RF/Microwave experts: Identify these mystery PCB structures

[G0HZU]

I had a rummage through some old backups and found a DOS version of Eagleware from 1993. It's version 4.1 and the whole design suite fitted on a couple of 3.5" floppies back then.

This is about the time that =M/FILTER= first appeared and although the screen shot below looks very dated this SW was absolutely amazing back in those days.

The user interface was improved a year or so later but I would have used this old DOS version most days when designing preselector filters for various microwave receivers back then. I haven't tried to use this old version for 20 years but I managed to put together a fairly similar BPF design in the image below. I didn't run the optimiser in the second image showing the simulated response so the passband response is a bit lopsided. The raw filter synthesis tool doesn't produce a perfect filter response even in the later versions of this SW and so it has a manual optimiser feature built in that can tweak/optimise the filter layout for you. I think the optimiser in this version is fairly crude and the later versions have a one touch optimiser button that is well set up in advance to optimise the filter to the basic filter response/ripple requirements.

Note that despite its age, this ancient DOS software uses a crude parallel port dongle for copy protection so I can't give out copies of it.

[T3sl4co1l]

Ah, cool!

So why can't I have that (and basically identical lumped element functionality) in my PCB design tools 

Tim

[KJDS]

Ah, cool!

So why can't I have that (and basically identical lumped element functionality) in my PCB design tools 

Tim

Because you have bought a cheap pcb tool.

Combine Mentor with ADS or Cadence sprectre RF, but these aren't cheap solutions.

Microwave office and PADS can also be combined

[nctnico]

A couple of years ago I have written a small program to convert a lumped element (L & C values) elliptic filter into a microstrip structure. I never got round to add other filter structures to it or include creating the lumped element values in it. There isn't much commercial demand for it.

[KJDS]

A couple of years ago I have written a small program to convert a lumped element (L & C values) elliptic filter into a microstrip structure. I never got round to add other filter structures to it or include creating the lumped element values in it. There isn't much commercial demand for it.

I did similar for a straight interdigital filter many years ago. It's quite a fun exercise, and with the ease of Sonnet or momentum then it reduces design times significantly.