RD6012P(W) Hybrid Adjustable Voltage Converter

I’ve already posted an announcement for this converter model, which is an evolution of the previous 6006P model, differing mainly in its lower output current and power.
Several readers have already asked me about this model. Well, it arrived, I took a look, tested it, and formed an opinion.

As usual, a reminder: to keep up with new posts on the blog, I recommend following me on Instagram, where I’ll post notifications about every new article and maybe write about something interesting now and then — link to my account

A few words by way of introduction.
No matter how I test the device or what I write about it, essentially this is a higher-power version of the well-known RD6006P, which “grew out of” the even better-known RD6006, so naturally they have a lot in common. To avoid repeating myself, I’ll just link to my reviews of all the previous RuiDeng models I’ve had.
DPH5005
DPS8005
DPS5020
RD6006
RD6012
RD6018
RD6006P

There are two key differences between the “P” series and the earlier models:
1. This is a hybrid converter: it first uses a DC-DC step-down converter, followed by a linear regulator.
2. Separate DACs/ADCs are used for current/voltage measurement and setting, rather than the ones built into the microcontroller, which improves accuracy.

I think it makes sense to compare the new model with its predecessor, and some of the differences are already apparent in the specifications table.


As always, the packaging is excellent: a large foam box plus an individual box made of thick cardboard.


The package contents are the same as for the previous models and include:
1. Converter
2. Cable with temperature sensor
3. A pair of fork terminals
4. Fuse
5. WiFi module (optional)


WiFi module, temperature sensor, terminals, and a 20 A fuse.


Overall view of the converter.


The front panel is identical to all the previous RD60xx series models, right down to the color of the plastic.


The same buttons for entering parameters directly, and screw terminals for connecting either a regular load (black/red) or a battery for charging (black/green).
As for the dimensions, the only difference between the modules is their thickness, which depends on the components installed. Width/height: 160 × 74 (167 × 81) mm.


But I think most people are more interested in what’s inside than in how it looks, so let’s take a look at the back of the unit.
And here it looks very much like the 6006P model. I’d even say it’s almost identical, apart from a few minor details.


Almost all the differences are in the input section.


1. The input common-mode choke is clearly larger than the one in the 6006; the reverse-polarity protection transistor is next to it.
2. The input terminal block has four contacts. A pair of 470 µF, 80 V capacitors is installed on the input.
3. Auxiliary voltage converters: everything here is the same as in the 6006P.
4. The DC-DC converter has been beefed up. The 6006P used a P-channel transistor plus a diode assembly; this model uses a synchronous rectifier built around a pair of N-channel HY3210s.
5. I didn’t measure the size of the power inductor. Subjectively, it’s about the same as in the 6006, but here it’s wound with two wires instead of one, as in the 6006.
6. The reverse-polarity protection relay for battery charging is rated for 16 A, but the more important difference is just to its left: the manufacturer added an inductor to reduce ripple after the DC-DC stage.


All the additions and changes to the input section meant the main PWM controller no longer fit on the main board, so it had to be moved to a separate, smaller board. But it’s still the same TL594 used in all the previous models.


The DAC/ADC and linear regulator section looks similar to the one in the 6006, but there are differences. For example, this model uses just one optocoupler for galvanic isolation, while the 6006 had three. There are other minor differences too, although most of the components are arranged in the same way.


Interestingly, there’s also an inductor on the output to reduce ripple. Apparently, at high currents it just couldn’t “keep up” without one.


The power board comes out of the case just like it does in the previous models.


On the underside are only the linear regulator transistors, shunts, and the heatsink for the auxiliary PWM controller.


The transistors and shunts are the same as in the 6006P, but there are twice as many here. Judging by the layout, the transistors are simply connected in parallel, which isn’t ideal. Ideally, they’d have small balancing resistors.


Overall view of the digital and analog section with the case removed.


Control board


I didn’t notice any differences from the 6006P here.


For testing, I connected the module to its “sibling,” the RD6018. The tests were either done in this setup or with the converter powered directly from the 800 W power supply I reviewed earlier.


On startup, the splash screen appears, followed by the main screen, which displays all sorts of useful and not-so-useful information:
1. Current time and auxiliary operating modes
2. Output readings: voltage, current, and power
3. Input voltage
4. Set values: voltage and current
5. Trigger protection settings: voltage and current
6. Operating modes, capacity delivered, and external sensor temperature.

You can also quickly switch to two other display layouts, but I personally prefer the main one.


Unlike all the previous models, this one has two operating modes: 0–6 A and 0–12 A. In the first mode, the output current goes up to 6.1 A, with higher accuracy. In the second, accuracy is lower, but the range extends to 12.1 A.
All well and good—I might even have forgiven the developers for not including automatic switching, if it weren’t for one “but.”
1, 2. The operating modes are switched using the up/down buttons, which also changes the meter/setpoint resolution.
3. The problem is that the protection limit doesn’t switch. So after switching from 6 A to the 12 A range, you have to load a preset memory slot with a higher limit; otherwise, you won’t get more than 6.2 A.


The settings menus in the latest converter versions are practically identical, as are the supported functions.


What’s more, the Russian localization errors carry over from one version to the next.


Let’s move on to the tests.

First test: input voltage measurement accuracy. For the most part, this test is pointless—in my opinion, this feature is rarely needed at all—but why not check it?
Overall, it’s pretty good, especially for a secondary feature. There was a slight positive error in the 10–60 V range, about 10–50 mV across the entire range, and it was practically linear.


Test two: voltage setting and measurement accuracy.
Everything looks very good here. There was a little "wobble" at the very lowest values, but we are talking about a range of 1–10 mV.


Test three: current setting and measurement accuracy.

First, the 0–6 A range.
The readings wobbled noticeably at low values, but above 1 mA the results were pretty good.


And the 0–12 A range.
This test was more of an add-on, so I ran it in abbreviated form. At first, I thought accuracy was poor at high currents, but then I realized that a 0.1% error relative to the multimeter isn't bad at all :)


Test four: output voltage versus load.
Everything is excellent here, even with a two-wire connection: the converter holds the output voltage without any problems across the 0.05–12 A load range.


However, during the test I ran into the same glitch as with the previous model. At one point, I noticed that the built-in voltmeter was showing the wrong value. And yes, its readings were low, even though the output voltage was set correctly and the current was also measured correctly. Restarting it with the button on the front panel doesn't help; it only resets the operating mode and returns to the first memory slot. Only completely disconnecting the power fixes it.


Test five: minimum input/output voltage difference.
I tested two input voltages (35 and 61 V) and two load levels (6 and 12 A).
The results showed that with a 35 V input, the minimum headroom needs to be 3.2–4.2 V, while at 60 V it needs to be a little higher: 5.5–6.2 V.

To see the difference, let's compare the measured input voltage with the output voltage the power supply was able to deliver.


Test six: ripple level at an output voltage of 30 V and load currents of 0, 4, 8, and 12 A.
Overall, the ripple itself is very low, though the "fuzz" was a little annoying. I captured the waveforms with the probe connected directly, i.e. without a C+C filter. With the filter, almost all the "fuzz" disappeared.

The converter was powered by an 800 W switching power supply.


Here the current was 12 A throughout, but the output voltage varied: 5, 12, 24, and 48 V.
The ripple was somewhat higher, but its main component (excluding the "spikes") was still no more than 5 mV.


Test seven: transient response.
1, 2. Applying/removing 5 V at the output.
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, 6. Transition from CV mode to CC mode and back, at 15 V with a 4.5 A load current and the current limit set to 3 A.

The main thing I noticed during the test was that the output voltage took a long time to drop. Otherwise, the results were quite good.


Test eight: connecting a TS-100 soldering iron. This test checks how well the converter handles a pulsed load of 0–2.8 A.
As with the RD6006P, I have no complaints.


Test nine: converter efficiency and power losses.
Input voltage was about 68–70 V, load current was 12 A, and output voltage was 5, 12, 24, 36, 48, and 60 V. The converter's no-load current draw when powered from 60 V ranged from 30 to 40 mA.

Output voltage — efficiency — power loss (efficiency and power loss for the 6006P)
5 V — 63.5% — 34.9 W (60.0% — 20.1 W)
12 V — 79.6% — 36.7 W (78.3% — 20.0 W)
24 V — 87.9% — 39.35 W (87.7% — 20.27 W)
36 V — 91.7% — 38.68 W (91.6% — 19.84 W)
48 V — 94.1% — 36.1 W (93.8% — 18.9 W)
60 V — 95.5% — 34.06 W (94.6% — 19.1 W)

Overall, efficiency is about the same as the 6006P's, and slightly higher in some cases. But since this model operates at twice the load power and current, its power losses were also almost twice as high, which inevitably affects the temperature.
Since the converter was lying on the table with its screen facing up, which hindered cooling, it heated up very quickly and significantly.


Test ten: temperature rise.
1. In this test, the converter was loaded at 12 A with an output voltage of 24 V, which, according to the efficiency results above, corresponds to the maximum power loss.
2–4. The initial temperature reported by the built-in temperature sensor was 23 degrees. After 5 minutes it was already 57, after 10 minutes it had risen to 61 degrees, and 20 minutes into the test it reached 63 degrees. I don't think there's any point in running the test longer.


1, 2. Thermal images taken 10 and 20 minutes into the test.
3. The inductor is the hottest component, while the DC-DC section's heatsink runs noticeably cooler, at around 60 degrees.


But don't get too comfortable: as I mentioned above, the temperature rise depends heavily on the cooling conditions. For example, after the previous test I placed the converter with its screen facing up, and within a minute it was already above 70 degrees. Another minute later, the temperature had risen to 80 degrees and the converter shut off its output.


This isn't really a test, just a bit of curiosity: I wanted to see how much voltage drops across the linear regulator.
Connecting directly to the transistor was extremely awkward, so I had to measure with the shunt and output inductor in the circuit. I don't think this makes much difference, though, since when I measured at different currents (1 and 12 A), the difference was roughly what you'd expect from the voltage drop across those components.
1. Output: 5 V, current: 1 A
2, 3. Output: 5 V, current: 12 A; as it warms up, the voltage gradually drops.
4. Output: 40 V, current: 1 A
5, 6. Output: 40 V, current: 12 A; here, though, the output voltage gradually increased.

On the 6006Р, the voltage drop across the regulator itself was about 1.5 V. It's a little higher here, but nothing significant.


Someone recently mentioned in the comments that they were having trouble connecting Android devices over Wi-Fi, so I decided to check it out.

First, download and install the latest version of the app, then launch it. A short guide appears on startup, explaining that you need to enable location services. I checked, and they were already enabled on my phone.


Next, open the software itself. The guide is shown only the first time you launch it.
It asks for access to your location and photos, then asks you to agree to its terms. There's no way around it, so we agree.
Next, launch “Distribution Network.” The app opened in Russian automatically on my phone; it would probably be better to switch it to English.


Now for the connection.
Select “WiFi” in the module settings and reboot it. The screen will show that it's waiting for a connection; from there, everything is done in the app.
1, 2. The biggest problem was getting the app to recognize that I was already connected to a 2.4 GHz network, because it stubbornly thought I was connecting over 5 GHz. In the end, I had to switch my phone from 2.4 GHz to 5 GHz, launch the software and let it complain, then switch the phone back to 2.4 GHz. Only then could I access the Initialize button.
3, 4. After initialization, wait for the module's screen to display its IP address, then tap “Confirm.”
5, 6. The software searches for something, finds it, and happily lets us know. That's it.


Here's what the process looks like on the module's screen.
1. Enable WiFi mode and reboot.
2. The screen shows that it's waiting for a connection. If needed, press the encoder to start the module in normal mode.
3. Screen after successful initialization, step 4 in the description above.
4. The first time, something went wrong here. It looked like the connection was established, but the app didn't connect to the module.
5. I rebooted the module. Lines scrolled across the display, including one error.
6. All done—the software is connected, as indicated by the WiFi indicator turning yellow.


Overall, there don't seem to be any differences from the previous software versions. The only thing I noticed is the switch between the 0–6 and 0–12 A ranges.


Now, about the computer software.
You can download it from the link on the product page, but when I launched it, it immediately offered to update. You can download the updated version here.


Next, I tried connecting the software over WiFi, but after a while I gave up on that tedious quest and connected via USB.
After connecting to the module, the software notified me that an updated firmware version was available, with a fix for the power limit function.


I updated the firmware; fortunately, it was easy.


Now let's take a look at the software.
There's not much to see here, since it's basically no different from the previous versions. The only change is that it now has a button for switching ranges.


Conclusions.
In my opinion, this converter turned out pretty well. No, I won't call it excellent—there are still a few shortcomings—but overall it looks quite respectable.
Compared with the previous model, the only difference is the higher current and, consequently, higher power. Otherwise, there are almost no changes; even the software bugs are still there.
That said, the hardware differences compared with the 6006Р are noticeable:
1. A four-terminal input block instead of a two-terminal one, and a larger common-mode choke
2. A synchronous rectifier and N-channel transistors
3. The power inductor is wound with two wires
4. Chokes have been added to reduce ripple: one before the linear regulator and another after it.
5. A pair of current-sense shunts and linear regulator transistors have been installed instead of a single shunt and transistor.
6. The PWM controller has been moved to a separate board.
7. A higher-current protection relay.

What I didn't like: the awkward range-switching system. As I said, I can live without automatic switching, but it's annoying that after switching from the 0–6 range to 0–12 A, the maximum current stays at the same value, i.e. 6.2 A...
It runs considerably hotter now, so you need to pay closer attention to case ventilation. I think if they release a 6018Р model, overheating could well become an issue. Though I'm probably hoping for something like a 6024 (without the Р).
And then there's the occasional output voltage measurement error, which has also carried over from the 6006Р version.

Subjectively, though, I think it's pretty good. It's nice to see that the manufacturer hasn't abandoned the idea of expanding the product range and is releasing new models. That said, at the time it was announced, it cost 20–25 dollars less. It's a shame the manufacturer raised the price :(
The seller sent over a discount coupon.
If I haven't got it wrong, the coupon gives you 7 dollars off.
https://rdtech.aliexpress.com/store/selected-sale-items/923042.html?es=0zoEZ8%2F9sMYBqdzmdP13eTtkm3LfdWSL


That's all for now. I hope this was useful, and as always, I look forward to your comments and questions.