Glenn and I had a conference call with Spencer about the design of the 4th generation linear cylindrical patch antennas.
We're really paranoid about getting the f_0 of the antennas right, since the CPAs are so narrow band. We thought of making up test patches, and then measuring the resonant frequency of the patch.
- UT80 Coax. Strip back 0.5 ".
- hit with hot air gun to protect end.
- use it as a sniffer.
- Test strip
- 1/2 wavelength long, width doesn't matter
- 5T spacing between everything
- Doesn't matter how you approach it
- Center is probably best
- Approach it you'll see a little dip
- Doesn't matter from what angle
- Don't get so close that it goes down in frequency
- Probably need to be at 0.1 dB/div
- "Grid dip oscillators (GDOs)"
- If you get too close, you change the resonance
- Couple with a few fractions of pF
- Sweep across desired freq and look at S11
Making the antennas more wideband
=5 wavelengths tapers are great.
- First level: First 100 -> 50 to taper. Optimum taper?
- Second level: 2 or 3 stage transformer
- Third level: 1 or 2 or 3 stage transformer
- "29 degree trick"
- Each T has an impedance that is geometric from 112 to 50.
- SolidWorks simulation SAT file.
Attached is a screencapture from a Goto meeting that we had, explaining some of the wider-bandwidth tricks.