strange ripple on 3€ 12V 1A power supply

[webgiorgio]

This switching power supply from ebay  (CD-01218) has a severe noise on the output voltage.
The input is at 50 Hz, and the noise "blocks" are in phase with the supply voltage.
The scope has 10 ms/division, and 1V/division.

What could be the reason for this bad voltage output? attached photo of the oscilloscope

[HB9EVI]

I would say: typical CE (=china export) phenomenon. Lousy design. I had once one where the big cap on the DC side blew up when nominal current was reached, likely because of a combination of poor quality cap and too high ripple/noise on the DC.

I don't know if it really worth it, but you can try blocking some more noise with chokes and further caps, but you'll likely fast exceed the 3€ with that material

[Zucca]

Test it again under load... and see what happens

[webgiorgio]

With a 3W load it goes worst, with peaks at 17 V.
With 10uF in parallel on the load it improves, but still a lot of spikes (+-1V).
Sure it is a low quality Power supply. It was working fine earlier, powering an arduino. The the Arduino got mad, I checked the PSU and found it guilty.
Seems like a fault.

Several years ago I purchased 10 pcs of this power supplies from ebay, and they all fail after 1-2 years of operation.
The curious thing is that those "blocks" of spikes are aligned with the peaks of the 230V supply voltage.
Maybe someone recognise that "shape" of the signal as a "typical" fault?

[HB9EVI]

could be an underdimensioned cap after the rectifier, not only it is lousy designed for its proper purpose, it obviously misses all measures against EMI as far as I see; I wouldn't like on mains running in my shack.

you go the old way: if you buy cheap, you buy double in the end. it's not worth troubleshooting; in the bin with it and get something decent

[Cyberdragon]

Show us the underside and close ups. I bet there's a lot more wrong with it (safety wise) than just ripple.