Proof-of-concept
Hall of Fame
other hi-tech products that never quite made it to market ...
Sun., 10-05-2003 17:50 UTC
Software: Experimental Browser for NWS RADAR (four NWS RADAR sites in and around the DFW area currently)
Six Meter Duplexer - 6M Duplexer Page
The
WB5WPA Six Meter Duplexer (Low-Band VHF) in it's basic configuration consists
of six band-reject coaxial "stubs" fabricated from The stubs are interconnected with 1/4 (electrical) wavelengths of RG/58 single shield (double braided/shielded cable isn't required at these frequencies) coaxial cable.
Attenuation of over 80 dB can be achieved using 1 5/8" Heliax with IL's
of less than 1 dB on a six-stub version depending on design.
High-Performance, Low-Loss Two Meter Intermod Killer
Swiftly
en route into the dustbin of technical history is this elliptical filter
composed of 10 reactive (LC) elements. This filter "Kills Intermod"
by effectively passing only frequencies below 148 MHz and attenuating those
above 148 MHz.
Outstanding performance is achieved by utilizing low-ESR caps fabricated directly onto a Duriod substrate. Duroid is a low dielectric-loss type of board material used widely at microwave frequencies. Bascially, this yields a design that uses caps that have no leads which unavoidablely contribute series inductance as found in normal 'leaded' components ...
This filter was designed to attenuate the commercial frequencies above
the 2 Meter Amateur band that include pager services, the NOAA Wx radio
transmitters and police and other 2-way services. The points of greatest
attenuation were chosen in the design of this elliptical fiter after an
extensive examination of the actual spectrum in my area using an IFR 1500.
The IFR 1500 was used as a synthesized and calibrated spectrum analyzer
and allowed me to record both the amplitude and frequencies of the potential
'interference sources' occupying the bands above the Two Meter (US 2 M
band:
The Intermod Killer contains no moving parts (no relays) and is completely passive (no transistors - no diodes!) and therefore does not require power for operation in transmit (or receive) modes.
In-band IL (Insertion Loss) as shown on the sweep GIF is less than .75 dB at the upper band (148 MHz) limit and out-of-band rejection (where it counts based on an actual survey of where the annoying pagers were busy churning out data) of 30 to 50 dB.
The two notches seen in the GIF are 35 and 55 dB deep respectively.
This filter is completely passive and requires no power. Those who have
tried it are impressed. It makes even an
This filter was 1) designed using EESOF's ESYN, 2) optimized using EESOF's
Touchstone and 3) subsequently swept and tuned on
Call for pricing (just kidding).
Click here to see the big Frequency (Bode) plot of the Two Meter Intermod filter
'Imod
killer' along with sweeps (caution: slow loading GIFs!)
Trunk
Tracker/GMSK data demodulator
Another one that never made it into production (although several protos are still in daily service!).
Let me tell you, there is NOTHING quite like writing 'your own solid code' and being able to 'see' ALL the trunking data without having to have blind faith in someone else's 'product' effort or backyard 'hack' described loosely in newsgroups by self-proclaimed 'lords' of hacking. One of the more memorable times in my life is when I first 'locked the loop' and tracked my first trunk groups in about 1992 (this was YEARS before commercial product was offered to the consumer!).
This was before I had incorporated the "rate 1/2 convolutional error correction code" routines too (much like as seems to have been done in the RS PRO-91; it falses regularly) and 10-bit 'block' parity check (normally performed over the received and 'error corrected' data to verify it 'corrected' okay) - it worked (example of early code) but was subject to some falsing occasionally. This would take the form of going to a 'channel ' or responding to a TG 'assignment' that didn't exist (owing to one or more bits errors in the received trunking control channel data LIKE the PRO-91).
Serious time and effort went into developing this item and I can't begin to describe the hardware and software co-development process that took place involving the intertwined use of a real-time 24-channel HP logic analyzer, several multi-channel real-time O-scopes (triggered at times by the logic analyzer to view the waveform present during a particular stage of data procerssing or to indicate the occurance of a critical interrupt), an EPROM emulator (loaded with the object output from the assember running on the development/compiling PC) and pulse-delay/pulse-stretch generators as well numerous 'software hooks' that could be caught by the logic analyzer and thereby indicate program logic state and data processing/ECD progress.
The box, when added to a scanner such as the Radio Shack Pro-2006 or Uniden BC760XLT demodulates 'wireless' data (of the GMSK variety) and formats it internally for analysis or "talk group" comparing against an internal list of desired 'talk groups' or outputs the data for further processing and logging to a PC. As such this box can be used as trunking design tool, as a system analysis tool (count the occurance of each and every different command capable of coming across the system) or as a learning aid (easily demonstrating the operation of a 'trunked' radio system within the context of an easily understandable conventional programmable VHF/UHF FM 'scanner').
The intent was to make this box was able to be added to a number of different scanners. It has two RS-232 ports for outputting data to a PC or receiving commands. Set up and control locally on this box is via front panel 'joystick' keys and the menu appearing on the alpha-numeric LCD display.
Accompanying software was written for the PC to allow viewing in real-time ALL data appearing on Motorola Type I and Type II analog trunking systems.
Link Summary:
Trunk Tracker page, shows some early hardware and describes it's basic operation. Main program 'Loop' - shows how main 'routine' responds to interrupts and then circulates though a series of flag-checks to determine the next course of action such as perform a mode change from setup channel to voice channel or vice versa. Main program 'code' - early code (circa 1993) prior to major program restructuring and incorporation of EDAC (Error Detection And Correction) routines such as the Rate 1/2 Convolutional Error Correction code and the 10-bit block parity code or the LCD driver routines. To view this later code you'll have to sign a non-disclosure clause (or wait until I have a clear sight of my death bed and I simply decide to post it) ...
Complex Impedance Mesurement
Measure the complex impedance (Z) of a one-port RF network (such as an antenna or a speaker or whatever) using simple components and a dual-trace scope (suitable for low frequencies) or an HP 8405A Vector Voltmeter.
This technique requres a litle math - but is usable from DC through as high as your construction techniques (and scope!) will allow.