China goes RF

Feb 23, 2014 63 Replies

Some time ago I bought a 12V 5A adaptor to power my regulated lab supply on ebay, ebay item number 110948131685, about 9$, 2$ shipping. Arrived some time ago, tried it yesterday, and it worked OK..... Left he lab supply on, went to do something else, noticed my analog video camera had huge RG interference hum bands. Thought its supply was defective, replaced it, did not help. tried things, got the scope, 500 kHz all over the place. switched things off, still 500 kHz all over the pace. Suspected some outside source, but then it hit me: That new AC adaptor was still on. Unplugged it,. RF gone. Tried various loads on it, even the slightest load and 500 kHz RF all over the place, VOLTS, yes VOLTS. Held the scope next to the adaptor, WOW, a real long wave transmitter!! Curious, opened it, it has 2 screws hidden behind 2 rubber stops, and decided to draw the circuit diagram. The first thing you notice about this adaptor is that it is extremely light. The reason turns out to be an other almost core-less transformer. very minimal ferrite core. Here is the circuit diagram:

formatting link



Interesting is that they use a MOSFET, The surprising thing is the NTC, have not seen an NTC in the mains since the old Philips tube color TVs... I do not understand D1, it catches only negative pulses, I would expect positive pulses there. The 2 transistors form some sort of SRC, that can get triggered on 'high voltage' or on 'high current' via the MOSFET source resistor. The TL431 is used as reference and voltage level sensor. The Vbe of the NPN is used as current sense threshold.



Nothing new here, except China now runs at 500 kHz, interfering with anything and everything in the vicinity, including AM (1 MHz, 1500kHz), or whatever else picks up the RF, and that RF is several watts. These things should be forbidden,



Component view, small transformer:

formatting link



PCB view, they did not forget to put a track in between the opto pins, actually found a lose piece of solder too:

formatting link



Bottom view, easily opened:

formatting link



No screening anywhere. Have fun, cannot use it in the lab. Hope nobody in the neighborhood buys one... if so forget about low level millivolt measurements.


Have you not got a spectrum analyser to take a closer look at precisely what its emitting? Your handwriting is terrible, btw. ;-)

On a sunny day (Sun, 23 Feb 2014 15:07:15 +0000 (UTC)) it happened Cursitor Doom wrote in :

Yours too.

But greatly improving!

D1 sure looks backwards.

How does it start up?

I bat it has all of the UL/CE/FCC stickers.

John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation

Gotta love those 4 diodes on the secondary side all tied together in parallel! :)

Maybe they have traces that act as balancing loads, if that is the case, they would make for some nice radiators.

Can of conductive paint sprayed on the inside?

Jamie

Might work but it's very expensive[*] and you can't paint holes. Any seams also have to be sealed. The first rule of EMI mitigation is to not radiate in the first place.

[*] cost my PPoE $10/unit on a 5"x4"x1" product. It wasn't done for EMI, though. Passing EMI testing was a piece of cake on that part, though it was an intentional radiator (i.e. wasn't cheap).

On a sunny day (Sun, 23 Feb 2014 07:55:40 -0800) it happened John Larkin wrote in :

Starts up normally, that is right away.

Only CE.

On a sunny day (Sun, 23 Feb 2014 11:10:04 -0500) it happened "Maynard A. Philbrook Jr." wrote in :

No, just srips to conenct all together.

That would help.

But I think both the mains elcos and the output elcos are high impedance at 500 kHz, and especially the harmonics of that. No small filter caps anywere in parellel with those. No mains filter, no output choke, not even a zigzag trace.

No, it's a fake one.

Thanks, Fred.

On a sunny day (Sun, 23 Feb 2014 16:34:39 GMT) it happened Jan Panteltje wrote in :

Thinking about it, maybe they count on it to avalanche, that would create some more RF too perhaps.

Yes, but how?

John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation

CE means Can't Enforce.

John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation

If D1 is indeed backwards, the mosfet may be avalanching, which could be noisy.

I still don't see where the gate drive comes from.

John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation

The primary filter components before the rectifier bridge are totally missing - not a single chance to fulfill the EC requirements. Does it have the CE sticker (interpreted as China Export)?

Tauno Voipio

What is this, readable schematics? :)

I think the 1M gate to GND may be misplaced, double check?

Can you read what value the Y1 cap ('2kV' I think it's marked) is?

How minimal is minimal? A hundred watts at 500kHz doesn't need much, and it should be a gapped but otherwise full loop type ferrite core.

The main thing missing is a common mode choke on the line side. Though there probably aren't enough layers of tape between primary and secondary, leading to excessively large P-S capacitance, which requires larger Y1 type caps to shunt it away.

The circuit is probably not too terrible overall, just made as skimpy as possible. I'm guessing everything gets damned hot at rated load, if it even delivers that.

Tim

-- Seven Transistor Labs Electrical Engineering Consultation Website:

formatting link

I've seen these circuits too in wall warts branded by Jameco Electronics. The capacitor across the transformer as the only RF filter is the signature of this bad design.

I used a Jameco adaptor on my phone to replace a hot wall wart. DSL quit working and there was a loud hum on the phone line. The junk adaptor was leaking 10mA of line power plus a whole lot more RF power. I could actually make tiny sparks by grounding a DC output.

On a sunny day (Sun, 23 Feb 2014 09:47:59 -0800) it happened John Larkin wrote in :

Oh, see what you mean. Been trying to look up that MOSFET, no luck so far. maybe it has some Cds capacitance that causes some current in teh transformer, causing a voltage in the secondary feedback winding, that then is amplified, etc.

On a sunny day (Sun, 23 Feb 2014 09:54:56 -0800) it happened John Larkin wrote in :

From the C8 1k from the secondary. Thats should be a positive pulse, as it cannot go negative due to the diode. Basicaly an upside down rectifier.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required