Honus wrote:
I'm receiving WWVH at 910 KHz below their listed frequency of 10000 KHz.
After a bit of research on the net, I've discovered that this is a "harmonic
image" that isn't being filtered out by my single I.F. design radio. What
does that mean in plain English?

Actually, it's just an "image". A "harmonic" is something else.

This will get a bit long but will hopefully be clear...

In the early days of radio, when you tuned your radio to 10000KHz, there
were several amplifier stages, all operating at 10000KHz, to increase
the strength of the WWVH signal until it was loud enough to drive the
speaker.

If you decided you'd rather listen to Radio Japan on 9525KHz, you had to
retune *all* of those stages from 10000 to 9525. At the time, each
amplifier stage required a different knob; you might be adjusting four
or five controls every time you change stations. Also, high-frequency
amplifier circuits above 2000KHz or so were a lot less efficient and
stable than low-frequency circuits.

So Edwin Armstrong invented the "superhetrodyne circuit". In this
circuit, when you tune your radio to 10000KHz, you tune a "local
oscillator" to 10000+455=10455KHz.

The output of this local oscillator is mixed with the signal coming from
the antenna. You get four outputs from this mixing process:

- 10000KHz, from the antenna
- 10455KHz, from the local oscillator
- 10455+10000=20455KHz, the sum of the two
- 10455-10000=455KHz, the difference of the two

These four outputs are then passed to an "intermediate frequency" (IF)
amplifier, tuned to 455KHz. The signal is amplified here, then
"demodulated" to audio and sent to the speaker. Selective circuits in
this IF amplifier reject anything that isn't 455KHz; the 10000, 10455,
and 20455KHz signals are removed, and all you hear is WWVH.

Next, you decide to change frequency to 9525 for Radio Japan. You
retune the local oscillator from 10455 to 9525+455=9980KHz. (again
assuming my math is right!) Now, your four outputs from the mixing process:

- 9525, from the antenna
- 9980, from the local oscillator
- 9980+9525=19505, the sum
- 9980-9525=455, the difference

You don't have to tune the IF amplifier: it's already tuned to 455KHz,
and now it passes the Radio Japan signal instead of WWVH. Whatever
station you tune, its frequency is "converted" to 455KHz, and amplified
in the IF amplifier, which is always tuned to 455.

Now, let's say you decide to listen to Radio Slobovia on 9090KHz.
(910KHz below WWVH-10000) You tune the local oscillator to
9090+455=9545KHz. And you get your four outputs:

- 9090, from the antenna
- 9545, from the local oscillator
- 9545+9090=18635, the sum
- 9545-9090=455, the difference

And you hear Radio Slobovia. But... there's nothing to stop WWVH-10000
from getting to the mixer. So you also get:

- 10000, from WWVH on the antenna
- 9545, from the local oscillator
- 10000+9545=19545, the sum
- 10000-9545=455, the difference

You have two different 455KHz signals! One comes from 9545 mixing with
10000, the other from 9545 mixing with 9090. You'll hear WWVH on 9090.

In practice, a receiver should have a "preselector". When tuned to
9090, this circuit should prevent the WWVH signal on 10000 from getting
to the mixer. If it doesn't get to the mixer, it can't mix with 9545 to
make 455. But preselectors aren't perfect, and if WWVH is strong enough
enough will "leak through" to be easily heard.

The difference between 9090 and 10000 is relatively small, it's hard to
make a preselector that can knock the unwanted signal down far enough.
Higher-quality receivers use an intermediate frequency much higher than
455KHz. For example, if the IF was 9000KHz (a common figure) then when
tuned to Radio Slobovia on 9090, the unwanted response would be on
27090KHz. It's much easier for the preselector to tell the difference
between 9090 and 27090 than it is between 9090 and 10000!

Is this 910 KHz harmonic only going to show up -below- the actual broadcast
frequency, or will it sometimes appear 910 KHz above as well? (I imagine

Only below.

As you might guess, it's possible (and common) to design the radio to
have the local oscillator 455KHz *higher* than the desired signal, in
which case the image would always be *above* the actual frequency.

Are these harmonics similar to radio harmonics?)

Yes, but what you're hearing isn't a harmonic. (there *are* harmonics
in radio. For example, you might hear Radio Slobovia on 18180KHz when
they're actually broadcasting on 9090. This kind of harmonic is usually
- but not always - the result of a problem at the transmitter, rather
than receiver design.)

And why is this
"image" so strong? My reception at 9090 KHz seems to be as good as that at
10000 KHz. (Which admittedly isn't great.)

Your radio's preselector isn't very good. Indeed, it's possible it
doesn't even have one. (this tends to be the first thing designers
leave out when they're trying to cut costs...)

Are there other harmonics of any given frequency that I haven't stumbled
across? For example, if I'm receiving WWVH at 910 KHz below their
broadcasting frequency, will I ever get them at say, 1820 KHz (910 X 2)
below?

Your particular example won't happen. However there are other spurious
receptions possible. For example, the local oscillator in your radio
will have harmonics. When the oscillator is tuned to 9545 to listen to
9090, it will also have some output on 9545*2=19090. And if there's a
strong signal on 18635, you might get:

- 18635 from the antenna
- 19090, the second harmonic of the oscillator
- 18635+19090=37725
- 19090-18635=455

and you'll hear the 18635 station too.

How common are these images?

If you hear it on one frequency you'll probably hear it elsewhere. It's
inherent in the design of the radio.

If there is a preselector, it's probably more effective on lower
frequencies. You may be less likely to hear these images on the 49m
(6000KHz) band and the 540-1700KHz AM broadcast band.

--
Doug Smith W9WI
Pleasant View (Nashville), TN EM66
http://www.w9wi.com