Enico MY-050500A charger

I decided to take this charger apart when I noticed some weird stuff happening with my phone when connected to the charger (touchscreen not working properly - probably high EMI from the charger).

Specifications, tests and measured parameters

  • Rated input voltage & frequency: 100-240V 50-60Hz
  • Rated output voltage: 5V
  • Rated output current: 500mA
  • Model no: MY-050500A (there are more chargers ending with -050500A that look similar and are also unsafe)
  • Measured output voltage: 5.19V (open circuit)
  • Measured output votage: 4.93V (10Ω load)
  • Inadequate isolation distance.
  • Very high EMI/RFI and extremely high ripple.
  • Measured ripple (after fixing the Y capacitor problem): short peaks still over 400mVpp
Measurement equipment used: UNI-T UT33D multimeter, Tesla BM 574 oscilloscope.

Mechanical build

This is just a generic USB charger that plugs into an European socket and has an USB output on the side.


The charger
The charger (click for full resolution)

The following photos are color corrected to make them somewhat easier to view:

Inside (click for full resolution)

Circuit board, top side
Circuit board, top side (click for full resolution)

Circuit board, bottom side
Circuit board, bottom side (click for full resolution)

The electronics

Schematic (click for full resolution)

Unlike many cheap chargers, this charger is not based on a self oscillating circuit. A special chip is used here, marked "DOCN DC1805A CG7P1C 1HJ". I haven't managed to find a datasheet. This IC is in an 8-pin surface mount package.

The distance between primary and secondary sides is in the ballpark of 2.5mm, which seems a bit too low for European safety standards.

The Y capacitor between the primary and secondary side (marked CY1) is also missing. The charger makes huge amounts of electromagnetic interference, interfering with my amplifier when any longer cable is connected to the USB output. It also interferes with the (capacitive) touchscreen on my phone. This capacitor is supposed to short the high frequency noise caused by spikes from the primary being capacitively coupled to the secondary coil, which causes problems with many devices. These capacitors are usually in the low 100s of pF range. I added a proper capacitor in place and it solved the problem. It must be a proper Y-rated capacitor for safety reasons!

The output capacitor (C3) is marked "Low ESR" and the ripple waveform seems fairly strange, it's mostly quite low and looking similar to a sawtooth-like wave with some distance between impulses (this is most likely a discontinuous mode flyback power supply), but there are big spikes/ringing on the positive-going edge of the "sawtooth". The distance between the positive pin of C3 and the rest of the circuit is fairly high.

Notice the "fuse" on one of the mains inputs, which is just a thin trace that is apparently designed to blow when too much current passes through.

There is also no proper filtering or inrush current limiting on the mains input


This is a product that apparently violates several standards and causes a lot of interference. It's hard to say how good/bad the insulation of the transformer itself is. To me it definitely doesn't seem like a well designed product. However, it's still better than many other Chinese chargers.


Advertisements (from webhosting):