Re: [FUNcube] M1GEO HF Upconverter Design
- Hi George You seem to have the test equipment. .....have you measured the
IMD products of the combination?? I can't believe the funcube will work all
that well being hit by a not very reduced 100MHz LO?? -50dB LO feed through
from a +7dBm LO means hitting the FCD with about 20mV , and the IMD
performance of the FCD is adequate but not too good. Just a thought rather
than be disappointed .
----- Original Message -----
From: "George" <george.smart1@...>
Sent: Sunday, September 25, 2011 11:22 PM
Subject: [FUNcube] M1GEO HF Upconverter Design
> Just to anybody who is interested, I have designed an up-converter for HF
using readily available parts. I have also published the schematic & PCB
for anyone wishing to make one.
> I intend to add diode-switching to a later design, allowing the
up-converter to be switched in and out electronically.
> Just thought my work may save someone else the effort.
> George, M1GEO.
> Yahoo! Groups Links
>...I have designed an up-converter for HF using readily available parts.Hi Georges
May I suggest you a small and very stupid enhancement to your nice "universal upconverter" ?
I build something very close to your design. I just added a "smoothing" Pi low path filter" on the oscillator output to reduce harmonics due to the "square" nature of the signal, and added a very soft high path on the output as you mentioned it on your blog. But this is not the point.
I used a 3.3 V oscillator and was unable to hear anything on the output of the transverter. Even with a signal generator injecting more thant -40 dBm on the input filter...
After some search, I found that the level of the LO input at the mixer input was 0.2 dBm... far too low. But when I disconected the mixer's input from the OL trace, the level of the oscillator jumped back to it's original power(around 12 dBm). I discovered that some oscillators cannot suffer a electrical grounding of their output (and the transformer of the LO input acts as a DC grounding)
It took me half a day to find this gremlin. The solution is quite simple : just add a series 300 pf cap (or whatever) after the attenuator and before the mixer, it will cut the DC junction and you'll find your 7 dBm signal back again.
Even if most TTL oscillators don't need such a precaution, adding this cap won't cause any trouble... and will avoid newcomers to search the reason why this +@!%**ing oscillator is down.
Thanks for sharing your experience, Georges
Hope this will help