[TenTec] Argosy 525 - three different RX mixers?

[Dr. Gerald N. Johnson]

On Thu, 2007-04-19 at 17:02 -0700, Kevin Purcell wrote:
> My point was that you aren't going to get a better match from 4 HP  
> separate diodes (which may or may not come from the same batch) than  
> the integrated ring so that mod is not worthwhile

Depends on the integrated ring's properties. Balance isn't the only
thing, dynamic range, insertion loss, and noise are factors in the
radio's performance, only indirectly affected by balance. If the
integrated diodes have inherently high noise compared to discrete
Schottky, then the integrated diodes will mask the RF stage noise. If
the integrated diodes have a significantly higher ON resistance the
mixer will have added insertion loss. Both higher noise and higher
insertion loss make for a noisier receiver. Excessive unbalance can lead
to more of the local oscillator's phase noise reaching the IF, but even
from separate batches, I've not detected a great deal of unbalance in
Schottky RF diodes. I'd prefer to use them from the same envelope and
for sure the same part number and maker, to have a chance of balance.
> 
> Do you agree?

I guess not.

I see advertisements from IC makers almost every week in microwave
magazines proclaiming their high dynamic range low noise mixers and when
I read down their specifications I find them showing a NF of some 10 or
12 dB but sometimes insertion gain and an input IP3 of + 10 dBm. Shoot,
the single diode 1N82A had a 12 dB NF back in 1956 for a UHF TV
converter, getting that NF for an IC 50 years later isn't an
improvement. Granted the IP3 is a lot better but 12 dB isn't LOW NOISE!
Passive mixers from MiniCircuits often show a NF of 5 to 6 dB with that
same 5 or 6 dB insertion loss.

If I was going to attack a discrete component ring diode mixer, I think
I'd toss it out to the transformers and drop in a $5 MCL mixer where the
transformers and the diodes were made to work together by a maker with
lots of experience because the radios I've seen with the radio maker's
transformers on the mixer tend to perform poorly. I recall such a mod
for the Heathkit SB-104 which had discrete transformers and diodes that
claimed to work significantly better. At least in the MCL mixers, the
diodes and transformers have been tested together.

But not having a 525 and finding the performance of my Corsair II
decent, I'm not ready to proclaim that scrapping the discrete or
integrated diode mixer with transformers to install a MCL mixer is the
optimum mod, but if I was unhappy with the original part's performance,
it wouldn't be hard to cut two signal plus LO traces and try that mod
and I sure would try it.
> 

> --
> Kevin Purcell
> kevinpurcell at pobox.com
> 
-- 
73, Jerry, K0CQ,
All content copyright Dr. Gerald N. Johnson, electrical engineer