logo elektroda
logo elektroda
X
logo elektroda

GaN 65W power supply/charger bought for four quid from Aliexpress - tests, thermal imaging, user ris

p.kaczmarek2  4 273 Cool? (+5)
📢 Listen (AI):

I invite you to test the boisterously advertised welcome gadget offered to us by AliExpress for just £4 with free shipping. Here I will check what performance this power supply has in practice, whether it really supports QC and reaches 65 watts, and whether the voltage at its output is safe for our mobile devices. In addition, I will also look at its PCB under thermal imaging to determine what heats up the most.

Let's start with the purchase. The whole thing comes under different names, to me it popped up as a "GaN wall charger with USB C plug". I already have reservations about the name itself, if only because essentially the charging circuitry is already in the phone, and here we have, as it already is, a power supply, although I'm more concerned about that GaN and 65 W....

Fast charging (QC standard - higher voltage?) also caught the title. Well, hard, we buy:

If problems arise, simply open discussions with the seller.

First impression First impression
I waited a long time for the package to arrive, but here it is - there's no mistaking it, it says 65W on the packaging and case too:

On the device we have its performance (?) - up to 7.2 A at 9 V, up to 5.4 A at 12 V and up to 3.25 A at 20 V:

All nice, but QC trigger can't even switch on 7V - what's up?

Let's not judge in advance. The QC trigger may have broken down for me, perhaps it's run-of-the-mill. Worse, at 5 volts I can also draw up to 2 amps and then the voltage drops:
USB tester showing 5.23 V and 1.127 A connected to white power adapter and PCB. USB tester shows 3.593 V and 2.563 A, connected to a disassembled USB-C power supply.
One might get the impression that we have a 5 V 2 A power supply here, but this remains to be checked.

Interior of the gadget
The whole thing is suspiciously easy to disassemble. I can get the impression that this device does the autodisassembly itself when plugged in and pulled harder at an angle.
Disassembled USB-C charger showing internal PCB and loose power plug
This is how the plug is connected to the board - on contact. Inside we have a small PCB with a load bonus:
Disassembled plastic USB power adapter with exposed internal PCB and plug
Here's the PCB itself. Everything is on a poor, single-sided board. It does not resemble a GaN power supply:
Close-up of power supply PCB with electronic components and USB ports
You can see by the pins from USB C alone that only 5 V is connected here. There are no derived signals responsible for QC negotiation.
Close-up of power supply PCB with capacitors, USB ports, and electronic components
I don't see anything else on the underside either - all we have here is a PSR (Primary Side Regulation - no optocoupler) flyback converter controller, a synchronous rectifier (on the secondary side - instead of a Schotky diode), and a small circuit to probably let the power supply report to the phone to let it know it can draw more than 500 mA.
Power supply PCB with visible ICs, resistors, and solder joints.
We only have a fuse at the input, and then there is a rectifier bridge. Then the minimum inverter circuit:
Close-up of a single-sided PCB with electronic components from a USB charger.
You can easily read that the inverter is based on the MT3612:
Close-up of MT3612 chip on a power supply PCB with SMD components
This fully dispels any doubts - according to the catalogue note, we have here up to 10/12 W depending on the version:
Document excerpt showing MT3612 chip specs and a table of model variants with power ratings.
9bb358427
Now we look at the secondary, low-voltage side.

There is only one more modern aspect here, as we have an MT6704 synchronous rectifier (based on a MOSFET) instead of a Schotky diode:


Practical tests
To start with a current of 1.5 A:

After half an hour the main PSR controller approaches 50 °C, the synchronous rectifier heats up a little less.

This is very strange, but after a momentary load of 1.5 A the voltage starts to jump to 6.5 V.



What happens, on the other hand, when we increase the current?
Electronic module with LED display showing 6.59 V and small cooling fan
At this load, the systems exceed the 50 °C limit.
Thermal image of PCB with marked hotspot reaching 55.3°C Thermal image of PCB with highlighted area reaching over 50°C
The voltage is also able to occasionally drop to 3.4 V, i.e. the overload protection is trying to protect the inverters.

Summary
This product is outright misleading and does not deliver what is promised. What's more, the whole thing is of poor quality, and after a load of about 3 A the voltage starts to jump in a strange way, up to 7 V, and the overcurrent protection does not always immediately engage.
In addition, the construction is flimsy, I would be afraid that at some time of pulling out of contact a part of the device will be left in my hands.
I wonder what the construction of the transformer is like, but I didn't want to destroy it to find out.
All in all, it is hardly surprising that I would totally advise against buying this product. If this is what a "welcome deal" or a "welcome gift" is supposed to look like, then I dread to think what the actual "cooperation" will look like. The shop is still promoting it and non-technical people are ordering and getting fooled.
Have you also come across this kind of electronic trap in online shops?

About Author
p.kaczmarek2
p.kaczmarek2 wrote 13909 posts with rating 11690 , helped 629 times. Been with us since 2014 year.

Comments

E8600 31 Jan 2026 16:47

Most importantly, they did not forget the rather gratis 'plasticine'. [Read more]

Krzysztof Kamienski 31 Jan 2026 17:30

Plasteline is added for two reasons. 1. To increase customer confidence in the product. 2. With loose mains sockets, gravity improves contact. [Read more]

%}