Thursday, November 05, 2020

Antenna noise canceller working well against bad switching supply noise on 40m.

HF noise is an increasing problem that plagues hams in the suburbs. Switching power supplies, solar inverters and plasma TVs are all culprits.

Here I'm plagued by occasional bad interference on 40m (but really all over HF) that I think comes from a switch mode power supply used by a neighbour. I've just built the antenna noise canceller by Terry, VK5TM and it seems to work very well.


Noise cancellers work by taking a sample of the noise, inverting the phase, matching the level with the signal and mixing them - thus cancelling the noise but leaving the signal. Here's another example where I switch it in and out:


Here's the part of the circuit that controls the phase of the noise signal that is then amplified and mixed with the main antenna signal:


I bought the board and kit of parts (AU$55) and it is a very professional package. The parts are well presented and I found none missing.


The board is excellent quality.




The kit went together smoothly following Terry's instructions. The only fault was mine, when powered up the relays didn't click even though the transmit LED came on - I had missed a solder joint on one of the coils.

I purchased a metal box from Jaycar, part HB5442, 120x61x102mm and the project fits well with plenty of room for wiring.


To effectively cancel noise you need a sense antenna that picks up the noise but not the signal. In my case I'm using a 20m dipole on the fence line as the sense antenna and an end fed for 40m for the signal. I think that having a sense antenna that doesn't pick up the band I'm wanting to receive might be a reason why it's working so well for me.

In my case the noise level on the sense antenna is very high and I have to set the gain control almost at the bottom of the range. I might try installing a smaller sense antenna on the fence specifically to pick up the neighbour's interference. The two phase controls are quite sharp and together can be used to null the noise very well. Signal strength is reduced but can be boosted with a pre-amp, the main thing is that the signal to noise ratio is greatly improved.

I was tempted to buy the MFJ-1026 for US$239.95 + shipping, but Terry's kit works well and if you can do a little soldering I recommend it.

Note that antenna noise cancellers do not work in all conditions. The best scenario is where there is a single source of noise that can be picked up close to the source. I understand that broadband noise may not be able to be cancelled but in my case the broadband noise I see between the switching supply bands are cancelled nicely.

Update: I'm having doubts

I've been playing with noise antennas and I'm starting to doubt that this is working properly. I think I get some noise cancellation but it's only when the main antenna gain is fully reduced. Bear with me while I experiment further.

Update 2: All good

It's all about the noise sense antenna. Originally I used a 20m dipole but the noise level was extremely high. Next I tried a short vertical connected via a balun - that's when things got weird. Finally I've simply clipped a meter or so of wire to the centre of the coax and the controls on the noise canceller work as expected.

Noise is reduced by playing with gain and the two phase controls.

No comments: