These days we amateurs are blighted by electrical noise, from switch mode PSUs through to plasma TVs. There are principally two ways for that noise to be getting in to the receiver. Firstly along the mains (see my article on the mains filter) and secondly through the antenna system be it on the feeder or from the antenna.
While tons of ferrite can dramatically reduce the mains and common mode sources of interference that which is picked up on the antenna is harder to eliminate because the noise is an electromagnetic wave in the same way that the signal we are listening to is. This is where the MFJ 1026 noise canceller comes in.
The 1026 works by allowing the phase and amplitude of signals from two antennas to be varied in relation to each other. You connect your main receiving antenna to the connector on the rear and use the auxiliary whip antenna to receive the local noise. In most cases the whip won't be good enough and so an auxiliary antenna is connected instead. Ideally if you have local QRN then you want the aux aerial to have a higher proportion of unwanted to wanted signal. Using the phase and amplitude controls you can get the two antenna paths to render the noise in antiphase thereby nulling the noise but leaving the wanted signal intact.
Initially I was disappointed in the 1026 because it didn't seem to make much difference. Then I reread the manual and very carefully tried again. The setting up needs to be done slowly but it is worth the effort in trying different combinations to identify the optimum positions. Make a note of these settings so you can quickly return when you change frequency.
In answer to my question, yes it does work. I saw a reduction in noise of 4 to 6 S points between in and out of circuit.
Being an Arduino fan I would like to mechanise the switches and controls to simply instruct the canceller to adopt new settings.