figure out how to download it.

The schematic doesn't contain anything fundamentally new: one op-amp forms the power supply itself, the other controls the winding switch, and the fan power control circuit is at the bottom. The “odd” op-amp supply is a little concerning, as is the center-tapped winding used to power the internal electronics, which serves no purpose in this case.
The potentiometers are also connected in a slightly unusual way, with two wires. Increasing the voltage/current corresponds to increasing the resistance of the potentiometer.


The power supply's main sections:
1. Green — the adjustable voltage and current regulator itself, the low-current section, and the power supply circuit
2. Red — the regulator's power section, rectifiers, and relays
3. Blue — the relay control circuit for switching the windings
4. Purple — fan control.


I'll skip the preamble and move on to assembly. Since the process description is more of an add-on, I'll hide this section under a spoiler.


The board is now fully assembled. Based on the build, I can say that there weren't any particular problems, but the board itself looks a little... unattractive. It just doesn't have that polished look.


It would also have been better to put the connectors along the edge of the board rather than in the middle. One minor drawback: it turned out that the power supply output is connected by soldering, not with a terminal block.



After soldering, it's best to wash off the flux—not so much because it could affect the electronics, but for the sake of appearance. If you like, you can then coat the board with Plastik-70 varnish.


The board is very easy to solder. I used flux-core solder and a basic temperature-controlled soldering iron.


This appears to be a photo of the prototype, found on the product page. The layout is simpler, but the heatsinks are noticeably larger.


So, I have some wires, output transistors, a bridge rectifier, and a few other small parts left over.


Now for connecting and adjusting the board.
1. 0-15-25-35 V — power transformer connections. Voltages are measured relative to the 0 point.
2. Bridge rectifier and transistors—self-explanatory.
3. 25 V and 35 V relay adjustment: set the voltage at which the corresponding additional windings are connected.
4. Temperature adjustment and thermistor: adjust the fan turn-on temperature and connect the thermistor. Thermistor polarity doesn't matter.
5. 12-15 V — input for an additional 12-15 V AC supply; one winding can be used.
6. Ammeter power — connect the power supply for the ammeter used to measure output current; regulated 12 V.
7. Fan — fan connector.
8. Current range — set the output current adjustment range.
9. Current set — adjust the output current (10 kΩ resistor).
10. LED CC — current-limit mode indicator LED.
11. Voltage range — set the output voltage adjustment range.
12. Voltage set — adjust the output voltage (10 kΩ resistor).
13. Output — output pads for connecting a load to the power supply.
14. Ammeter — ammeter connection; if unused, short it with a jumper.

Now for the adjustments.
Winding switching voltage.
1. Turn the resistors all the way to the left, or close to it—alternatively, until both relays switch off.
2. Set the output voltage to about 9-10 V, then turn the 25 V resistor to the right until the first relay switches on.
3. Set the output voltage to about 20-22 V, then turn the 35 V resistor to the right until the second relay switches on.
4. That's it.

Output voltage/current adjustment range.
1. Turn the voltage adjustment resistor all the way to the right.
2. Turn the corresponding trimmer until you get the desired output voltage, for example 35 V.
3. Do the same for the current adjustment. You can use a multimeter as the load.

To increase the current, turn the trimmer to the left; to increase the voltage, turn it to the right.

Fan turn-on.
1. With a load connected, heat the heatsink until it's hot enough to burn your hand—about 50-55 degrees.
2. Turn the resistor to the left until the fan switches on. You can raise the temperature to 60-70 degrees, but use a thermometer to measure it.
By the way, the fan is controlled by a fairly powerful transistor, which seems to have been chosen mainly because of its large package. The fan has a primitive control circuit, so it doesn't switch cleanly on and off: the transition is gradual, and it can run at low speed. However, the temperature range from off to full power is quite narrow.



If your transformer has only two windings—for example, one from an amplifier power supply with a pair of 18 V windings and a center tap—you can use that too, although it will of course run hotter. In this case, replace the second relay with a jumper.


Connect the two left terminals of the potentiometers together, then connect each potentiometer using two wires.
The thermistor also uses a two-wire connection. After soldering it, insulate the joint with heat-shrink tubing.
The input for the additional power supply is designed for a winding with a center tap, which I find extremely inconvenient. You can connect the two outer connector pins together and power it from a single 12-15 V winding; it will work just as well.


I didn't use the fan or ammeter/voltmeter wires. I twisted the rest together before soldering to keep things neat and reduce interference. The black heat-shrink tubing was included.


A brief aside: there’s a spot on the board for a bridge rectifier, but at 5 A it would quickly overheat, so I decided to mount it off the board. That’s why this photo shows not only the transistors, but also the bridge rectifier.
The TIP3055 transistors are rated for 15 A, 60 V and 90 W. In this power supply, each transistor carries 2.5 A, handles up to 50 V and dissipates up to 35–40 W, so there’s still a little headroom.


For testing, I used a relatively small heatsink. For actual use, you could easily use a CPU cooler from a reasonably powerful processor. Since the windings are switched, even in the worst-case condition (a short circuit), it will dissipate around 75–80 W, which is quite comparable to a processor.
The transistors are electrically isolated from the heatsink. If you don’t do this, the thermal resistance will be lower, but the heatsink will be at the positive rail of the power supply.


You could say we’re ready to test :)


For the tests, I used a fan with a three-pin connector. In this case, connect the red and black wires as shown in the photo.


The manufacturer posted an example of how to use the board with a somewhat uncommon but interesting ammeter-voltmeter on the product page. I couldn’t find one when I was writing this review, though. It apparently measures up to 5 A, and it’s reasonably priced.


On the other hand, I saw an equally interesting little meter from another seller. I’ve been wanting to buy one and play around with it for a while, especially since it measures current up to 10 A and voltage up to 95 V, and can connect to a computer for monitoring. But it costs 13 bucks — link .


Anyway, I’m getting carried away. I’m connecting the board to the tried-and-tested set of two transformers, plus a small one for the auxiliary supply. Together, the transformers provide three voltages in multiples of 12 V. By the way, the board manufacturer recommends 15+10+10 rather than 12+12+12, as I roughly mentioned in my review of the high-power adjustable power supply. That combination of voltages is more optimal.


Now let’s see what this little board can do.
1. The minimum setting is −0.1 V. Yes, negative voltage—I’ve come across this before.
2. The maximum is 21 V with the range trimmer set to minimum.
3. Next, I tried adjusting the maximum voltage with the trimmer, but got only 26 V—not much.
4. At first I thought I’d solder in some resistors to test it, but then I remembered that increasing the resistance of the adjustment resistor increases the voltage or current setting. So I simply pulled out the connector and got the full output without any trouble.
5. The minimum current is 0 A. The CC indicator LED is lit, and the power supply’s output resistor acts as the load.
6. There were no calibration issues here; I set it to 5 A.


Then I turned the trimmer further and, again without any trouble, got 6 A.


But I wasn’t happy with the output voltage limit, and needed to do something about it. I suspected the auxiliary supply, so I measured the transformer output and found it was only 11 V. I tried another transformer with an output of around 24 V, and easily got as much as 42 V at the output.
The auxiliary voltage is regulated by a 12 V regulator, which needs at least 15 V at its input. There’s also a 15 V zener diode on the board. But with an input of 11 V, it’s difficult to get more than 15–16 V, so the voltage was sagging.


After that, I wanted to check the maximum output power this setup could deliver. But about 20 seconds into the test, there was a loud bang, and I got this little surprise....
Yes, when I replaced the transformer, I completely forgot about these capacitors, so the result was entirely predictable: they had about 32 V across them.


But “the show must go on,” so I replaced the casualties with more reputable Samwha 1000 µF, 35 V capacitors.


In the end, I got over 200 W at the output, with a load current of 5 A and a voltage of 41 V. I think that’s pretty good.


Next, I’ll test how well the output voltage stays stable as the load current changes. The results here are also pretty good, although the voltage did fluctuate a little. This may have been due to the contact between the load and the board, since the load was connected to the multimeter probes, which were simply inserted into the holes in the board.
Test at 1, 2, 3.5 and 5 A.


The board gets noticeably warm during operation. The power resistors get the hottest.
1. At low voltages, the auxiliary power resistors heat up. They are connected in series with the 6.2 V and 15 V Zener diodes. The one nearest the edge of the board gets especially hot; it supplies the 6.2 V Zener diode.
2. If you set the output voltage above 20–30 V, the 2.2 kΩ resistors in the upper-right corner start getting very hot. The temperature of one depends on the output voltage, while the other depends on the input voltage, which reaches its maximum when the output is above 20–22 V. I think it would be better to replace them with something around 3.3–4.7 kΩ.

In both cases, the resistors reach around 100–110 °C.


And finally, a test to measure the output ripple. Unfortunately, there is some, at a frequency of 100 Hz. In both cases the load was around 4 A (an automotive bulb), but in the first test only the original input capacitors were installed. In the second, I connected another 10,000 µF capacitor in parallel with them, using wires about 10 cm long.
The ripple was 50 mV in the first case and 25 mV in the second.


In my view, the output ripple is caused less by insufficient input capacitance—which I think is perfectly adequate here—and more by the somewhat odd feedback circuit (marked in red).
I also don't like that there's a full 100 µF capacitor on the output (marked in green). I think it would be better to reduce it to 10–22 µF. It doesn't really affect the ripple, but it does affect the current surge when switching from CV mode to CC mode.


Video version of the review


And, of course, a few conclusions based on the assembly and test results.
First, a few words about the kit itself.
There aren't many complaints, but there are some. The resistor knobs were missing, the insulating pads are awkward to use, the bridge rectifier needs to be mounted on a heatsink, and the capacitors are mediocre quality.
But there are advantages too: everything goes together without much trouble, and it actually works, delivering even more than the specified 35 V at 5 A. I was able to get up to 42 V and 6 A, and I don't think that's the limit.

Based on the test results, the only real criticism is the high ripple level, but I think there's a chance it can be improved.

Overall, the kit is a little rough around the edges, but in my opinion it's more interesting than the well-known 30 V, 3 A board, which I reviewed some time ago. The key differences are:
1. Output voltage up to 35 V, and it can actually be increased further.
2. Current up to 5 A, which can also be increased.
3. Input capacitor capacity of 6,600 µF, compared with 3,300 µF in the 3 A version.
4. The 3 A power supply had one power transistor; this one has two.
5. The transformer windings can be switched between three stages.
6. Fan control based on temperature has been added.
7. The current-sense shunt is on the positive rail, not the ground rail.

There is only one significant drawback: the reviewed version has higher ripple, most likely due to shortcomings in the circuit design.


Product on AliExpress - link