On Mon, 10 Feb 2014 20:24:07 -0800, lw1ecp wrote:

Hi! Somewhere I read about unfortunate combinations of IF and LO
frequencies. Of course, you must make sure no LO harmonic will equal the
IF across its tuning range. Or f(antenna) = IF. Apart from these obvious
cases, there are more subtle combinations that are sadly discovered when
the receiver has already been designed and built. A famous example is
the birdie when tuning 910kHz with an IF of 455: twice the antenna
frequency beats with the 1365kHz LO giving also 455.
Can anyone point me to some article on how to predict undesirable cases?
I am planning to tune 14 to 14.3MHz with IF = 10MHz and LO = 4 to 4.3MHz
(commercial rigs would oscillate at 24 to 24.3). Am I going in a bad
direction?
Thank you!

The 1976 ARRL handbook had a nomograph. But it didn't get carried on.

There are computer programs, and you can do this with a spreadsheet or on
paper.

Basically, you figure out the receiver tuning range for all harmonics of
the LO, then you check those ranges against all harmonics of the passband
of the input filter.

Then you estimate the attenuation -- i.e., if you have a diode-ring mixer
with good symmetrical excitation then you can pretty much count on the
odd harmonics being attenuated by factors of 3, 5, 7, etc., (or 10, 14
and 17dB down) and the even harmonics being at least 20dB, maybe 30dB
down.

You'll kind of have to guess at the input frequency harmonics -- that'll
depend a lot on the distortion properties of your front end.

Then put it all together into a chart of possible receptions, and the
amount that they're attenuated.

Easy, isn't it?

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com