[Elecraft] Re: KX1 mixer discussion

[Geoffrey Mackenzie-Kennedy]

Ron, AC7AC, wrote:

An interesting idea for noise blanking, Geoff: turning the L.O. on and off.
But keying an oscillator like that always produces sidebands around the
oscillator frequency. That has its own issues when that oscillator is used
as an L.O.

I wasn't aware of any dynamic range issues when using a noise blanker. The
big issue I've noticed is the appearance of the inevitable mixing products.
They are not normally a bother, but I have noticed an odd signal from time
to time when using the noise blanker and, sure enough, it disappeared when I
turned the blanker off.

------------------------------------------------------------------------
Ron,

You are quite correct about the 'key click' issue!! After 10 years I still 
have not got around to proving the reason, but I think that the 
characteristics of the mixer, the fact that its 'normal' LO injection is a 
well defined square wave from a flip-flop, and that it is a flip-flop being 
keyed rather than an amplifier or oscillator could be preventing the keying 
sidebands becoming an issue. I suspect that there are sidebands far enough 
down so as not to be noticeable to the ear nor cause operating problems, but 
might be seen by a spectrum analyzer at the mixer's IF port.

The dynamic range issue can arise in an 'in-line' IF noise blanker if any 
part of its circuit, such as an amplifier, forms part of the load presented 
to a preceding stage such as a post-mixer amplifier. In this case such an 
amplifier within the noise blanker would receive a large share of the 
wideband output from the post-mixer amplifier, so its IIP3 must be 
adequately high to avoid compromising the IIP3 of the post mixer amplifier, 
which in turn could compromise the dynamic range of the entire receiver.

BTW I use IIP3, input third order intercept, as a measure until a more valid 
measure of performance is agreed upon to take care of the cube law 
'misbehaviour' exhibited by some devices.

73,
Geoff
GM4ESD