Re: G3WDG Boards

Andy G4JNT
 

Yes, I remember that filter - liked the dual mode cavity concept.

But I also recall the 'JVL transverter that used iris coupled cavities in standard WG16.   The LO filter was designed for 20MHz bandwidth, and the signal one for 60MHz BW.  These seemed incredibly narrow, but Mike explained that due to the very high unloaded Q of WG you could get away with high loaded Qs and he'd used these values to make the transverter useable with a 28MHz IF.   My 'JVL transverter didn't need a retune over the 4 or 5 years I used it before upgrading to a WDG transverter so that really high Q tuning stayed stable.

Iris coupling was horribly tedious to build, for a non-metalworker like me, what with having to cut clots in the WG and prepare accurately drilled holes in copper shim then solder them all up.   But around then, 1990 ish when I worked in a real RF lab at Marconi Space (as was), someone in the filter group showed me post coupled designs, and some software he'd written to design them.   NOW,  I know he took the equations from MYJ and implemented then in Basic (the original interpreted GWBASIC !)

Post coupling would be MUCH easier to homebrew - it was pointed out to me - as all that needs is a pillar drill and bits of thick copper wire.  You can't get such low values of coipling for narrow bandwidths, and I seem to recall a version of the 60MHz 'JVL signal one was really pushing a post coupled design, with the 20MHz one not feasible.
But going on memory, a third order 0.5dB Chebyshev that was suitable for as 144 IF was a practical possibility 

I still have a listing of the Basic SW converted for PowerBasic 16, including a 16 bit compiled .EXE of it,  but I seem to have damaged the code in the interim as it generates error messages now.  May be worth studying MYJ, comparing with the Basic S/W and redoing from scratch



On Mon, 9 Dec 2019 at 15:24, Chris Bartram G4DGU <chris@...> wrote:

Hello Andy

What we as a 10GHz community could do with, is an off the shelf, or easy to build, 10GHz bandpass filter suitable for image rejection at 144 IF.

The problem trying to use a 144MHz IF at 10GHz is akin to trying to use top band as an IF for 2m! It is very difficult indeed to design a simple, low-cost filter to reject the image adequately. That is particularly true if you wish to make it free from adjustments. I have been there!

My solution, which was available off the shelf, was a simple two-pole filter using orthogonally polarised circular WG resonators sharing the same length of guide. The design produces 50dB+ rejection at 10008MHz with between 1 and 2dB loss at 10368MHz.  I came up with that I when I designed my transverter. The design did go through a number of iterations in which I tried to eliminate adjustments. I almost got there, but the very accurate machining which was required put the costs of manufacture through the roof. The final version, while extremely reproducible, did need to see my VNA for set-up.

I have looked at other solutions using eg. resonators realised in some forms of substrate integrated waveguide. Some look promising, but other activities have interfered, and I'd like to get back on the air before I decide to do any more work on the project!

73

Chris G4DGU

Join UKMicrowaves@groups.io to automatically receive all group messages.