Print deze pagina

PA0JEB AC mains pollution monitor

Because I have a lot of QRM on my QTH, I wanted to inspect the AC mains pollution noise to see if there are any noise sources in my house.

To do this you need either very expansive test equipment or when you're happy with just a indication, you could build it your self like Wim PA0JEB did. Being Dutch  I decided to make the AC Mains pollution monitor designed by Wim PA0JEB which was published in the March 2018 issue of  Electron (the club magazine of the Veron). Below you find the building result and a few first measurements.

For this project I could salvage all the parts from my junkbox and stock.

The PCB was made by hand (milling away the un-needed copper with my Proxxon drill/grinder). It aint pretty but it gets the job done.

Boxing it up because working with mains can be deadly.

And after some first remote tests (switching it on a few feed away) and measuring the outputs with a DVM, I was confident enough to hook it up to my Rigol Oscilloscoop.

The osciloscoop showing the first measurement in my house. The yellow line should be a clean 50Hz sine wave, but there is still some noise. This has to do with using just a 1H coil with the 470nF capacitors making the PI filter cutoff to high.

Using the "PI-Type LC Low-pass filter" example on CircuitLab.com

 

and filling in the used values (L = 1H, C = 470nF) the frequency response confirms my expectations.

Fc = 1 / (pi x square root ( L x C) Hz  (note C = C1 + C2 for PI filters)

Fc = 1 / ( pi x sqrt(L x C)) = 1 / (pi x sqrt( 1 x 940E-9)) = 328Hz (-3dB point)

 

The article stated that a coil of serveral H was needed an Wim even used the primary side of the small transformer. When I measured a small transformer I had laying in my junkbox it indicated a value of  around 10H.

Filling in the used values (L = 10H, C = 470nF) in the simulation above I get this frequency response.

Fc = 1 / ( pi x sqrt(L x C)) = 1 / (pi x sqrt( 10 x 940E-9)) = 104Hz (-3dB point)

So I have to look for something bigger orhave to wind my own to get the correct response.  Or I could look for capacitors with a 4.7uF capacity.

The frequency response is the same as using a 10H coil. (L = 1H, C = 4.7uF)

Fc = 1 / ( pi x sqrt(L x C)) = 1 / (pi x sqrt( 1 x 940E-8)) = 104Hz(-3dB point)



The FFT presentation of the timedomain signal shows a peak at aprx 9.75 Khz. So there is some leaking around on the mains.

But the noise is well visible so I now can test variouse objects in house to see if they are generating noise on the mains. 

One thing I did see were repeating pulses which may indicate PowerLineCommuniction in my area. That something to look into.


Previous page: uBitX - videos
Volgende pagina: Tools