DC-DC Step-Down Module Rated for 10 A

Before New Year, readers asked me to review a couple of converters.
Well, it wasn't really any trouble, and I was curious too, so I ordered one, received it, and tested it.
I was actually more interested in a slightly different converter, but I still haven't gotten around to it, so I'll cover that one another time.
Today, though, I'm reviewing a simple DC-DC converter rated for 10 A.

I apologize in advance to everyone who has been waiting a long time for this review for the considerable delay in publishing it.

First, here are the specifications listed on the product page, along with a few clarifications and corrections.
Input voltage: 7-40V
1, Output voltage: continuously adjustable (1.25-35V)
2, Output Current: 8A, 10A maximum time within the (power tube temperature exceeds 65 degrees, please add cooling fan, 24V 12V 5A turn within generally be used at room temperature without a fan)
3, Constant Range: 0.3-10A (adjustable) module over 65 degrees, please add fan.
4, Turn lights Current: current value * (0.1) This version is a fixed 0.1 times (actually turn the lamp current value is probably not very accurate) is full of instructions for charging.
5, Minimum pressure: 1V
6, Conversion efficiency: up to about 95% (output voltage, the higher the efficiency)
7, Operating frequency: 300KHZ
8, Output Ripple: about the ripple 50mV (without noise) 20M bandwidth (for reference) Input 24V Output 12V 5A measured
9, Operating temperature: Industrial grade (-40 ℃ to + 85 ℃)
10, No-load current: Typical 20mA (24V switch 12V)
11, Load regulation: ± 1% (constant)
12, Voltage Regulation: ± 1%
13, Constant accuracy and temperature: the actual test, the module temperature changes from 25 degrees to 60 degrees, the change is less than 5% of the current value (current value 5A)

Let me put that into more understandable terms.
1. Output voltage adjustment range: 1.25-35 V
2. Output current: 8 A; it can handle 10 A with additional cooling from a fan.
3. Current adjustment range: 0.3-10 A
4. Charge indicator cutoff threshold: 0.1 times the set output current.
5. Minimum difference between input and output voltage: 1 V (presumably)
6. Efficiency: up to 95%
7. Operating frequency: 300 kHz
8. Output voltage ripple: 50 mV at 5 A, with an input voltage of 24 V and an output voltage of 12 V.
9. Operating temperature range: -40 ℃ to +85 ℃.
10. Quiescent current: up to 20 mA
11. Current regulation accuracy: ±1%
12. Voltage regulation accuracy: ±1%
13. The parameters were tested over a temperature range of 25-60 degrees, and the change was less than 5% at a load current of 5 A.

The order arrived in a standard plastic bag, generously wrapped in polyethylene foam tape. Nothing was damaged in transit.
Inside was my test subject.


No complaints about the appearance. I just turned it over in my hands and couldn't really find anything to criticize. The workmanship is neat, and if the capacitors were replaced with name-brand ones, I'd say it looked good.
There are two terminal blocks on one side of the board, for the power input and output.


On the other side are two trimmer potentiometers for adjusting the output voltage and current.


Looking at the photos in the store, the board seems fairly large.
I deliberately took the previous two photos as close-ups as well. But you get a sense of its size when you put a matchbox next to it.
The board is really small. I didn't check the dimensions when I ordered it, but for some reason I thought it would be noticeably bigger. :)
Board dimensions: 65x37 mm
Converter dimensions: 65x47x24 mm


The board is double-sided, with components mounted on both sides.
No complaints about the soldering either. Sometimes large contacts aren't soldered properly, but you can see in the photo that this isn't the case here.
The components aren't numbered, though. I don't think that's a big deal; the circuit is fairly simple.


In addition to the power components, the board also has an op-amp powered by a 78L05 regulator, as well as a simple reference voltage source built around a TL431.


The board uses a powerful PWM controller, the XL4016E1, and it is even electrically isolated from the heatsink.
I don't know why the manufacturer isolated the chip from the heatsink, since this reduces heat dissipation. Perhaps it's for safety, but since the board is usually installed inside something, it seems unnecessary to me.


Since the board is designed for a fairly high output current, a powerful diode assembly, the MBR20100CT, is used as the power diode. It is also mounted on the heatsink and electrically isolated from it.
In my opinion, this is a very good solution, but it could be improved a little by using a 60 V assembly instead of a 100 V one.

The inductor isn't very large, but this photo shows that it's wound with two wires, which is already a good thing.


1, 2. The input has two 470 µF, 50 V capacitors; the output has two 1000 µF capacitors, rated for 35 V.
Going by the list of claimed specifications, the output capacitors' voltage rating is right on the limit. But it's unlikely anyone would step down from 40 V to 35 V, not to mention that 40 V is the chip's absolute maximum input voltage.
3. The input and output terminals are labeled, though the labels are on the underside of the board. That's not a big deal.
4. The trimmer potentiometers, however, aren't labeled at all.
The one on the left adjusts the maximum output current; the one on the right adjusts the voltage.


Now let's take a closer look at the claimed specifications and what we actually have.
I mentioned above that the converter uses a powerful PWM controller—or, more precisely, a PWM controller with an integrated power transistor.
I also quoted the board's claimed specifications above. Let's take a closer look.
Claimed: Output voltage: continuously adjustable (1.25-35V)
No questions here: the converter will output 35 V, and theoretically even 36 V.
Claimed: Output Current: 8A, 10A maximum
Here's where I have a question. The chip manufacturer clearly specifies a maximum output current of 8 A. The chip's specifications do include a line for maximum current limit: 10 A. But that's nowhere near the maximum operating current; 10 A is the absolute limit.
Claimed: Operating frequency: 300KHZ
300 kHz is certainly nice—you can use a smaller inductor. But sorry, the datasheet clearly specifies a fixed frequency of 180 kHz. Where did 300 come from?
Claimed: Conversion efficiency: up to about 95%
Well, this one is fair: efficiency is up to 95%. The manufacturer actually claims up to 96%, but that's theoretical and depends on the ratio between the input and output voltages.


Here's the PWM controller's block diagram, along with an example implementation.
By the way, you can see here that for an 8 A current, the inductor should be rated for at least 12 A—that is, 1.5 times the output current. I usually recommend a 2× margin.
It also shows that the output diode can be rated for 45 V. Diodes rated for 100 V usually have a higher voltage drop, which reduces efficiency.
If you want to improve this board's efficiency, you can salvage 20 A, 45 V diodes—or even 40 A, 45 V ones—from old computer power supplies.


Initially, I didn't want to draw the schematic: the top of the board is covered with components, solder mask, and even silkscreen. But then I realized that redrawing the schematic was quite doable, so I decided to keep up the tradition :)
I didn't measure the inductor; 47 µH is taken from the datasheet.
The circuit uses a dual op-amp. One half is used for current regulation and stabilization, the other for indication. You can see that the second op-amp's input is connected through a 1:11 divider. The description actually specifies 1:10, but I don't think the difference matters.


First no-load test. The board was initially set to an output voltage of 5 V.
The voltage remains stable across an input voltage range of 12–26 V. The current draw is below 20 mA, since the power supply's ammeter doesn't register it.


The LED lights red when the output current falls below 1/10 (1/11) of the set value.
This indicator is used when charging batteries: if the charging current drops below 1/10 during charging, the battery is usually considered fully charged.
That is, set the charging current to 4 A, and the LED stays red until the current drops below 400 mA.
One caveat: the board only indicates that the current has dropped. It doesn't cut off the charging current; it simply continues to decrease.


For testing, I put together a small test setup using the following equipment:
Adjustable power supply
Electronic load
Oscilloscope
Multimeter
Non-contact thermometer
Thermal imager
A pen and a piece of paper—I lost the link :)

But during testing, I eventually had to use this adjustable power supply as well, because my experiments had disrupted the linearity of the current measurement/setpoint in the 1–2 A range on the high-power supply.
So first I tested the temperature rise and assessed the ripple level.


This time, the testing was done a little differently than usual.
I measured the heatsink temperatures at points close to the power components, since the components themselves were difficult to measure due to the dense layout.
I also checked operation in the following modes:
Input — output — current
14 V — 5 V — 2 A
28 V — 12 V — 2 A
14 V — 5 V — 4 A
And so on, up to 7.5 A.

Why did I test it in this somewhat unusual way?
1. I wasn't sure how reliable the board was, so I increased the current gradually, alternating between different operating modes.
2. I chose 14 V to 5 V and 28 V to 12 V because these are among the most commonly used modes: 14 V (approximately the voltage of a passenger car's electrical system) to 5 V (for charging tablets and phones), and 28 V (the voltage of a truck's electrical system) to 12 V (another commonly used voltage).
3. Initially, I planned to keep testing until it shut down or burned out, but my plans changed, and I came up with a few ideas for the components on this board. So I only tested it up to 7.5 A. In the end, though, this didn't affect the validity of the test at all.

Below are a couple of sets of photos showing the 5 V, 2 A and 5 V, 7.5 A tests, along with the corresponding ripple levels.
Ripple at 2 A and 4 A was similar, as was ripple at 6 A and 7.5 A, so I'm not including the intermediate results.


Same as above, but with a 28 V input and 12 V output.


Thermal performance with an input of 28 V and an output of 12 V.
You can see that there’s no point increasing the current any further: the thermal camera is already showing the PWM controller at 101°C.
I use a rough limit of my own: component temperatures should not exceed 100°C. It really depends on the components, of course. Transistors and diode assemblies can be operated safely at higher temperatures, for example, but it’s best not to exceed this limit with ICs.
The photo doesn’t show it very well, of course. The board is very compact, and it was a little easier to see in the live view.


Since I figured this board might be used as a battery charger, I checked how it would perform with 19 V at the input (a typical laptop power supply voltage) and 14.3 V at the output at 5.5 A (typical charging parameters for a car battery).
Everything went fine here—well, almost. More on that later.


I put the temperature measurements into a table.
Based on the test results, I wouldn’t recommend using the board above 6 A, at least not without additional cooling.


I mentioned above that there were a few quirks. Here’s what I mean.
During testing, I noticed that the board behaves a little strangely in certain situations.
1.2 I set the output voltage to 12 V with a 6 A load. After 15–20 seconds, the output voltage dropped below 11 V, so I had to adjust it.
3,4 The output was set to 5 V with 14 V at the input. I raised the input to 28 V, and the output dropped to 4 V. The photo on the left shows 7.5 A, and the one on the right 6 A, but the current didn’t matter: when the input voltage is raised under load, the board “drops” the output voltage.


After that, I decided to check the device’s efficiency.
The manufacturer provided graphs for different operating conditions. I’m interested in the graphs for 5 V and 12 V output with 12 V and 24 V input, since those are closest to my test conditions.
Specifically, the claimed figures are:
12 V input, 5 V output
2 A — 91%
4 A — 88%
6 A — 87%
7.5 A — 85%

24 V input, 12 V output.
2 A — 94%
4 A — 94%
6 A — 93%
7.5 A — Not specified.


The next test was fairly straightforward in principle, but there were a few quirks.
The 5 V test went without a hitch.


The 12 V test, however, had a few quirks. Here’s what happened.
1. 28 V input, 12 V output, 2 A: all good
2. 28 V input, 12 V output, 4 A: all good
3. Increase the load current to 6 A, and the output voltage sags to 10.09 V
4. Adjust it back up to 12 V.
5. Increase the load current to 7.5 A; it drops again, so I adjust it again.
6. Reduce the load current to 2 A without readjusting, and the output voltage rises to 16.84 V.
At first, I wanted to show how it rose to 17.2 V with no load, but decided that wouldn’t be a fair comparison, so I included a photo with a load connected.
Yeah, that’s disappointing :(


While I was at it, I also checked the efficiency when charging a car battery from a laptop power supply.
But this test had its quirks too. First, I set the output to 14.3 V, ran a temperature test, and set the board aside. Then I remembered that I also wanted to check the efficiency.
I connected the cooled-down board and saw an output voltage of about 14.59 V, which dropped to 14.33–14.35 V as it warmed up.
So, in practice, the board's output voltage is unstable. This kind of variation isn't critical for lead-acid batteries, but you absolutely must not charge lithium batteries with this board.


I ended up with two efficiency test results.
They are based on two sets of measurements, although the results don't differ very much.
P out is the calculated output power, with the input current rounded off; P out DCL is the output power measured by the electronic load. Input and output voltage were measured directly at the board's terminals.
So I got two efficiency measurements. Either way, the efficiency is roughly in line with the claimed figures, though a little lower.
Here are the figures from the datasheet again:
12 V input, 5 V output
2 A — 91%
4 A — 88%
6 A — 87%
7.5 A — 85%

24 V input, 12 V output.
2 A — 94%
4 A — 94%
6 A — 93%
7.5 A — Not specified.

And here's what I got in practice. I think replacing the power diode with a lower-voltage equivalent and installing an inductor rated for a higher current could squeeze out another couple of percentage points.


That seems to be everything, and I can already guess what readers are thinking:
Why all these tests and confusing photos? Just tell us the bottom line: is it any good or not? :)
And to some extent, readers would be right. The review could probably be two or three times shorter if I removed some of the test photos, but I'm used to doing it this way, so please bear with me.

So, here's the verdict.
Pros
Quite well made
Compact size
Wide input and output voltage ranges.
Charge-completion indicator (when the charging current drops)
Smooth current and voltage adjustment (you can easily set the output voltage to within 0.1 V)
Excellent packaging.

Cons.
At currents above 6 A, additional cooling is recommended.
The maximum current is 8 A, not 10 A.
Poor output voltage regulation; it may vary with load current, input voltage, and temperature.
Sometimes the board started to “sing.” This happened over a very narrow adjustment range: for example, when I changed the output from 5 to 12 V, it gave a quiet whine at 9.5–10 V.

One more thing to keep in mind:
The board only indicates that the current has dropped; it cannot stop charging. It's just a converter.

My take: honestly, when I first picked up the board and turned it over, looking at it from every angle, I was ready to praise it. It was neatly made, and I had no major complaints. When I powered it up, I still wasn't inclined to criticize it much. It gets hot, but they all do; that's normal, basically.
But when I saw how the output voltage jumped around in response to just about everything, I was disappointed.
I don't want to investigate these problems, since that's the manufacturer's job—they're making money from this. But I suspect the issue comes down to three things:
1. A long feedback trace that runs almost around the entire board
2. Trimmer resistors installed right next to the hot inductor
3. The inductor is located directly above the section containing the “sensitive” electronics.
4. Non-precision resistors are used in the feedback circuits.

Bottom line: it's perfectly suitable for undemanding loads, definitely up to 6 A, and works reasonably well. One option is to use the board as a driver for high-power LEDs; it should work well.
Using it as a battery charger is questionable, and in some cases dangerous. A lead-acid battery may tolerate these fluctuations, but you must not charge lithium batteries with it, at least not without modifications.

That's all. As always, I look forward to your comments, questions, and additions.

The product was provided by the store for review.