
John's homebrew pages
Simple band-pass filter for 23cm
The improved 23cm transverter
had been tested in action but I wanted to add one final improvement - a bandpass filter.
There were two reasons for this. First, it is a Good Thing in that
it significantly reduces the radiation of out-of-band signals if any
are produced by the transverter. I don't think I really had any
problems with out of band sugnals, there is already good filtering on
the transverter main baord, and the driver and power amplifiers are
both tuned, not broadband. However, it was also an interesting
'exercise for the student' since I hadn't built a proper metal
filter(tuned lines in air are less lossy than microstrip) for these
wavelengths before.
The second reason is that the transverter receive side is completely
open, i.e. broadband - there is no filtering in front of the two MMIC
devices at the front end. Again this had not given me any problems, but
if I operate close to stronger microwave sources it could well do, and
so filtering these out before they get into the transverter is also a
good idea.
One problem was that the box containing the transverter, though
quite large, did not really have room for full size quarter wave
resonators at 23cm. I found a design for a short tuned line filter for 144MHz in the RSGB VHF/UHF
Manual (3rd edition) and scaled the lines from this design to 1.3GHz,
which gave a line length of 19mm - much shorter than a quarter wave. I had some 3mm brass tube, so
calculated the inductance of 19mm of this, which resonates at 1296MHz with a 1.7pF capacitor.
I had found a few piston trimmers in some surplus equipment bought
at the Galashiels rally. These trimmers have low loss for microwave
use, and I picked out two from a two stage 1500MHz bandpass filter
built with microstrip lines. I also found a design of a microstrip
1.3GHz filter in the RSGB Radio Communication Handbook (10th edition).
I hoped - not having the
specification, having tried to find piston trimmers on the web marked
like the ones I had - that 1.7pF would be within range of the piston
trimmers.
Here's the filter as built, without its covering tinplate lid (more
Hula Hoops consumed!). The trimmers are screwed into the bracket I
extracted off the board they came from; this is soldered to the back
filter wall. The lines are soldered between the base groundplane and
the trimmers, with the input and output connections tapped just below
the line centres.
The design seems to work well; I tried a prototype with adjustable input taps, and
found that the input and output tap positions on the lines were not too
important. I don't have the test equipment to make accurate measurements of
the insertion loss of the filter, and in fact once the filter is tuned
no significant loss is detectable with my home-brew through-line power
meter.
So that's the final addition to the new 23cm transverter which can
now be given lots of exercise in the RSGB UKAC contests and from a few
SOTA summits.