DC-DC converter, as sometimes happens.

Today I’m going to write not only about the product I tested, but also about how sometimes you plan one thing and, for some reason, end up with something completely different.
Anyway, if you’re interested, read on.

Recently, my colleague ksiman posted a review of the “half” version of this converter—the same little board, but without the display. So in a way, these reviews complement each other.
In the comments, I mentioned that I was also planning to review this board. The review said that things didn’t end very well (or rather, they ended very badly). Things didn’t go entirely smoothly for me either, although the outcome was better. More on that later; for now, let’s take a look at my version of this DC-DC converter.

Anyway, I spotted this little DC-DC converter and wanted to get my hands on it and see what it was like. I ordered one for review and received it some time later, but somehow never got around to checking it out, so I put it aside for a while.

Eventually, I finally got around to it, took a bunch of photos, and had a good look at it.
It arrived in a small sealed bag.


It’s quite small, smaller than a matchbox.
The manufacturer lists the following specifications:
Input voltage: 5V-30V
Output voltage: 0.8V-29V
Output current: 5A maximum (a heatsink is required for currents above 3A)
Conversion efficiency: 95% (maximum)
Switching frequency: 300KHz
Output ripple: 50mV (maximum)
Operating temperature: -40℃ to +85℃
Size: 51 x 26.3 x 114


There are connectors on the sides for connecting the power supply and the load.
The assembly is neat; I have nothing bad to say about it.


There are two trimmer potentiometers on top: one adjusts the current, and the other adjusts the voltage.
The current can be adjusted from 0.06 to 5.5 A.
The voltage range is 0.82 to 30 V.
There’s also a red LED next to the trimmers that indicates when the converter switches to constant-current mode.


The back of the board is practically bare; the only component there is a 50 mΩ shunt resistor.
By the way, it’s worth noting that in devices of this type, where heat from the IC is transferred to the board, it’s standard practice to use lots of plated through-holes connecting the two sides of the board to improve heat transfer. Unfortunately, that hasn’t been done here, so attaching a heatsink to the back is ineffective.


As I mentioned above, the converter consists of two boards. The DC-DC converter itself is no different from the one in the review I mentioned earlier. The difference between the two versions is that mine has a display board attached.
It’s connected using mounting standoffs.
The two on the left are the converter board’s input; the ones on the right are its output.
This connection lets you monitor the output voltage and measure the current flowing through it.
The design is quite convenient and simple.


The converter is built around an XL4005E1 PWM controller. It’s rated for an output current of 5 A and an input voltage of up to 32 V.
The datasheet makes it look like a pretty good IC, but in practice it turned out to be quite “delicate.”
It’s also worth mentioning the SK86 diode. According to the datasheet, it’s rated for a maximum current of 8 A. To be honest, I don’t understand how it can dissipate the power generated at that current.
In any case, the manufacturer chose a fairly powerful diode; quite often they use something worse.


This photo shows the section responsible for adjusting the current limit and indicating when charging is complete (you can see two small LEDs on the right).



There are two indicators at the top.
The upper one, blue, displays the output voltage. Up to 10 V, it uses the format 1.23; above 10 V, it shows 23.4. The last digit displays the symbol — V
The lower indicator, red, displays the output current in the format 1.23, with the last digit displaying the symbol — A.
There is an RX-TX connector on the left. That was one of the reasons I ordered this board: I wanted to try connecting it to a computer, but unfortunately, I couldn't get it to work :(
I have no idea what the connector on the right is for.


The board is, let's say, just about passable. It looks okay, but there's clearly some sloppy workmanship.


The board is fitted with:
Microcontroller 8s003f3p6
Shift register 74hc164 for controlling the display
Presumably an sgm8592y op-amp
7130H voltage regulator


Now for a small caveat. This is the second board; the first died a heroic death during testing and while I was preparing the review. I can't say exactly what killed it, but this is what happened: the input voltage was around 28–29 V, with a 10 Ω resistor connected to the output. I gradually increased the voltage across the resistor using the board's trimmer, then there was a small click and the input voltage appeared at the output—the power transistor had failed.
It may have been a defective board, or perhaps some ripple or something else, but I wouldn't recommend pushing the input voltage too high, even though the datasheet specifies 32 V, with 35 V as the maximum.
It's better to keep it below 25–27 V.
After that, I ordered a second board, since I'd already done quite a lot of work preparing the review.

When first powered up, the board is set to an output voltage of about 5 V and a current of about 1 A.
In the photo, the board is connected to a 24 V power supply from my recent review.
If you turn the voltage adjustment trimmer all the way up, the no-load output voltage equals the input voltage.


There's not much else to say about the board, so let's move on to testing.
The following equipment will be used for testing:
The board under review.
24 V power supply.
Non-contact thermometer
Oscilloscope
Electronic load
Pen and paper :)


The test procedure was as follows:
I measured the temperature rise and output voltage ripple at the following voltage settings: 5, 10, 15, and 20 V. At each voltage, I set the load current to 1, 2, and 3 A.
First, I measured the performance at 5 V with currents of 1, 2, and 3 A, at 10-minute intervals. Then I let the board cool to room temperature and repeated the cycle at the next voltage. That made 12 measurements in total.
The multiplexed display made things more difficult: I had to take a ton of photos and then pick one that showed as many digits as possible. The display's refresh rate is quite low, so there's a slight but noticeable flicker.
First, a no-load check: there is practically no ripple.
The oscilloscope probe is set to 1:1.





The entire test cycle took about 3.5 hours.
The temperatures recorded:
I monitored the temperature of the PWM controller, diode, inductor, and output capacitor.
When testing it, I decided to try 3 A, as stated on the store page. I figured if I burned it out, so be it—I’d have a couple of these lying around. But the experiment showed that the converter handled it, and the chip didn’t go into protection. The PWM controller reached a maximum temperature of 110.2 degrees.



So, to sum up:
Pros.
Neat construction—not perfect, but quite good.
The converter passed testing at currents up to 3 A, although temperatures were high.
Current and voltage measurement accuracy is quite good, with no particular complaints.
Low ripple; the maximum recorded was about 60 mV at a switching frequency of 300 kHz.
Compact design.

Cons.
Gets very hot at currents above 2–2.5 A.
Be careful not to exceed the input voltage, or install a protection TVS diode on the input.
The inductor is wound with thin wire.

In my opinion, it can be used quite normally at currents up to 2 A. I was a little disappointed that I couldn’t figure out the RF/TX signals. The converter can be modified without too much effort: rewind the inductor with thicker wire and about 1.5 times fewer turns, or replace it with a more powerful one (that’s better). Replace the diode with a more powerful one, and better still, move it to the back of the board. That would improve the thermal performance.
The claimed 95% efficiency is probably unattainable, but I think the actual figure is somewhere close, with the important caveat that it depends on the operating conditions. At 3 A, the board dissipated about 4 W of heat (roughly estimated), which means very low efficiency at a 5 V output. Efficiency gradually improves as the output voltage rises, although a step-down converter shouldn’t have such a steep dependence.
So, what can I say? I spent money on parts and a lot of time assembling the power supply board and putting everything together, but in the end I got a power supply with these specs:
Output voltage: 0.85–24 V.
Output current: 0.06–2.25 A.
Not much, but it’s still useful. I simply didn’t need to build such a powerful power supply.

I hope the information I’ve provided was useful.

The product was provided by the store for review.