Before doing serious work with these four analog noise reducers, I thought I should get to know them properly with some lab tests.
Generally, these four devices have similar connections and controls. Similar but different. You will find that each has connections for two antennas (Main and Aux) as well as an Output connection to the receiver. All but the ANC-4 provide gain or attenuation controls for each antenna. My ANC-4 only provides gain control for the Aux or noise antenna. NCC-1 adjusts antenna gain with stepped attenuators and a balance control. All have phase adjustments, although the X-Phase really just does a combination of phase and balance.
In order to phase signals to create nulls, you need some control over both amplitude and phase of the signals you are mixing together.
All four cover all of HF, and the NCC does Medium Wave as well. Both the ANC-4 and MFJ-1026 have high-pass filters to reduce signals below HF, i.e. below 1.9 MHz you get a lot of attenuation to prevent overload or mixing products from local broadcast stations. With the ANC-4, you can bypass this MW filter with jumpers. Unfortunately, the MFJ-1026 requires circuit modifications if you want to use it below HF.
For my tests, I connected my TinySA as a 15 dBm signal generator into a power splitter, and fed two identical signals into each antenna input. My test frequencies were 1, 5, 10 and 15 MHz. I also did some additional tests with a 10 dB attenuator to reduce signal into the AUX or noise port. I did this to replicate real world conditions where noise probe antennas typically emit much lower signal strength than the main antenna.
Finally, I used my Siglent DSO to monitor and measure output from each device, using sine wave and FFT display. The FFT math provided direct dB measurement of noise reduction.
Analog Noise Reducers Lab Results
Here are the highlights. I may provide more details later, but suffice it to say all four analog noise reducers provided significant nulls when input signals were phased together. No surprise there.
Deepest nulls came from the NCC-1 as expected on all frequencies. On my MFJ-1026 I needed to kick in the preamp to get enough nulling amplitude on the AUX circuit. Nulls on the NCC-1 showed up at 40 dB or so, with the others coming in around 15-20 dB. Controls on the ANC-4 and MFJ-1026 were a bit finnicky, but much smoother on the X-Phase and NCC-1.
You have an increased risk of IMD on all devices except the DX Engineering box if you don’t keep control of gain. This is pretty easy to spot on the FFT scope display when hooked up to the analog noise reducers.
My main accomplishments with the workbench were twofold. First, I convinced myself that I actually knew how these devices were supposed to work. Second, the tests demonstrated they were working at least close to expectations. Each of these boxes performs differently. READ THE MANUAL FIRST!