A Tail of Spectrum Pollution.

Last autumn I installed a new antenna system for two metres and immediately notice a huge rise in wideband noise and birdies across the whole two metre band, other bother!

As with all these things I tend to go at it like a bull in a china shop and only afterwards realise I should have taken some measurements of noise or even of the local beacon, oh bother!

Well the first thing to do is turn the mains power off to house and take a quick look a the waterfall on the rig compared with it on.

Mains switched off.

And with the mains switched back on again.

Mains switched on.

Wow! Ok I wasn't expecting that.  

Next thing to do is turn the mains off again and slow turn each circuit back on by the fuse box, in my case everything stayed looking good until I hit the sockets that power all my AV equipment and WIFI / Lan access point at which point all the birdies flew back in. ;-)

Next up I switched off all the AV equipment and the router and slowly switched each item back on the first item happened to be the access point and low and behold the waterfall filled up again.

I then unplugged everything from the access point other than the power supply and again the interference vanished so wasn't being caused by the WIFI, SMPS or the unit itself.  

Now I should mention I run CAT5 cabling to various devices such as the media server, sky box etc ... 

For each CAT5 cable I wound 6 turns around a type 43 ferrite torrid, available from the RSGB shop,  at both ends of each CAT5 cable in turn checking each time if the birdies had reduced.   

What I found was any cable run over half a metre in length need to have ferrites fitted and with the longer run cables, in excess of  a metre, need two of more ferrites at each end to get enough impedance to reduce the birdies. 

After what seemed like an extraordinary length of time, hey these things take ages to do from a wheelchair, I finally managed to get a handle on the birdies to the point they're almost no existent. 

Noise after ferrites fitted.

Ok this is looking good a lot better than it did originally, and again being a wally I should have a taken a measurement of the overall noise floor and not just glance at the waterfall.

So I next setup SpectrumLab to log my local noise pollution towards the West for a few days to measure the general noise level. 

SpectrumLab with Plotter logging Spectrum Noise.

To calibrate the system I took a long plot to start with of a calibrated 50ohm load @ 290K with a 2.7kHz bandwidth.

A 50ohm load at 290K is equal to -174dBm/Hz so in a 2.7kHz bandwidth -140dBm.

I only I had a couple of days measurement I wrote a quick Python program to display the results in dBm's and S-Points.  

   

Now S-Points, a totally bloody useless measurement of anything if you ask me. ;-)  I had a quick look at S-Meters on Wikipedia and there appears to be two IARU technical recommendations for S-Meters one for HF defined as -73dBm for S9 with a 50 ohm impedance and another for VHF as being -93dBm for S9 again with a 50ohm impedance system  both with 6dB increments between points. 

So I did two plots one with the HF specification 

Scaled with HF S-Meter Specification.

and another with the VHF spec.

Scaled with VHF S-Meter Specification.

The first thing we can see if the chart using the VHF S-Meter specification looks horrendous while the HF doesn't look that bad and I'm inclined to go with the HF chart instead.

The second thing we can notice is at around 15:00 everyday and until 08:00 the following morning there appears to be an outside light being switched on which is raising the noise floor by 6dB so that'll  need to be tracked and sorted out.   

I still need to take some measurement with the house mains off again just to see if the noise floor drop at all but I do now feel as if I getting a handle on situation even if there's plenty left to do.