DPS-5020 Adjustable Voltage Converter and Enclosure

Today I have an adjustable voltage converter on my workbench with an output voltage of up to 50 V and a current of up to 20 A. As far as I know, this is the most powerful converter in RuiDeng's product lineup. What's more, its specifications are very similar to those of the ZXY-6020 I reviewed earlier, which I'll be comparing it with today.
While I'm at it, I'd like to say a big thank-you to one of my regular readers, without whom this review probably wouldn't have happened.


This is far from my first review of this board—or of this company's products in general—and not even my tenth, so I doubt I'll have anything radically new to say. Still, I'll do my best to share my impressions, especially in comparison with a module from another company. Think of this as a small addition to the reviews already out there. :)

In my previous review, I complained that the converter arrived in a simple little box. This time, though, I got the "full package": the large foam box familiar to other users, with a separate package inside containing the converter and its enclosure.


The converter and all its accessories are packed separately. I have to give the manufacturer credit for designing proper packaging that really does protect the product on its way to the customer.
The packaging also includes information about the product range and brief specifications for the other models.


Specifications for the entire RUIDENG converter lineup


This model is sold in three configurations:
1. Converter only
2. Converter + USB adapter for connecting to a computer
3. Converter + USB + Bluetooth adapter.

This review covers the last and most complete configuration. In my opinion, though, the manufacturer should have made an adapter board that lets you connect both adapters at once. Otherwise, including two types of adapter makes little sense, since you'll only use one or the other. In my review of the ZXY6020 power supply with its board, I showed how to connect both adapters at once, although the USB adapter there was a little different.


1. The package included a pretty good instruction manual in Chinese and English, with explanatory illustrations as well as text.
2. It also came with a couple of ribbon cables for connecting the modules and fork terminals for connecting wires to the power board.
3, 4. I covered the USB and Bluetooth boards in more detail in my review of the 8005 board, so here I'll just show what they look like.
In short, we have a Bluetooth board and an RS232 TTL-to-USB board. Both have the same connector for connecting to the converter, and the connecting cables are included. The USB adapter board also has galvanic isolation; for obvious reasons, the Bluetooth module doesn't need it. :) As far as I can tell, all these modules use standard RX/TX signals plus 5 V power.


The front-panel module looks similar—actually, it's virtually identical in appearance to the DPS8005 module—so I'll reuse some information from my previous review.


The front has the control buttons and encoder; the back has two connectors for connecting to the power board.

Controls
1. Three buttons
Up — moves through the menu, changes a parameter, and enables voltage adjustment mode. Press and hold to activate the first memory slot.
SET — enters/exits settings mode. Press and hold to display memory slot M0 and reset the settings to their default values.
Down — moves through the menu, changes a parameter, and enables current adjustment mode. Press and hold to activate the second memory slot.

2. Display for showing parameters and the control menu.
3. Encoder for adjusting parameters. The encoder is push-button operated: pressing it changes the adjustment step size (on the main screen) or activates a menu item (in settings mode). Press and hold to lock/unlock the controls.
4. Output enable button


Let's remove the back cover. We'll have to do that anyway, since it gets in the way when connecting one of the connectors. Inside, we find just a bare board with connectors marked LCD and KEY: the first is for the display, the second for the keypad.
The ZXY6020 had a slightly different layout, with the power board separate from the "brains" with the buttons and display. Unfortunately, in this case I have to give ZXY the credit: it is much harder to disrupt its operation with interference. With this unit, I’ve seen the display and controls malfunction when the connection wires run alongside power cables, as induced interference disrupts the low-level control signals.


And this is the converter itself. As far as I can tell, everything is here: the power stage and the “brains.” Only the controls and display are external.


The board is very compact.


Compared with the ZXY6020 kit, it looks tiny. Part of the reason is that the ZXY uses a large passive heatsink, while the DPS has active cooling. Even so, the difference is substantial.


One of the board’s long sides has connectors for the display and control module, as well as a small connector for a Bluetooth module or USB converter.


1. The input terminals are decent enough, but you can’t properly clamp a bare wire in them. You’ll have to use the included “fork” terminals, otherwise the wire gets pushed out.
2. The input has three 470 µF, 63 V capacitors. In my opinion, their voltage rating is cutting it close, especially given the pulsed nature of the load. If you use the board at the full input voltage of 60 V, I’d replace them with something more suitable rated for 80 or 100 V.
3. The output has three capacitors and a current-sense shunt made from three wires. There are two 470 µF capacitors before the shunt and one 330 µF capacitor after it, all rated for 63 V. Here, though, they operate at up to 50 V, so there’s more headroom. The total output capacitance is 1270 µF, which imposes some limitations when working with “delicate” loads such as LEDs. It’s not recommended to connect this kind of load in CC mode with the output voltage set much higher than expected. For example, if an LED module is designed for a 10 V drop, don’t set it above 11 V. Better still, connect it first and then gradually increase the voltage.
4. The output has a storage inductor wound with three fairly thick wires. It’s smaller than the one used in the ZXY, partly because that unit was rated for over 1200 W, while this one is rated for 1000 W.


The input voltage regulator powers all the board’s electronics, including the microcontroller.


The heatsink is quite small, with a fan mounted on top. The fan is controlled automatically according to the temperature beneath the heatsink; there’s a temperature sensor on the board for this purpose.


As with the ZXY, there’s nothing interesting on the underside.


1. I don’t understand this part. There’s a pair of MOSFETs at the input. The only thing I can think of is reverse-polarity protection on the input, but why is it needed here...
2. The low-voltage section is powered by XL7005A. It’s a PWM controller rated for input voltages up to 80 V, but honestly, I don’t like this circuit. In power supplies based on ZXY boards, I added a zener diode in series with the supply to reduce the voltage reaching the auxiliary power supply circuit. This reduces the lower end of the input voltage range, but improves the reliability of the power supply. Something like a 1.5KE10 should do.
3. A small 3.3 V regulator, presumably to power the microcontroller.
4. And here’s the microcontroller itself. There seems to be a small op-amp on the underside.


Let’s see what’s under the heatsink. All you need to do is remove two screws and disconnect the fan’s power connector. By the way, the mounting screws weren’t very tight, so I recommend checking them before putting the unit into service.
It would also be a good idea to replace the thermal pad with thermal paste, but unfortunately that requires the transistors to be exactly coplanar, which isn’t always the case. Besides, the whole approach of drawing heat from the transistor packages is flawed, since the material they’re made of is a poor thermal conductor.


Under the heatsink are four power transistors and a few other components.


The power stage uses a synchronous rectifier topology: two transistors act as the input switch, while the other pair replaces the diode. Circuit design - step-down converter with synchronous rectification.
This design significantly improves efficiency, and ZXY definitely falls short in this respect, since its design is much simpler.
All four transistors AOD2810 are 80 V N-channel MOSFETs with an on-resistance of 8.5 mΩ and a current rating of up to 46 A.
The temperature sensor marked RT1 is visible in the center.


1. The main PWM controller is a TL594, an improved version of the well-known TL494. It's a little strange to see such an old chip in a fairly new device, since these chips are outdated and there are much more interesting alternatives available.
2. Next to the PWM controller is a dual rail-to-rail op-amp, the SGM8582. I haven't fully analyzed the circuit of this converter, but the output voltage (through a divider, of course) is fed to this op-amp, so I assume it's in the current and voltage measurement circuit used by the microcontroller.
3. Judging by the circuit, this is a MOSFET driver. Since N-channel MOSFETs are used, it must be a high-side driver.
4. I couldn't figure out this section. At first I thought it was a low-value resistor used as a shunt, with a synchronous rectifier controller IC next to it, but there's another small MOSFET beside the resistor...
No, this is most likely the synchronous rectifier control section after all, but I didn't quite understand how it works.


I found some more components under the shunts. The board is very densely populated, while the underside is empty. In my opinion, some of the components could have been moved there.


1. More components under the shunts
2. LED indicator for blown output fuses.
3, 4. And here are the fuses themselves. They're quite unusual given their size and 20 A rating. There are two in parallel on the output and one on the input, since under normal operating conditions the input current won't exceed 20 A, while output current spikes can make this arrangement necessary.


Connecting everything is very straightforward. The two cables between the front-panel module and the converter are marked on the board, so just match the labels. It's practically impossible to connect the Bluetooth or USB module incorrectly, since there's only one connector on the board.
Remember, the connecting wires must not run alongside the power wires!


For the initial tests, I used my usual adjustable ZXY6005-based power supply.


The input voltage range is 6–60 V. At an input voltage of 6 V, the current draw is 70 mA. Raise the voltage to 60 V and the current drops to 10 mA, but power consumption increases by 1.5 times.


The controller firmware is version 1.6. The 8005 I showed recently had the previous version, 1.5, but I didn't notice any difference.
The settings menu is identical to the DPS8005. The only differences are the output voltage and current ranges, as well as the power range, which makes sense. All the other settings are exactly the same.
>

Since the settings and controls are identical to those of the previous converter, I'll copy the description and image from my previous review. If you'd like a more detailed explanation of the controls and their features, it's best to read my DPS8005 review. There's a lot of text, and there's no point in copying all of it here as well.



As before, I'll note that both the measurement and setting of the output parameters are quite accurate.
First, input voltage measurement. This parameter isn't very critical, but even in the worst case (maximum input voltage), the error was only in the last four digits; across the rest of the range, it was 1–3 digits.


Here I checked two parameters at once: the accuracy of setting the output voltage and the measurement error of the converter's own controller. Again, I think the results are good. The largest measurement deviation was two digits at the end, and the setting error was even smaller. The maximum difference in the set voltage was 0.01 V.


Current accuracy is a little worse. For example, instead of 10 mA I got 20 mA, and instead of 20.1 A, just 20.73 A, but accuracy was better across the rest of the range.
The built-in meter performs worse in this mode, which confirms what I said above: the output parameters can be set more accurately than the converter's controller can measure them.


Next, let's check the ripple level, which I think is even more important than the accuracy of setting voltage and current.
The “competitor” based on the ZXY6020 will serve as the power source.


Input voltage is 60 V in all tests; output voltage is 25 V.
1. No load
2. 2.5 A
3. 5 A
4. 7.5 A
5. 13 A
6. 20 A

To me, the ripple level is very high. But worse is the ripple waveform: instead of short spikes, it's a clear sawtooth. More on that later.


Let's try it at the maximum power for my power supply, around 700 W.
Input voltage: 60 V; output voltage: 34 V; load current: 20 A.


Peak-to-peak ripple is about 160 mV.


Knowing that these boards are sensitive to the power supply, I decided to test it with another one. In this case, it was a 48 V 5 A power supply, but I had raised its output voltage to about 52 V.


Output voltages of 21 and 30 V, with loads of about 135 and 270 W, respectively; current was 6 and 9 A.
Basically, nothing changed.


To finally put this test to rest, I used a linear power supply, which definitely shouldn't introduce any interference.
To be fair, when I tested operation with no load and saw the ripple, I tried disconnecting the power supply from the mains so it would run from its input capacitors, but there was no difference.

Output: 3 V, load current: 16 A; input: 20 V, with a current of about 2.6 A.


1. No load.
2. Voltage: 3 V, current: 16 A.

Well, the ripple did decrease, but the output voltage is low, and the power is only about 50 W.
Overall, the sawtooth at the output is entirely the converter's “doing,” not the power supply's, and the power supplies had little effect on it.


And the final test: thermal performance. To test it, I connected two electronic loads at once and loaded the converter to about 650 W, which is 65% of its maximum power.


I checked the board with a thermal camera after five minutes, then again about 15 minutes later. Nothing had changed.
I didn't measure the efficiency specifically, but the heat generated was enough to give me a good idea, and I can say for sure that the board is highly efficient. Even at 65% load, the components were barely warm, and the inductor was at body temperature.
The hottest component on the board is the power converter for the electronics.



Let's set the converter aside for now and move on to the enclosure.
Along with the converter board, the store offered an enclosure. I decided to get it as well, even though I'm skeptical of designs with an external power supply, as I've mentioned several times in the comments.

Link to the enclosure, price: $23.37 (option 2).



The enclosure is quite neat. It has all the necessary holes for the switch and terminals, as well as a hole for the USB connector—that's the difference from the standard version.


Inside the enclosure was a large bag of fasteners and other accessories.


The design is as simple as it gets: two U-shaped pieces that form a box when put together. The cover was held on with a couple of screws.


The package includes four bags of assorted bits and pieces.


1. One bag contained wires and a fan.
2. The fan is rated for 5 V and measures 40 mm.
3. The included wires have a cross-sectional area of about 2.5 mm². For comparison, the standard wire on the left is 2.5 mm², and the one on the right is 1.5 mm².
4. The second bag contained terminals, feet, and a power switch.
5. The power switch is the biggest concern: it's not rated for switching DC, as I'll explain a little further down.
6. There was also a little bonus: a pair of simple alligator clips.

Now, about the switch. I've mentioned before that a relay—or a switch, for that matter—rated for switching DC above 30 V is not only a tricky thing to find, but usually expensive as well. The reason is that when a DC circuit is broken, unlike an AC circuit, an arc can form, much like in welding. The result is pretty straightforward: the switch or relay burns out in a spectacular fashion.
Here, the switch is switching up to 60 V. The situation is somewhat mitigated by the capacitors downstream, which prevent a large voltage from appearing across the contacts immediately, but this is still strongly discouraged.


There were also seven bags of fasteners, with each type packed separately.


And, of course, there's the fan power board, which also serves as a kind of adapter between the terminals and the wires to the converter and switch.
1. Top
2. Bottom
3. It uses the same IC as the main converter.
4. It also has a 20 A fuse.
5. The kit didn't include a socket for connecting the fan, so I had to rummage through some boards I'd prepared for desoldering and remove one from there.
6. Of course, you could just solder the wires directly to the board, but I think a connector is better.

By the way, it would be quite easy to add temperature-dependent fan speed control here, like I did in the power supply based on the ZXY6005. There, a thermistor was placed in the feedback circuit of the converter that powered the fan.


Now I can start fitting almost everything in the photo into the enclosure.


First, the USB adapter for connecting to a computer.
This is simple: plastic standoffs and nuts, plus a couple of screws. Everything fit perfectly.


The switch and terminals.


1. The terminals have a non-round plastic insert to prevent them from turning. The second plastic washer has a matching recess.
2. After installation, there should be one nut left on each terminal.
3. Position the rear terminals so that the red one is at the top; otherwise, you won't be able to connect the fan power board, or you'll have to install it upside down. I positioned the switch so that it turns on to the left. That way, when the power supply is in its normal position with the screen facing me, the switch turns on to the right. That seemed more logical to me.
4. The terminals can be positioned however you like at the front, but I also put the red one at the top.


The fan, board, and wires.


This is where I found a design flaw. The switch also connects to the board, but the holes in the board for its wires are clearly too small. Using wires with a different gauge would be wrong, so I simply soldered them onto the top of the pads.


The trickiest part here is not overheating the switch contacts while soldering the wires. It’s better to use a slightly overheated soldering iron. I tucked the fan wire underneath it so it wouldn’t get caught in the blades.


1. This is all that’s left of the supplied wires—not much, of course, but it should be enough.
2. The kit included seven fork terminals for the wires. It’s best to tin the wire first and not leave the stripped end too long.
By the way, the board also came with four “forks,” so now I have a whole five in reserve :)


We can install the board in the case now, but you’re probably wondering what those bits of metal are at the bottom left—they don’t seem to have been included in the kit.


Yes, I decided to install the board a little “unconventionally.” Since I didn’t have any proper 25 mm standoffs on hand, I cut some from aluminum tubing. Ideally, the standoffs should be 26 mm tall, but 25 mm would work too.
But note that the plastic standoffs must stay in place! The board must not make contact with the case, since not all the pads under the standoffs are connected to ground.


This seemed like the best way to install the board, since it solves three problems:
1. Cooling is not compromised (and may even be improved)
2. The power wires are as short as possible.
3. The power wires don’t cross the wires to the display and control buttons, which could cause serious problems.

There was another option: install the board with the components facing up and the inductor toward the front panel. In that case, the display module connectors would be near the back, and the wires could run under the board. But I wanted to do it the way shown in the photo.


And here’s what the inside of my power supply based on the ZXY6020 looks like.


Although there’s a ground plane on the underside of the circuit board, I twisted the wires to the USB adapter together anyway to reduce interference pickup.
All that’s left is to glue on the feet and screw the case together.


That’s it. I only have one complaint about the appearance: there’s a gap at the top caused by the metal being bent slightly out of shape. Otherwise, it looks perfectly fine.


Let’s power it up. Everything seems to work. But I’ll point out right away that the fan is quite noisy, and it makes noise whenever the power supply is on.
This gave me the idea of connecting it to the fan power connector on the converter board, but increasing the heatsink size by about 2–3 times. That would “kill two birds with one stone”:
1. Automatic speed control would work
2. The transistors would get about the same cooling as with the original fan.



Questions about this and similar converter models.

While I was preparing this review, some of my readers already knew I was working on a review of this board and asked me various questions. I’ll try to answer them now.


Just a few words about the software. It’s practically the same as the DPS8005 software, except that when I tried to launch the 8005 software with this board, it complained loudly and told me to find another version—which is exactly what I did. You can download the software for both Windows and Android here.

I installed the program, and it launched without any problems. By the way, in this screenshot, a battery is connected and the external power supply is disconnected, so the voltage at the board’s input is lower than at its output.


Now we connect the power supply, and everything returns to normal. The power supply was set to turn on automatically; the output voltage is the battery voltage, and the charging current is 1 A.


The program’s advanced settings tab.


I was left with the included alligator clips, a couple of screws, five terminals, and a Bluetooth adapter.
The adapter is worth thinking about some more. For example, you could install it in place of the USB port or connect them in parallel, as I did in my ZXY6020 review.


And here’s a group photo. Of course, these aren’t all the power supplies I own, but the ones shown here have a lot in common.
ZXY6020S
ZXY6005S
DPS5020

By the way, the ZXY6020 and ZXY6005 differ in more than just output current. The ZXY6020 has more settings, although at first glance they might seem identical.


Here are my conclusions after inspecting and testing it.
First, the converter.
It has a very interesting design, genuinely high efficiency, and a good combination of features. I liked the module’s own software features less than those of the ZXY, but its PC/tablet software offers much more. The ZXY is easier to operate and has a higher screen refresh rate, making it more convenient to use. However, these modules are quite large, and the price is about twice as high.
The DPS has the better circuit design: its high-side switch is an N-channel MOSFET with a dedicated driver. It also has synchronous rectification and better fan speed control.
If someone made a kind of ZXY+DPS hybrid, that combination would probably come close to the ideal power supply—but alas.

What I didn’t like about the DPS was the fairly high output ripple, as well as the possibility of interference getting into the wires running to the display and buttons.
Also, there’s just one 20 A fuse on the input. Near the end of my experiments, when the unit was already in its case, it blew with a snap. I applied 52 V, simply flipped the switch, and the fuse blew. Apparently, there was a poor connection in the switch, and the contacts had partially welded together. After replacing the fuse with a jumper, everything worked as before.
Personally, I think it would be better to use two fuses in parallel on the input as well, and to include spares for replacement.

Now for the case and what comes with it.
The paint job is neat, it looks pretty good, and the fan even has holes shaped like nuts, which makes installation very convenient. All the necessary fasteners, connectors, and other parts are included, so assembly is quite pleasant.
But there are some drawbacks, too. The top cover doesn’t sit flush against the bottom half in the center, which detracts from the overall finish. There’s no fan connector on the board, the holes for connecting the switch are smaller than they need to be, and the wires are barely long enough. If you install the board as intended, the wires probably won’t reach at all.
The fan is noisy and runs constantly. It’s better to replace the stock fan, and swap the stock heatsink for a larger one.
A separate word about the switch: you shouldn’t use it at all, because it isn’t rated for switching DC above 20–30 V, especially at currents up to 20 A. It’s better to switch the 220 V AC supply on the input side of the power supply.

And I’ll say it again: personally, I’m not too keen on the idea of having the power supply and converter as separate units. But I liked the case, and I even considered building a roughly 300 W switching power supply to fit the space provided. That would let me use a compact DPS8-8005 with it and get a small, convenient adjustable power supply, like the ZXY6005. Ideally, I wouldn’t use the converter’s power stage at all, but would take only its feedback circuit and combine it with the power supply’s feedback circuit.

That’s about it. I’m sure I’ve forgotten something. If you have any questions, I’ll be happy to answer them. I hope you found this review useful.