"Stephan Grossklass" wrote in message
...
Pete KE9OA schrieb:

Another thing that can cause that (2 X 2nd I.F.) image is
2nd LO leakage coming back into the input of the 1st mixer.

Happen to have a little example with numbers? I still have difficulty
imagining this.

(see rant below)

Let's take a typical rx with 55.845 MHz 1st IF and 455 kHz 2nd IF, plus
a desired frequency of 10.000 MHz. Furthermore, 1st IF filtering be
perfect in suppressing 910 kHz images and a SW lowpass be present.
1st LO = 65.845 MHz.
2nd LO = 55.390 MHz = const.

Even if there is some 2nd LO leakage into the 1st mixer input, it should
be treated exactly like any other ordinary signal (with a signal level
much smaller than that of the 1st LO), i.e. mixed to an unexciting
10.455 MHz. Obviously for another signal to be observed as a +910 kHz
image we need to have it on 54935 kHz, after the roofing filter, but I
assumed this would be perfect. Alternatively, we'd need to generate a
10910 kHz signal in the 1st mixer and leak that back to the 1st mixer's
input (but how likely is that?). I guess I'll grab a piece of paper
now...

Stephan

You would think that it would be the case, but it just isn't..............my
first dual conversion receiver had a 1st I.F. of 10.7MHz and a 2nd I.F. of
455kHz. Low-side injection of the 10.245MHz 2nd LO at a level of +7dBm into
a diode ring mixer.
Upper sideband = 10.7MHz................lower sideband = 9.79MHz. I had a
910kHz rejection of only 35dB, which is pretty bad for this type of design.
I went through all of the calculations and had the folks at Piezo Technology
stuck. They never heard of such a thing. I tried different filter
termination impedance, tuned circits at 10.7MHz, atteunator pads to provide
a resistive termination for the crystal filter. It was out of desperation
one day that I tried a copper shield around the 2nd LO and the image dropped
down below the noise floor of the system. The offending culprit was WOPA
1490 interfering with WILL 580. WOPA was in Oak Park, Illinois, while WILL
is in Champaign, Illinois, at the University of Illinois.
Once I went to lower power 2nd LOs the problem hasn't shown up. Also, with
the selective diplexer I am using after the 1st mixer (BW =13kHz @ 6dB), the
additional selectivity helps out quite a bit.
When I told some of my co-workers at Collins Radio Co. about it, they hadn't
heard of this mechanism either.

Pete