[Elecraft] K3 Birdies

[David Gilbert]

I recently completed K3 #2623 and everything went very smoothly, except 
that I'm really surprised at all the birdies I can hear on the main 
receiver with the antenna input simply looking into a precision dummy 
load (or open circuit ... no difference either way).  20m is clearly the 
worst, with over 50 discernible birdies ranging from barely readable to 
S4 on the meter.  The majority are S0 and S1, but I had three each on 
20m and 40m that were S2 or louder.   They are definitely synthesizer 
artifacts since their pitch jumps in large steps as I tune across them. 

All of this is in the main receiver ... I can find only a very few 
really weak birdies in the subreceiver.    I didn't rigorously check all 
of the bands, but 15m and 10m are virtually clean, even on the main 
receiver, and 80m and 40m have roughly ten noticeable birdies each (main 
receiver only).

I had previously taken notice of the comments on this reflector and in 
the manual about shielding and good connections, so while building the 
rig I made a special effort to make sure all connectors were firmly 
seated and all ground screws were tight.  The mail list archives have a 
comment from one fellow who improved his birdie situation by 
recalibrating the synthesizers, but I tried that and it didn't make any 
difference.  Most of the relevant postings suggested adjusting cable 
positioning.

So I popped the lid off the K3 this evening and (using a thin wooden 
dowel) moved various cables around while listening to the birdies.  I 
was able to make most of them completely go away by positioning the 
cables in one place or another, but it was amazing to see how sensitive 
the placement was.  In the case of one cable, literally a couple of 
millimeters made the difference between an S4 birdie and ESP copy.  And 
it wasn't just making sure that a particular cable was close as possible 
to a grounded shield or furthest from some other point.  In some cases, 
the optimum position was a particular distance from something ... it was 
like tuning an L-C circuit.

While I was at it I double checked every cable connector and they aren't 
the problem ... wiggling them had no impact on the birdies.

I kept playing around with cable positioning until I got the best 
possible combination.  In most cases, nulling the strongest birdies put 
almost all of the weaker ones below the noise floor, but in a few cases 
nulling one birdie made a different one stronger. Of the birdies I could 
previously hear on 20m, roughly 30 are now either gone or barely 
readable.  Of the remaining ones, all except one are S1 or less.  I 
ended up with only one really objectionable birdie hitting S4 on 14186.6.

I'm pretty sure that one or two of the cables are going to "relax" and 
shift over time, so I'll probably see some of the birdies reappear, but 
in all reality none of them are likely to be disruptive.  On CW I can 
use a narrow bandwidth to make all of them go away, and on SSB the notch 
will do the same.  And since they all tune so much faster than real 
signals, it doesn't take too much to distinguish them and mentally 
ignore them.

Still, I'm wondering if this is normal behavior, or whether all these 
birdies might be a symptom of something else out of whack.  Comments 
appreciated.

73,
Dave   AB7E




-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.qth.net/pipermail/elecraft/attachments/20090210/66c38699/attachment.html