WZ5005 Adjustable Voltage Converter: Could It Be an Alternative to DPS Converters?
The converter didn't arrive on its own; it came bundled with a 48 V, 350 W Mean Well power supply, which was bought specifically to work with this converter. I'll tell you about the power supply another time, though, since today we're looking only at the converter—although I'll have to mention the power supply along the way.
I bought the converter on Taobao. This AliExpress link goes to one listing with a reasonable price. There's another link too; it's a little more expensive, but has more reviews.
The converter's packaging is somewhat reminiscent of the popular RD60xx series converters.

The converter is usually sold in four configurations:
1. Converter + wires
2. Converter + wires + UART-to-USB module
3. Converter + wires + RS485 module
3. Converter + wires + WiFi module
Prices on AliExpress usually range from 25–26 for the basic configuration to 30–31 for the version with a WiFi or RS485 module.
I have the simplest version: just the converter and wires for connecting a load.

You can find the full specifications for the converter, as well as download the software and manual, at this link.
It looks very similar to the DPS5005, but is noticeably larger.

The included wires are about a meter long, with banana plugs on one end and alligator clips on the other. The wire gauge isn't specified, but the cable is marked 29×0.15, which presumably means 29 strands, each 0.15 mm².
The soldering is decent, and the alligator clips are pretty good too. To check the wire gauge, I measured the resistance of the wires—conveniently, they're just the right length. I got 35 mΩ per meter, equivalent to a wire with a cross-sectional area of 0.5 mm². So the 0.15 in the marking refers to the strand diameter, not its cross-sectional area.
The wires are flexible, but I would have preferred a larger cross-section.

They didn't even include a manual or any kind of leaflet, so let's move on to the converter.

The case is designed for panel mounting, just like the RD60xx converters.

Dimensions:
external length and width — 120×74 mm
panel cutout length and width — 116×70 mm
case depth — 35 mm
overall depth — 65 mm

The key differences from the DPS/DPH are the 1.8-inch display instead of a 1.44-inch one, many more controls, and much more convenient buttons.
The display is at the top of the front panel. To its right are buttons for setting the voltage, current, protection thresholds, and output enable.
In the middle is a joystick for moving the cursor up, down, left, and right, plus an OK button. These buttons are also assigned to the memory slots. To the right is a push-button encoder; press it to enter the menu.
Below are a pair of terminals for connecting a load and a button to turn the converter itself on.
The display is recessed quite far into the case, but what caught my eye more was the bezel design. It gives the impression that this is the screen of a KVN-49 television.

The inside also bears a noticeable resemblance to the RD60xx. The design consists of two boards: a power board and a control board.

1. The lower board has two connectors for USB/WiFi/RS485 modules. Below them are contact pads with VCC/GND/SWDIO/SWCLK lines routed to them.
2. There's no input EMI filter, but, as on the RD60xx, there is a fuse, a capacitor, and reverse-polarity protection.
3. There are two input voltage converters. The main one is built around an XL7015, with an input voltage rating of up to 80 V, which gives plenty of headroom for the specified 55 V input.
4. A pair of power components is covered by the heatsink, with a relatively small inductor on the left.

The modules connect to a PC in much the same way as the RD60xx. The only difference is that WiFi uses one connector and the other interfaces use another. There is one detail, though: the UART-USB model has a bracket for mounting on the front panel and what looks like a USB-C connector.

The PWM controller and its surrounding components.
Unfortunately, I couldn't identify the controller. I checked the ones I know, but almost all of them have ground on pin 12, while this one has it on pin 11; pins 3 and 5 are also connected to ground.

1. Three op-amps, a voltage regulator, and a transistor that is most likely in the current generator circuit, in parallel with the output. You can see a capacitor by the upper-left op-amp that was clearly soldered on later. Also, the voltage regulator is in an unfortunate spot, right opposite the latch.
2. At the output, there are two 25 mΩ shunts connected in parallel, a 15 A fuse, and more signs of rework. There are also solder splashes on the board and even on one of the electrolytic capacitors.
3, 4. The input and output have 330 µF, 63 V capacitors made by Gemcon: one at the input and three at the output. Two are before the shunt and one is after it.

I removed the upper board after all the tests so as not to compromise their accuracy. The procedure is already familiar: desolder the power terminals, pull them out, then remove the board.

The underside of the upper board is empty, apart from the connector to the control board.

By the time I was taking the oscilloscope measurements, I already knew it used synchronous rectification; you could tell from the shape of the output ripple. But I was a little surprised to find paired transistors. I was also surprised by how little thermal paste there was—there was barely any. I had to apply my usual KPT-8 afterward.
Four N-channel RU7570L transistors are installed, rated for a maximum voltage of 70 V with an on-resistance of 9 mΩ.
There is also a temperature sensor under the heatsink.

There aren't many components on the control board: a couple of voltage regulators, a 24C02 EEPROM, and an STM32F103 microcontroller.
Part of the reason there are so few components is that the USB module is separate, so there is no galvanic isolation or converter here.

The module really does look a lot like the RD60xx. I'd even call it a little brother: even the encoder has the same knob and the same little dots around it.

The enclosure also looks like it was designed on the same principles as the boards inside. I wouldn't be surprised if it was developed by the same people who made the RD60xx.

For the initial tests, I used the same Mean Well power supply shown at the beginning, but raised the voltage to the recommended 55 V. The converter initially powered up with the output enabled, though that may have been how the person who sent it to me had set it.
Even the power button works like the one on the RD60xx: while the converter is in standby, its brightness smoothly cycles from 0 to 100%.

1, 2. On startup, the logo appears, then the converter switches to operating mode.
3. The menu is in Chinese by default, but it's easy enough to find the language setting.
4. There are two language options: Chinese and English.
5, 6. Besides the usual display, there's also an option to show the readings as a graph. But the manufacturer wanted to be different and made the “waveform” move from left to right instead of right to left, which immediately catches your eye.
7. I immediately turned off the annoying beeper that confirms every click, as well as auto power-on. The short beep on startup remains, though.
8. This is also where you select the connection mode—wired or WiFi—and set the baud rate and address.
9. Brightness has 7 levels, from 0 to 6. The backlight is quite bright, and the display is still easy to see even at the lowest setting.

The adjustments work in the usual way: press the corresponding button to enter the fine-step setting mode, then move through the steps to set the value. There are a few quirks, though. The power buttons are at the top right, while the left/right navigation buttons are at the bottom left, which is a little awkward. It gets more annoying because the converter exits setting mode after about 30 seconds. So you adjust the voltage, get distracted for a minute, want to tweak it some more, and have to press and select everything all over again...
1. Voltage setting
2. Current setting
3, 4. Overvoltage or overcurrent protection settings.
There are no questions about setting overcurrent protection: it can be handy, since the protection is triggered without switching the converter to CC mode. Overvoltage protection, on the other hand, is needed much less often, though I suppose it could be useful in some situations too. In any case, it's a useful feature.

1. There is also a display mode for accumulated capacity in Ah, energy in Wh, and input voltage. Use the up/down buttons to switch to it; holding down a memory button activates the corresponding memory slot.
2. I've already written about the graph display mode, but here's a little more detail. It doesn't have auto-scaling: the vertical scale for voltage and current is shown at the bottom. By default, it's 2 V and 200 mA per division, so if you increase the values beyond that, the graph becomes useless.
3. Use the left/right buttons to select the parameter, then turn the encoder to change its value so the graph fits on screen.
4. Pressing the voltage, current, or protection setting buttons brings up the corresponding menu on the right, just like in normal mode. You can close it by pressing the button again, or it will disappear on its own after half a minute.

The converter has a fixed minimum difference between the input and output voltages, so you can't set the output higher than the limit allows.
1, 2. For example, with an input of 50.76 V, it let me set the output to 47.16 V, while at 53.75 V, the maximum was 50.02 V. That's a difference of 3.6–3.7 V.
3. While I was adjusting the settings, the converter unexpectedly shut off. More precisely, it shut off because the Mean Well power supply tripped its protection, which seemed odd.
4, 5. If the input voltage is lower than the set output voltage—for example, if the input was higher when you set it but then dropped—the converter will shut off when you try to activate the output. The output voltage drops very gradually, though; I was able to take a photo of the process without any trouble.
6. Turning the encoder down doesn't reset the voltage to the maximum limit for the current input, so you need to turn the encoder left or right to make it recalculate the limit for the voltage currently applied. After that, the converter works correctly.

When switched off, the output simply disconnects, with no load from the converter, so the voltage drops very gradually.

When the output is active, the voltage changes quickly in either direction. That would be good, except that the Mean Well power supply ends up tripping its protection. Since it doesn't restart, unlike after an overcurrent, and its latched protections are for overheating and overvoltage, there's probably some kind of spike coming from the converter.
The encoder works well: nothing skips, and you can change the settings very quickly.

While testing the converter, I noticed that with no load and the output active, the current shown on the display wasn't zero. It would sometimes read 2, 3, or even 6 mA.
There's a zero-adjustment option in the menu, which I haven't seen on other converters. It was probably added to compensate for a circuit design flaw.
After the automatic correction, everything was back to normal.

A few measurements.
Test 1: checking the accuracy of output voltage setting and measurement.
I set the converter output to 0.1, 0.5, 1, 5, 10, 20, 30, and 51 V. I wouldn't call the results perfect, but they're fine.
Note that you can't set the voltage below 0.1 V, even though the adjustment step is 0.01 V. The output is either 0 or 0.1 V and above.
The test was performed with an input voltage of 55 V.

Test 2: checking the accuracy of current setting and measurement.
The load here was the multimeter's shunt and the wires connected to it. I set the current to 1, 2, 5, 10, 20, 50, 100, 200, and 500 mA, then 1, 2, and 5.1 A.
The accuracy is worse than for voltage setting, but it's fine for an inexpensive converter.

Test 3: output voltage regulation under varying load.
As always, I compared the output voltage under a light load (about 50 mA) and a nearly full load (5 A here).
I could say the results are good, and even excellent in some modes, if it weren't for a few “quirks.”

The problem occurred at all voltages, but was more noticeable at 20–30 V and above.
1. At 36 V with no load, everything looks good; the voltage barely changes.
2, 3. Under load, it started fluctuating between about 35.995 and 36.040 V. The difference isn't large, but I don't like that it's happening at all.
4, 5. Tested close to the maximum operating point. Again, you could call it excellent, but the “wobble” was still there, even though the photo happened to capture similar readings.
6. The power supply worked perfectly with a 4 V difference between input and output, and its efficiency seemed pretty good too.

Test 4: output ripple from the converter.
As always, I started by testing with no load, then at 33, 66 and 100%. The average voltage was 25 V, since ripple is usually close to its maximum at this setting.
Measurements were taken with 1 µF and 0.1 µF capacitors connected in parallel with the probe. I can say that everything looks quite respectable: at maximum current, the peak-to-peak ripple was 75–80 mV.

Here, the current is the same in every mode—5 A—but the output voltage varies: 5, 12 and 48 V.
During the previous measurements, I also noticed some voltage “wobble,” but it was more pronounced here, so I checked it at different timebase settings.
1, 2. 5 V, 5 A
3, 4. 12 V, 5 A
5, 6. 48 V, 5 A.
As you can see, the voltage is actually fluctuating over quite a wide range. This is what caused the voltage variations under load that I mentioned earlier.

And the usual tests:
1, 2. Applying and removing 5 V at the output. When switched off, the voltage drops very gradually.
3, 4. Response to connecting/disconnecting a load. The output was set to 12 V, with a 30 Ω resistor connected continuously and a 3.3 Ω resistor connected in parallel with it (or disconnected). So this shows the response to a current step from 0.4 to 4 A and from 4 to 0.4 A.
5. Transition from CV to CC mode: 15 V, load current 4.5 A, current limit set to 3 A.
6. Same load and settings, but in the opposite direction: transition from CC to CV.
The first two cases look fine: no overshoot. There’s none when powering up the module or switching it on from standby either, which is good.
The third and fourth cases are also pretty good, with a small overshoot of about 0.5–0.6 V when the load is removed.
But the transition from CC to CV really surprised me. I even had to change the vertical scale because the waveform simply wouldn’t fit on the screen. With the output set to 15 V, the overshoot reached 28 V—almost double, which is already dangerous.

“Soldering iron test.” A TS100 soldering iron is connected as the load to check how the power supply responds to continuous changes in load current.
Interestingly, the results here are very good—better than with the RD60xx series converters.

Test 5: efficiency.
The test was run at a load current of 5 A, an input voltage of 55 V, and output voltages of 5, 12, 24, 36 and 48 V.
The resulting efficiencies were 86.8, 93.1, 95.3, 96.7 and 97.4%, respectively, which is pretty good. No-load current at 48 V output was 32 mA, and just 16 mA with the output disabled.

There were no overheating issues either. I ran it for about 20 minutes at first, then for another half hour or so. In the end, the heatsink temperature was 53–54 °C, making it the hottest component. Both the inductor and the heatsink were comfortable to touch, so there’s still some thermal headroom. This is one area where the RD6006 module surprises me, since it has a fan-cooled heatsink.

I tested the software only in its desktop version, connecting through a UART-to-USB converter from one of my electronic loads. Since there’s no Wi-Fi module, I didn’t try the smartphone app.
Well, what can I say? The software is almost completely dull, and everything is so primitive it’s hard to recognize. RuiDeng can sleep easy for now :)

On the plus side, I found a reproducible shutdown bug. During testing, I used an RD6018 power supply, but for the efficiency measurements I switched to an RD6006 to improve accuracy. If, for example, you set the output voltage to 50 V in the software and then immediately change it to 5 V, the RD6006 almost always (about 90% of the time) trips its output overvoltage protection.

I connected a multimeter to see what it would show in max/min mode. An oscilloscope would have been more appropriate here, but by then I’d already cleared it off the desk.
On the left is one of the 10% of cases where the transition from 50 to 5 V works correctly: the spike reached only 63.36 V. The next attempt produced a larger spike, and the RD6006 shut down.

Here’s what it looks like over time: first I enable the output at 5 V, then raise it to 50 V, and finally lower it back to 5 V.

I have one theory about why this happens, and the likely culprit is the synchronous rectifier and its incorrect control. If we look at the step-down converter schematic (on the left), we see a classic step-down converter. But if we mirror it, we get an equally classic step-up converter. In other words, So when the output voltage drops sharply, the lower transistor keeps switching, acting as the power switch in the step-up converter. The upper transistor does turn off, but its parasitic diode remains in the circuit.
As a result, all the energy stored in the output capacitors flows through the boost converter to the module output and, consequently, to the output of the power supply feeding the whole device. And since Mean Well has overvoltage protection, it trips.
The simplest solution seems to be installing a diode between the power supply and the converter, but you absolutely must not do that, because the voltage spike will be even higher: there will be nothing to absorb its energy. The voltage could easily exceed 80 V, followed by a failure of the PWM controller in the power supply circuit and the need to buy a new converter.
You could install a TVS diode or some kind of shunt voltage regulator, but in my opinion it’s easier to increase the capacitance of the power supply’s output capacitors. Also bear in mind that the energy fed back can be higher if the converter is powering a load with a large capacitance.
As I was writing this, it occurred to me that the problem could be even worse if a battery is connected to the converter output...
In that case, probably the only fix is to install a diode between the DC-DC output and the output terminals.

When the power is cut, the converter keeps running for a while and continues sending voltage and current readings, then it shuts down and communication is lost.

And now, as always, the verdict.
I’m not even sure what to say. At first, I was going to praise it—the converter itself is pretty good. I especially liked its efficiency, low ripple, and stable operation. But there are a lot of “caveats.”
1. Output voltage “wobble” under load.
2. A very large voltage spike when switching from CC to CV.
3. An input voltage spike when the output voltage drops sharply.
4. Controls that aren’t very convenient
5. Poor build quality in places, soldering residue, and no thermal paste.
6. Software that’s too basic
7. Add your own reason.
In my opinion, although the converter is inexpensive and probably worth the money, it’s still very much a work in progress. I think the manufacturer should seriously consider making some improvements.
That’s all for now. I hope you found this useful.
инструкция на wz5020, корпус для wz5005 модель для печати, wz5012l нагрузки на выходе нет а уходит в защиту, схема wz3605e, wz5005,, форум wz3605e, преобразователь dc dc wz 5020l инструкция, https://www.kirich.blog/obzory/preobrazovateli/965-reguliruemyy-preobrazovatel-napryazheniya-wz5005-neuzheli-alternativa-preobrazovatelyam-dps, wz5012l инструкция по эксплуатации, wz5020l schematic, в wz3605e заданное и текущее значение напряжения не совпадают, dps5005 kirich, разъемы wuzhi wz5005, wz6012, wz5020l зарядное устройство, wuzhi wz5005 блок питания с али, преобразователь напряжения wz3605e отзывы, wz5005 прошивка, wuzhi wz5020, wz5005l инструкция на русском языке, инструкция wz5020l, dps 500c програм. пониж. преодраз. цена харак. отзывы видио покупателей, пресс wz схема, wz3605e схема, модуль wz3605 не. показывает. выходное напряжение
Questions and answers
How does the WZ5005 differ from the RuiDeng DPS5005?
The WZ5005 has a 1.8-inch display instead of 1.44 inches, more controls and more convenient buttons. It is also noticeably larger than the DPS5005.
What issues did the WZ5005 show during testing?
The author notes output voltage fluctuations under load, a strong overshoot when changing from CC to CV, and an input voltage spike when the output voltage is reduced abruptly. Soldering marks and too little thermal paste were also found.
What is the author's conclusion about the WZ5005?
It showed good efficiency, low ripple and stable operation, but the author considers it unfinished and says it needs further development.
This model is included in the summary table: Comparison of adjustable converters for a bench power supply
The figures from this review are in the measurement database: Converters
Discounted products
You may be interested in
Discounted products
All the content on the site and the advice are free, but we’d be grateful for your support of the project and the channel!
Comments5
Only registered readers can comment. It takes a minute to sign up, and you can log in with the same account you use on the forum.
Log inRegister




Comments are published in their original language.