This weeks Solder Social was a great learning experience!
Yeah, I’ve started calling our BiTX20 building group the Sewing Circle with Soldering Irons. A Soldering Circle! What’s the difference? A bunch of old bitties sitting ’round gossiping and trying to make ourselves sound important to one another. It’s a hoot!
We saw firsthand what a couple of picofarads can do to / for a filter. The BiTX20 transceiver front end has a nice bandpass filter that doubles as the front end filter for the receiver and the driver filter to the 5W PA. It’s a fairly important part of the circuit because if it’s not tuned well, performance will suffer on both receive and transmit.
First off, the documentation is a little contradictory to the markings on the board. In the photos of the board layout, the filter capacitors are marked 68pf. They are also 68pf in the supplied schematic. BUT, on the board itself, and in the written text, it calls for 33pf. In the text it states one could use either a 33 or 47 pf. We figured it was because of the tuning range of the inductors.
Some of us had already built the filter with 39 pf caps that appear to have been substituted for unavailable 33’s or 47’s. Maybe it was in the middle of the range?
Contemplating the filter
Dave KC4ZVW had his Filter / Amp assembly ready first so we applied power and checked voltages per the instructions – everything was within tenths of volts, fantastic!
Next we went about setting up the Tracking Generator and Spectrum Analyzer. Thanks to Tom KD4WOV for helping us get that set up. I’ve never done that on a filter before.
Wow, what a mess! It had huge insertion loss and a lousy pattern. But after fiddling with the caps a bit Dave found a nice compromise between pattern and insertion loss.
At last, a decent pattern with low insertion loss!
Mario KU5E had used 33pf capacitors and came up with completely different results. He ended up removing the 33’s and putting in the 39’s. What a difference! We spent quite some time fiddling with the pattern and insertion loss with the 33’s and were never able to get less than 15dB of pure wasted signal.
As usual, Vern Ki4SDY was an excellent host and generously provided some caffeinated beverages, although I’m not sure this group really needed any stimulants ;]
Best we could get with the 33pf caps!
Now this is a good pattern!
Once I finally figured out how to take a picture of the spectrum analyzer screen these shots came out pretty well. Next week I’m going to bring my desktop tripod so that I don’t have so many blurry shots. These screens sure tell the story. We were able to get these tuned to with a dB or so insertion loss.
Overall the filter looks like it has better than 20 dB overall rejection, with less than 3dB insertion loss in about a 100 kHz passband. We don’t think we have this filter optimized – I will likely try several values there. Perhaps a variable?
a higher resolution shot of the pattern
Everyone is invited to our weekly Solder Social at Verns barber shop in Longwood, near the intersection of Ronald Regan Blvd and Church Street. Vern is the second door east on Church.
Hope to see you all there next week!
