Since I am NOT ever intelligent enough to leave things well enough alone, I was compelled to solder on two edge mount SMA's on my circuit board that a friend of mine bought from K5GNA for use with a Transystem AIDC 3033 (in my case-- alternatively a AIDC 3733 or even a Drake...) and measure the filter/PCB with a HP network analyzer again... You think that I'd learn....
Anyway, here is the "public service info" that will shows that the Murata filter is indeed adequate for use as the SSB filter to get the Transystem converters to have a decent image response.
Plot of S12 (or S21)-- the filter isn't 100%
symetrical-- but it is close enough for all practical purposes with 50
Ohm in and out.
Geeze, I guess I should post the return loss graph as well.. Well I didn't take the time to make this one pretty, but here it goes:
Another thing I thought of is to ignore K5GNA's advice on not putting two Murata's in the Transystems, and taking off the dipole, and using the 50 Ohm adapter. On a 3033 this should give a 2.4 db NF/27 db gain approximately. On a 3733 this should give a 2.4 db NF/35 db gain, more or less. This proceeded by a DEM preamp with a 0.6 dB NF would put the NF of the Transystem in the league of the SSB Electronic units. Would the extra 16 dB of gain kill us with oscillations? (41-50 dB of Gain?) Maybe.. but if that happens attenuation could be added at the second Murata filter, or likely at the 2M IF. I don't know if I'm going to do this to mine. Probably not as I'd like to use the BBQ dish and the dipole with mine. But its a thought. When my life gets together again, I'll run these theoretical numbers through the AO-40 spreadsheet.
Good luck on your AO-40 adventures. Hopefully I'll get my station together enough to join you in the next year or so..
73 de Fred W0FMS