[Elecraft] L34 on K2 tuning question?

[Geoffrey Mackenzie-Kennedy]

David Woolley wrote:

> Geoffrey Mackenzie-Kennedy wrote:
>
>> for tuning L34 using the quote 'weak' 7 MHz birdie - more like a angry
>> elephant. After doing some work to reduce the number of K2  receiver
>
> It's not all that strong compared with sky noise with even a small 
> antenna.

To avoid any misunderstanding about my use of the term 'strong', the level 
of the internal spurious signal at 7 MHz in my K2's receiver was 
approximately 10db above the noise floor at the output of the product 
detector when viewed with a suitable spectrum analyzer, with the receiver 
connected to a 40m reference dipole at 70ft. The background noise here on 
40m is usually 'below S1' with either dipole or beam, storm noise excepted. 
Receiver IF bandwidth was set to 400 Hz ,which is not, of course, the 
receiver's noise bandwidth because of the actual response of the K2's IF 
filter.

In my book a receiver internal spurious signal of this magnitude is 
certainly not 'weak'!!

 > I am pretty sure that the birdy is a harmonic of the main
> microcontroller instruction rate and is entering before the mixer.

I agree that this is one source, and one point of entry for this signal.is 
K16 / K17.

>It's very unlikely that a post mixer birdy would move in the same way with 
>tuning setting.

Unfortunately there is evidence of multiple mixing due to the high level of 
signal from the various oscillators, VCO, BFO etc found right across the RF 
Board and elsewhere.This makes it highly likely that some products at IF 
will move in the same direction as the signal mixer's injection frequency.

 > What I would have liked is an explanation of the theory behind the
> separate noise and signal peaks.  The possibilities I can think of are:
> - some of the noise is coming in on the image frequency, but that
 <snip>

At the risk of oversimplifying, my thinking was as follows. If the birdy as 
observed was in fact the resultant of two signals, one of which had entered 
the signal chain before L34 and the other after L34, and the point at which 
the birdy was being observed was considered to be the summing point, then 
tuning L34 would result in a change in amplitude of the observed birdy. This 
change would be due not only to any change in the raw addition of signal 
power as L34 is tuned, but also due to the relative phase of the two signals 
at the summing point - which relationship would be changed to some extent as 
L34 is tuned. The noise power which determines the noise floor on the other 
hand is dominated by the 'legitimate' noise power coming down the receiver 
chain, and I have found no evidence that any part of this noise signal 
enters the receiver chain after L34. I should mention also that in my K2, 
the level of the VCO, 12.xxx MHz reference and spatially coupled BFO signals 
are high in the region of the two crystal filter

The situation becomes quite different if a legitimate input signal is used 
to adjust L34 because the effect of phase change when tuning L34 is taken 
out of the equation. I have no evidence that any product produced by a 
signal introduced at the antenna finds its way into the signal chain after 
L34 when the applied signal level is useful for alignment purposes. However 
at high levels of input signal this is no longer true.

>   I'm not sure why people should be saying that
>   one should be specifically optimising for noise above the birdy.

I believe that the birdy was being suggested in lieu of a signal generator. 
Elecraft's XG1 should do the job.

73,
Geoff
GM4ESD