GE Superadio III AM/FM radio Image Rejection Intermodulation on MW

[gmcjetpilot]

This is a general topic of strong-near AM/FM stations and specific to the GE Superadio III single conversion superheterodyne.

I am not sure if this falls under the topic of: image channel interference or IF breakthrough, co-channel interference, intermodulation, cross-modulation? I've been fighting a wicked Image from a strong station on 850 Khz. Day time the station has an Image below it, at approx 780 Khz. The only stations adjacent to 850 Khz I can get reliably by day is 750 Khz and 910 Khz. At night I can get 780, 800, 830, (850), 860, 900 and so on. The improved performance at night is of course the stronger night time signals, but also the 850 Khz station drops power at night

When the strong/close BCB station goes to night power it helps. The Image at approx 780 Khz is almost non existent, if I rotate the radio so the large internal loop stick antenna is null. The 780 Khz station has some superheterodyne squeal I can't tune out. The distant 860 Khz station gets some intermodulation or cross talk from 850 Khz.

The difference from 780-850 is 70 Khz. I can see where 455 Khz x 2 = 910 Khz. The difference of 910 Khz and the strong station at 850 Khz is 60 Khz.... this might explain the ~780 Khz image?

Also I have some strong FM stations at 93.9 Mhz and 96.1 Mhz and at night (not day) it seems there is an "Image: at about 88.5 Mhz and 90.5 Mhz respectively. Keep in mind these are approx Image Freqs with the analog dial. So the difference is about 5.5 Mhz.

The images do not seem to match the math for image issues?
Is there a way to align the radio to improve this?
Is there a filter or modification, more "pre-selection" to improve image rejection?

 

[fg]

This could be an explanation:

Tuned at 780 kHz, the preselector might fail to reduce the strong 850 kHz signal sufficiently. The mixer (or any RF amplifiers ahead of it) then would get overloaded, generating the 2nd harmonic of the latter signal -- which would be at 1700 kHz.

If the local oscillator runs 455 kHz above the 780 kHz frequency, the L.O. frequency will be 1235 kHz. This 1235 kHz L.O. signal then would mix with the 2nd harmonic of the 850 kHz signal and generate a modulated signal at 465 kHz (this could be closer to 455 kHz if the numbers aren't accurate). So what we have here probably isn't an image rejection/"double spotting" problem (it can't be, since the 455 kHz I.F. frequency is much greater than 50 % the frequency displacement between 850 kHz and 780 kHz).

Provided the analysis is correct, improving the front-end filtering and/or the dynamic range of the mixer and that of any RF amplifiers preceding it should cure the problem. Also consider adding a wavetrap for the 850 kHz signal.