[Elecraft] K2 FM BCI - one step closer (and one step back)

Sun Jun 9 14:30:01 2002

One can easily simulate the FM broadcast interference by using a
grid-dip meter as a transmitter at around 100 MHz. It leaks right into
the K2 in the 20 m band just as has been reported for FM stations. No
wires attached except for power, no covers opened. Ferrites on the power
supply connection has no effect. Turning on the preamp makes it a little
stronger, but not the 12 dB or 2 S-points or so that it is does at HF.

With my old Heathkit grid-dip meter, it is enough to put it next to the
K2 on either side. I then put a 56 ohms resistor on the end of a coax
cable from my signal generator at its max output of 0.1 V rms. This is
not enough not be noticeable with covers on the K2, but fine for tracing
with covers off. It turns out that the most sensitive parts are right in
front of the mixer, in particular RFC12 and the area adjacent to it (5
LED bars lit). Also the two relays K16 and K17, and the 20 m relay K4
are sensitive (3-4 bars lit).

Frequencies: 99.9 MHz FM modulated, 100 kHz deviation, 1 kHz tone, is
noticeable at several frequencies in the 20 m band, but in particular at
14.077 - 14.078 MHz. This is as predicted in my email from May 30 and
indicates mixing with the 5. harmonic of the VFO. At this setting it is
also sensitive at 90.04 MHz which is the image frequency. Using my
grid-dip meter I get sensitivity at about 81, 71, 62, 52, and 43 MHz
also (accuracy assumed to be +/- 0.5 MHz or so), they can probably be
computed from mixing with 4., 3. harmonic and so on and their images.

But what picks up the signal? I tried to replace RFC12 with a 1k
resistor, little change. I then wanted to ground the input of the diode
mixer TUF-1, soldered a strap turned on power .... and the lights went
out. I had connected it to 12 Volts instead - no more mixer - input
transfomer burned off!!!!! - end of experiments!

So here are results for someone else to continue; either a component
that acts as an antenna for a magnetic field (the covers stop electric
fields I think?) has to be replaced, or the VFO signal has to be cleaned
up even more.

73, Sverre

------------------------------------------------------------------------
--------
Sverre Holm - LA3ZA [email protected] 
Thu May 30 03:04:01 2002 

Could there be some undesired mixer products in the K2? 

Say that the BC signal enters the K2 by radiation after the input LP/BP
filters, and that there is a small 5. harmonic in the oscillator. This
will mix almost perfectly with the broadcast signal. The following
calculation could indicate what goes on based on the K2's IF of 4915
kHz.

Displayed frequency                  14082 kHz
Oscillator frequency: 14082 + 4915 = 18997 kHz
5. harmonic of oscillator            94985 kHz
Mixes with 99.9 MHz to give an IF of  4915 kHz

A bandwidth of +/- 50 KHz for the interference translates to 1/5 in the
setting of the K2 dial, i.e. 14082 +/- 10 kHz which almost takes us down
to your observed worst frequency of 14068 kHz.

A similar calculation near the 40 m band and the 8. harmonic gives 6958
kHz as the worst frequency, and in the 80 m band 3720 kHz and the 11.
harmonic could give this effect. If this is the cause, most likely the
even harmonic effect (40 m) would be weaker than the odd harmonic effect
in the 20 and 80 m bands.

Does this make sense and correspond to anything observed?

73 and good hunt, 
Sverre

Sverre Holm, LA3ZA
Asker, Norway (59.8N, 10.4E)
---------------------------------
www.qsl.net/la3za