/       /       /    Low-profile GFA20A-12 12 V 20 W power supply from SANMIM
Support the project on Patreon


Low-profile GFA20A-12 12 V 20 W power supply from SANMIM

$6.5
Go to store
A second power supply from SANMIM, this time a 20 W, 12 V model from its low-profile, or “flat,” series.


Low-profile power supplies aren’t needed very often, and the selection is usually very limited. But I can say that in some situations they can make life a lot easier, especially when using a regular power supply would significantly increase the size of the enclosure. So I think these can be useful too, and decided to order one.

The power supply costs 174 UAH, or about $6.50 for one, and around $6 each if you buy two. Mine came in a plain zip-lock bag; I don’t know how the manufacturer ships them.


This series comes in two versions, GFA15A-хх and GFA20A-хх, rated at 15 and 20 W respectively. The хх indicates the output voltage.
The only key difference between the power supplies is their output power. Both are fixed-output models, although if absolutely necessary, you can adjust the output voltage slightly by changing the values of the feedback resistors.


Like the previous power supply, this one comes with an instruction manual, unfortunately in Chinese again.


The power supply is available with six output voltage options: 5, 9, 12, 15, 24, and 36 V. But the 20 W version is only available in the store with a 12 V output. In my opinion, a 5 V version would also be very useful, at the very least.
The specified operating temperature range is -30 to +70 degrees, but the datasheet includes a graph showing how the maximum output power depends on ambient temperature and input voltage.


No complaints about the exterior; everything is very neatly done.


Though I suppose I’ll nitpick: the output rectifier assembly’s heatsink isn’t secured at all. It’s held in place only by the stiffness of the assembly’s leads. A bit of a blunder, though...


A drawing with the power supply’s overall dimensions. Almost everything matches the actual measurements; the only difference is that it’s thinner than specified: 20 mm, not 21.5 mm, which is even better.


Compared with a matchbox.
Interestingly, some of Mean Well’s power supply dimensions are based on inches. For example, the board is 2 inches wide and 3 inches long, so there’s some standardization of dimensions. This one has no such relationship.


Components and circuit design.
1. The input and output connections are brought out to a screw terminal block. There are five terminals because the ground connection is brought out as well.
It has a full mains filter, including a 10 mm, 470 V varistor. There’s also a thermistor, an X capacitor, and a common-mode choke.
2. All the noise-suppression capacitors are Y-type, which is a big plus. The power supply is safe to use without a ground connection, too.
3. The input capacitor is rated at 47 µF and made by Aishi, just like the one in the previous power supply.
4. The circuit design is very similar to the previous power supply, with the main difference being the use of higher-power components. For example, it has a CNB7N60 high-voltage transistor rated for up to 600 V and 7 A. There’s also a difference beyond just the power rating: it looks like this one has proper thermal protection. At least, resistor R6 looks very much like a thermistor, both in appearance and in how it’s connected. Nice.
5. The output has four 470 µF, 25 V capacitors made by Aishi, as well as an output inductor to reduce ripple.
6. The output rectifier assembly is rated for 100 V and 20 A, which is a huge margin over the specified 1.7 A.


The capacitors are a little below their rated capacitance, although their ESR is actually quite good. But as I wrote earlier, “factory Chinese” is better than counterfeits pretending to be top-tier brands.


There are no components on the underside, but I’ll point out the large number of safety slots in the board—another point in its favor.


All right, let’s connect the power and give it a try.
The no-load output voltage is a little over 12 V, but I think we can safely ignore that difference :)


With a specified load current of 1.7 A, the power supply easily reached 2.5 A and even a little more. It shuts down instantly on overload, with no gradual transition.


A load test, combined with temperature measurements and a look at how the output voltage changes with temperature.
1. Load current: 600 mA, about 33%
2. 1.2 A, about 66%
3. 1.8 A, a little over 100%; I decided to round it slightly.
4. Since the power supply behaved itself, didn't shut down or overheat, I increased the load current to 2.4 A, or about 140% of its rated load.
5, 6. All the tests went very well. After almost an hour and a half of testing, I removed the load and measured the output voltage, then waited for the power supply to cool down and checked it again. The difference in output voltage between the “hot” and “warm” power supply was about 0.01 V, which suggests the feedback circuit resistors were chosen correctly—or simply that they are located well away from the components that heat up.


With no load, the power supply draws just 0.2 W; at 50% load, 13 W; at 100%, 25 W; and at 2.6 A (150% load), about 37 W.


Efficiency graph. Input voltage: 230 V; the values along the bottom correspond to the load current.


The overheating test followed the standard procedure I use for my reviews: several 20-minute stages, in this case at 0.6, 1.2, 1.8, and 2.4 A, for a total of 1 hour 21 minutes.


At a load current of 2.4 A, the highest temperatures were recorded at the transformer and output rectifier assembly, at about 70°C—well below their maximum ratings. In fact, I couldn't get this power supply to overheat, even at 140% load. I think the only way to trigger the protection would be to put the power supply in a compact enclosure with poor heat dissipation and load it to the same 140%.


Ripple was measured with the probes connected directly to the output terminal block. The oscilloscope's bandwidth limit was set to 20 MHz.


1. No load: about 25 mV peak-to-peak, not bad
2. 50% load: 20 mV, excellent
3. 100% load: 25 mV, excellent
4. 150% load: 35 mV, excellent, with one caveat that I'll get to later.
5, 6. Low-frequency ripple with no load and at 100%. Interestingly, the peak-to-peak value is higher with no load.


Here's the caveat I mentioned above.
Looking at the waveform above at 150% load, you can see that the high-frequency ripple is modulated by a low-frequency component.
A closer look with a different time base shows the ripple being modulated at around 25–28 Hz. I think I've seen something similar in Mean Well power supplies, if I'm remembering correctly.


This type of ripple only appears at loads above 2.3 A. More precisely, it's not there at 2.3 A, but is already noticeable at 2.35 A, and its amplitude only increases from there. But that's an overload condition.



Conclusions.
I have a few minor complaints: the capacitors have less capacitance than specified, and the heatsink for the output rectifier assembly isn't secured at all. Otherwise, it's all good: plenty of power headroom, low temperatures and low ripple, input and output filters, thermal protection, and generally neat construction.
In my opinion, this power supply is closer in quality to Mean Well products than to its Chinese “cousins,” so I can definitely recommend it.

This model is also available on AliExpress, in all output voltage versions. Unfortunately, it costs about $8.70 ($4.80 for the unit + $3.88 shipping).

That's about it. I hope you found the review useful. And as a reminder, I'm also interested in which power supplies you'd like me to review.
$6.5
Go to store
This page was found when searching for:
пульсация на осциллограмме, схема gk8ff-12vdc-s-a, как увеличить ампераж блока питания, nw 20a12 baaa, sm gfa20a схема подключения, источник питания бп gfa20a 12, 20.12 30 gfa, gfa20a 12, gfa20a, низкопрофильные блоки питания, блок питания низкопрофильный, низкопрофильный блок питания, блок питания 12 вольт, тест блока питания, технические обзоры, компоненты блока питания, устройство блока питания, блок питания отзывы, блок питания на 12 вольт

Questions and answers

How much load can the SANMIM GFA20A-12 handle?

Rated for 1.7 A, it also ran at 2.4 A, about 140% of its rated load. Overload shutdown occurred at 2.5 A and slightly above.

What ripple did the SANMIM GFA20A-12 produce?

The measured ripple was about 20, 25, and 35 mV at 50%, 100%, and 150% load, respectively. Low-frequency modulation appeared above 2.3 A.

How hot does the SANMIM GFA20A-12 get under load?

At up to 2.4 A, the transformer and output diode assembly reached about 70 °C. The author could not overheat the supply during a 1-hour-21-minute test.


This model is included in the summary table: Comparison of power supplies: ripple, efficiency, and temperature rise

The figures from this review are in the measurement database: Power supplies

Calculate manuallyAll calculators

Discounted products


You may be interested in


Discounted products


Support the project

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!

Comments2

Comments are published in their original language.

  1. 1 June 2019 03:50#1
    Спасибо за ваши труды.
    +3
    1. 1 June 2019 10:48#2
      Вам также спасибо :)
      0

Discuss on the forumHave a question or some experience to share? Discuss the article with the author and readers: the forum is a convenient place to share schematics and photos.
Go to the forum

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