[Elecraft] K3s and DEMI transverter

[David Olean]

Comments below.....

On 6/2/2021 9:32 AM, W8JH wrote:
> I am looking to add the DEMI (Q5) transverter to my K3s and have no
> experience with transverters other than the internal one in my KX3.  In the
> future, I may wish to progress to 2m EME as well.
>
> Here are some choices I need to make and I would appreciate any advice:
>
> Common IF vs Split IF.  I'm thinking split IF with a K3s?
I would advise using split signal paths.  All my transverters are rigged 
up for a 0 dBm 28 MHz signal level on TX.  I always plan a system that 
has power levels such that if you leave the exciter rf drive level at 
max by mistake, it still will not impact anything or blow anything out.
>
> Common RF or Split RF.  Common RF for one antenna connection?
>
> IF drive level of -20 to 0dBm vs 0dBm to +20dBm.  I think the K3s
> transverter output is 0dBm so I'm not sure what to think here.
>
> PTT polarity low instead of high I'm thinking.
PTT set for ground on transmit standard for all bands.
>
> High stability Local Oscillator vs. standard.

This is an individual choice as all crystals are not created equally. I 
have a K3 with the garden variety ref oscillator. It is more than stable 
enough for digital modes.  The problem is typically the LO in the 
transverter itself. Crystals drift as temperatures change. I  have a 10 
GHz transverter with a crystal oscillator and the drift from cold start 
to warmup is about 2 kHz at 10 GHz. After it is warmed up, I see only a 
few hundred cycles of ambiguity from day to day.  It was stable enough 
to make me not change over to a synthesized LO since sideband noise on 
the synthesizers was not so hot. Others have had more trouble with 
crystals drfiting all over the place. Not all crystals are created 
equally.  Cleaning the crystal blanks affects drift drastically. 
Production methods vary.

Recently, I made some digital contacts on 222 MHz and found that my old 
crystal oscillator in my DEMI transverter drifted a bit. I saw a 10-12 
Hz drift after each JT65 transmit sequence.  The problem is the power 
amplifier inside the trnsverter that heats up the crystal over time. My 
solution was to put a small 12 volt fan on the heatsink just above the 
hybrid amplifier location. It cured the drift. After employing the fan, 
people assumed that I was running a ref locked local oscillator. 
Measurements indicate a drift of just a few Hz over time.   At higher 
frequencies, (over 432 MHz) some sort of reference oscillator 
arrangement is needed for digital modes. Cheap 10 MHz sources are 
available on EBay. Just make sure that whatever ref locked oscillator 
you use, that it does not produce bad LO noise sidebands.  I have a 
proportional oven 10 MHz crystal oscillator made by Wenzel Associates. 
It was cheap on EBay, and more than stable enough for any digital modes. 
You do not need a rubidium standard at all.

>
> I could get a custom IF and I think there is some slight advantage to a
> 14mHz custom but I'm guessing the 28mHz standard IF is good enough?

I use 28 MHz on 50 thru 432. On 902 and above, I standardized on a 145 
MHz IF for everything up to 10 GHz.  24 GHz has double conversion.

73

Dave K1WHS

>
> Thanks in advance for any advice or references.
>
>
>
>
>
> -----
> 73,
>
> Joe, W8JH
>
> K3s, KPA 500, KAT 500 and  KX3 happy user.
> --
>