Strange issue with wall DC adapter - ripple/ground...

[elcrni]

hi all,

I came across a strange thing i can not explain or fix (hobby electronics). I have a sensitive RF receiver, i've been working on it for some time now and it worked beautifully, been receiving and decoding a clean RF signal. I've been using old laptop AC/DC adapter for the whole time.
This morning i've set my receiver in an enclosure and decided i would go with regular 9V 2A wall adapter... then a strange thing happened, my RF reception was all over the place and instead of very clean and regular "blinks" of a LED every time signal goes HIGH (1Hz signal) now i have random and erratic blinks/signal HIGHs...
Then i took another similar dapter, also 9V 1V and everything works like a charm.... then, i've took 5V 4A adapter, removed DC/DC buck converter and again my signal is all over the place...

So in total, i have tried 6 very similar wall adapters, 2 work, 4 does not...

Then a strange thing happens, when i connect the one that does not work, and then take a jumper wire and connect GND of that adapter to the GND of my bench power supply, it works!

Every single adapter that produces chaotic RF signal reception, if i just connect the GND to bench power supply or even to a GND of another adapter, it works.

This led me to believe that there is something wrong with the ground rail on those "faulty" adapter.

I would really appreciate any hints or tips on what am i looking at here and or how to fix it.

Many thanks,
Alek

P.S. This seems very strange to me but could easily be something trivial or common :-)

[penfold]

It could be a matter of "common mode" noise, it's quite a big problem with typical wall-warts, normally a bigger problem with more compacted designs which doesn't allow the designer much scope for screening and isolating noise.

A mains-frequency transformer-based adapter may be the best option if you cannot tolerate a permanent earth connection. You could use a common-mode filter or ferrite ring on the power leads, but for the filter to be most effective you would need some earth connection.

[elcrni]

Thanks penfold!
not sure i quite understand what the "mains-frequency transformer-based adapter" could be :-)
Also, not sure what one adapter that works has that the one that does not work doesnt have? or this a design difference that i cannot easily solve...

what baffles me is that i have 2 very similar adapter, similar price, size... and one works while the other one doesnt. None of those 2 have a ferrite ring...

Many thanks,
Alek

[elcrni]

just reading about isolated and non-isolated adapters... is this what is making issues here?
and if so, is the one working isolated or non-isolated?

many thanks
Alek

[G7PSK]

Cheap switch mode power supplies very often have the filter capacitors missing and cause a lot of bother with RF interference to amature radio operators. a fair number of LED lamps do the same.

[penfold]

...
not sure i quite understand what the "mains-frequency transformer-based adapter" could be :-)
...
what baffles me is that i have 2 very similar adapter, similar price, size... and one works while the other one doesnt. None of those 2 have a ferrite ring...

By mains frequency transformer-based ones I'm thinking of the old-style big bulky ones which use an iron transformer designed for 50/60Hz rather than the modern switchmode ones operating at 100's kHz. There will be some capacitance between primary and secondary windings (because they are relatively close to oneanother) of any transformer, this allows a proportion of the primary voltage to pass through and appear on both the positive and negative output terminals (common mode noise). At 50 Hz this capacitance between windings doesn't allow anywhere near as much noise through as it does at 100kHz+ switchmode frequencies. There are a few things that designers can do to prevent this noise and between two adapters at a similar price point I wouldn't expect the results to be too similar, hard to predict from the outset which one would be better.

By earthing the negative terminal of the output, you provide a path for this noise current so that it doesn't interfere with your circuit so much.

[elcrni]

thanks guys!

I think i understand the issues much better now. I did some further testing with the scope and discovered the base of the problem.
So, i had one group of wall wart adapters that just produced erratic signal and another group that produced steady signal.
Those that produced steady signal still had issues with signal width and with the scope i have discovered what is happening.
Then, i took an old laptop adapter that felt much heavier and most importantly, it had an E-Type plug meaning that i had a connection to earth ground, which seems to be very important for sensitive RF circuits.

Bellow is the image of how my signal looks:

A: wall wart that produces steady signal but with bounces from time to time, that bounce is then interpreted as eg 2ms signal received instead of 200ms.
B: laptop charger with E-Type plug produces steady, stable and bounce free signal all the time.
Since it always works when i connect ground to my scope, or to my bench power supply, i have figured this should be the issue, earth ground, and seems i got it right.



E-type plug



At the end, seems like regular wall warts are not going to do for this project but luckily lots of relatively cheap laptop adapters with E-Type plug available.

Many thanks,
Alek