In days past, long distance phone circuits were bundled together
with frequency division multiplexing (FDM).
Frequency selective voltmeters were used to monitor and troubleshoot
these systems.
Now that analog FDM systems have joined NTSC TV in the history books,
these instruments have appeared at hamfests, surplus dealers and Ebay.
The HP 3586 frequency selective voltmeters appeared in 1980 for about $12,000. This microprocessor driven beast tips the scales at fifty pounds. A 10 MHz input allows the 3586 to be locked to an external frequency standard such as a Trimble Thunderbolt GPS Disiciplined Oscillator (GPSDO).
The 3586 includes a high quality tracking generator which can be used for driving radios or transmitters. I plan to use this to stabilize the 30 MHz reference in my ICOM 706MKiig enough to make 144 MHz WSPR possible.
The 3586B has WECO input connectors. The Bell System must have bought a tone of these. Click on the picture to see an ad hoc solution - a paper clip for the signal. Another jack provides a signal ground connection. An adapter has been ordered. Some have replaced the WECO connector with a BNC connector.
The 3856's GPIB interface allows me to write Linux programs to
scan an arbitrary number of frequencies and record the signal strength
and carrier frequency of each.
As I wrote this, 1190 KEX is on a carrier frequency of 1189970.2,
somewhat out of the FCC regulated 20 Hz tolerance.
The rest of the locals I've checked are within 10 Hz.
(KEX has returned to its assigned frequency within 0.1 Hz.)
Left: plot of signal strength from two nearby NDB navaids
showing daily signal strength variation.
Left: plot of carrier frequency from two nearby NDB navaids
showing daily frequency variation.
Two turns of very small guage wire around L601 allow the 3856B's tracking generator output
to injection lock the main oscillator on my ICOM IC706MKIIG transceiver.
Click HERE for more pictures
Last Saturday was a wild night.
I decided to test the L7 on 160 meters.
Since my Nye Viking Antenna Impedance Matching Network
does not cover 160, I put my old Dentron SuperTuner on the line.
The Dentron easily matched the Cobra Senior antenna.
One look at the SuperTuner shows the "Super" refers to the frequency coverage or matching range rather than power handling. Not unexpectedly, snapping sounds could be heard from the little Dentron as I spoke into the mike. Some of the sparks were between plates on one of the capacitors. Other sparks emanated from the switch. Time for something bigger.
I got out my Nye Viking "Antenna Impedance Matching Network" PI network antenna tuner. This is apparently an early version with "homebrew" caps that could be switched in the output when needed to make a match. The caps weren't up to the task in the 80s when I was using a random wire for 75 meters. This time I decided to dismantle the output caps to see if they could be repaired. I found a poor solder connection on one of the caps. This might have been part of the original problem. When I reassembled the output caps I put the plastic part that showed thermal stress in the top cap, which is the last one to be switched in.
Lo and behold, the tuner worked. Not only that, but it actually seemed to match the antenna on 160. That was the good news. The bad news was that the strap to configure the tuner for balanced output had loosened. The tuner was actually matching something like a Marconi antenna with an extra loading coil (the extra ladder line in a loose loop).
With some 500 watts coming out of the L7 there was plenty of signal to reach OZ on 160 meter WSPR. Unfortunately, there was also plenty of RF going into the shack and the rest of the house. The SIIG keyboard freaked out. Linux started to forget its name. A broken ground on the CI-V cable allowed RF to short out a surface mount transistor in my ICOM 756pro.
Fortunately, the fried transistor was relatively accessible and I was able to replace it with a regular sized transistor. As shown in the picture, the repair wasn't pretty but it works.
Configured for balanced output, the Nye Viking tuner only tuned down to about 3 MHz. I don't think it was ever represented as covering 160.
After some experimentation, I found that reconfiguring the balun and adding a junkbox coil to the output allowed me to match the Cobra Sr. on 160. The balun is now a 1:1 matching transformer at the input of the matching network. The next high power test ended when one of the "homebrew" Nye Viking capacitors started hissing. Just as when the tuner was new, those caps just couldn't handle the power. I replaced the lot with a pair of 3000 volt transmitting caps in series. A 110 second run at 600 watts did not cause any of the parts to become warm.
This new configuration results in a nearly balanced output. The case is now at elevated RF so the result is not a perfect balance.
Disconnecting the serial and parallel cables from the Linux server
was apparently enough to keep Mother happy during a 600 watt test on 160.
Click for larger picture
For this contest I recorded a spectrum analyzer video
of the contest activity as received at Caddyshack..
An old Kenwood TR-7400A was pressed into service
instead of the usual ICOM 706mkIIg.
I chose the old Kenwood because the innards were more accessible.
I changed the tone pad input to a line level audio output.
I drilled a hole in the back heatsink for a jack for a 455 KHz pickoff.
The towel is Princess' supervisory location.
(She can be seen in a picture below.)
The Advantest U3641 spectrum analyzer video screen output and audio
from the Kenwood were captured with Hauppage WinTV.
The 5+ GB file was compressed to 212 MB MPEG-4 file with Linux FFMPEG.
The recording was started at exactly 7:55.
The time displayed by the spectrum analyzer is somewhat bogus.
The recorded audio goes dead when I transmit.
I hope to have a fix for this oversight the next time.
Click HERE for compressed video (212 MB file)
If the video doesn't play, right click and then choose "save link as ..." option.
2-29-2012 Fifth Wednesday Contest Spectra
These measurements were made with my prime 2 meter antenna connected to a Ramsey 2 meter one transistor preamp feeding an Advantest U3641 spectrum analyzer.
I wrote software to control the analyzer, download and format the data.
Tek 2712 Spectrum Analyzer
RF Pickoff for a RACAL-DANA 1992 nanosecond UNIVERSAL COUNTER
This counter needs a clean RF signal to get an accurate count. A bit of pink wire inside the shielded SWR sensor gets enough RF to run the counter.
Interfacing the Racal 1992 Counter with GPIB
E. F. Johnson Viking Ranger 160-10 meter CW/AM Transmitter
LEDs and Smarts for the CDE/Hy-Gain Rotator Controller
MODS for the MFJ-931 Artificial Ground
Rubidium Frequency Standard and GPS Disciplined Oscillator
New Life for my BUD (Big Ugly Dish)
Slower Speed CW (QRS)
My first memories of radio involve a ham who lived across the street. He had his station in the attic of his parents' house. He could send code across the country. He could light a light bulb by placing it near the transmitter's tank circuit. That was incredible. I remember looking through a catalog and picking out a microphone for his transmitter. I don't know if he ever got that microphone, or if it would have worked with his station. I was about four then.
In early grade school I built a crystal radio from a kit and tuned in the local stations. I also had one of those rocket ship crystal radios.
By 6th or 7th grade I was converting old radios into transmitters
by connecting the oscillator section of the tuning condenser to an antenna and applying audio
to the cap of the 6A8 converter tube.
An old radio with shortwave bands allowed me to listen to Radio Australia on 9580 in the mornings.
I took over my folks' big Zenith console radio before I got into Hi-Fi.
This radio had a 78 rpm changer,
two 12 inch woofers and a treble speaker, biamped.
I played LPs with a Columbia turntable with a pull-out tonearm.
Back then Hi-Fi meant being able to play an entire side without the needle getting stuck.
By high school I had a Miracord changer, Eico 20 watt amp (built from the kit) and an Acoustic Research AR2 speaker. The Eico had tremendous bass response - I once connected a speaker to the phono input and another small speaker to the output. When I wiggled the input speaker, the output speaker would wiggle in response until its little magnet was demagnetized.
My high school broadcasting adventures are chronicled under Radio Free Moscow.
I made 6 meter transceivers out of some old tube type car radios and a few extra parts. The receiver used a super-regenerative detector which became a modulated oscillator for transmit. The pair of radios could communicate over a half mile using standard car antennas.
In 1963 I passed the code and got novice and technician licenses. I dumped my old W9SEX and W9QRM calls for WA9JGQ.
I converted a surplus VHF business band transmitter down to two meters, adding paper clips to the final tuned transmission line to obtain resonance in the two meter ham band. A turret type TV tuner was modified to convert two meters to HF. I shortened the elements on a channel 2 Yagi to make a wide spaced two meter beam. A Heathkit 55 watt hi-fi amp provided modulated B+ for the transmitter. A KnightKit R-100 receiver and Ameco Nuvistor 2m converter kit listened on two meters. There wasn't much 2 meter activity in the Wisconsin Northwoods back then, but I did manage to eke out a contact with a ham in Appleton, 110 miles away. The signals were very weak, and we could not identify each other. Some years later, both of use were studying electrical engineering at the University of Wisconsin, Madison. We found ourselves in the same circuits lab where we just happened to compare stories and determined that we had made that contact,
I studied a study manual for the FCC First Class Radiotelephone license my sophomore year in college. Over Thanksgiving vacation my roommates and I took all the FCC exams we could at the New York FCC office. This was the last chance to take these exams for free. Shortly afterwards my roomies received failure notices from the FCC, but I received nothing. Weeks later I received my First Class Radiotelephone license.
I got a job monitoring the transmitters at WISM and WISM-FM weekends using my FCC First Class Radiotelephone license.. Back then a First Phone had to be on duty to check the directional array antenna currents.
Over the summer I worked for WISC-TV substituting for engineers as they took their summer vacations. WISC-TV was the only VHF station in Madison. I built a TV camera from a kit and converted a military surplus TV transmitter to the 420 MHz ham band. I designed my own video modulator for this transmitter as the original modulator had a poor bandwidth. I also experimented with flying spot scanning using a small photovoltaic cell as detector.
The summer after graduation I worked at the Gisholt Machine company. I designed a silicon transistor servo amp that worked with a cheap motor to replace a $5000 unit with germanium transistors that required its own air conditioner. The heat test chamber was not hot enough to affect the new servo amp. The amp and motor were installed on a turret lathe. It worked perfectly for about a month, regulating the feed rate so precisely that the shavings were extraordinarily uniform. One afternoon the lathe was working on a 400 pound cylinder, taking an inch off the diameter and a 4 inch drill down the middle. The feed rate took off and the work piece came unchucked at 60 rpm, busting $200 worth of tooling and rolling across the floor. I had been asked to design a servo amp with one millivolt sensitivity but there was no mention it would be used in series with another servo system. The phase shift caused by connecting the two systems in series reduced the margin of stability to where the vibration caused by the aggressive metal cutting disrupted its operation. That taught me a lesson - don't just design to the specs you boss gives you. You must understand the entire system.
I married during graduate school and moved into an apartment. I built a cavity resonator out of a coffee can to pull out Chicago FM stations WFMT (98.7)and WEFM (99.5) in between strong local stations.
After moving to another apartment I built a 50 foot long Yagi antenna to receive Chicago's WFMT classical music station. It gave much better reception than a Radio Shack FM antenna, good enough to listen in FM multiplex stereo.
When I first started work at Tektronix I built a RTTY demodulator using a double peaked active filter and obsolete RTL logic parts. I developed an 1800/110 bps modem for Tek but it was not productized. I designed the vector generator and serial interface for the Tek T4002 graphics terminal. I was project engineer for the 4010, 4012, 4014, and 4081 terminals.
When I moved to a large houseboat in 1975 I dabbled in two meter FM and SSB.
I left Tektronix for a small company called Sidereal where I designed their early Micronet terminals. The first three models used a Motorola 6800 CPU. I designed boards for Telex, TWX/DDD, 1200 baud, BRPE paper tape, remote batch, CRT display, and floppy disk interface.
The river was not a good VHF location. I practiced code and passed the tests for an Advanced ham license. I got a Drake C line, Hustler vertical, Dentron amplifier, and some long wire tuners.
I gravitated to RTTY and made my own Terminal Unit from the Sidereal Micronet computers I had designed for work. I wrote my own terminal program called CTU (Chuck's Terminal Unit). The CTU program synced its idea of the time to the 60 KHz WWB signal using a background software phase locked loop. Much of the time CTU would operate on 3627.5 KHz autostart. Hams could address my station and get a RTTY distortion report. They could also forward a message to local two meter RTTY. That activity languished when I left Sidereal as the tools to support the CTU program were no longer available.
Memorable contacts include RTTY to Antarctica and QRP to Alaska with a HW-7. I bought a Dentron MLA2500 linear amp primarily because the FCC was about to mandate removing the ten meter band from amps. I didn't use the amp much because the high power RF was more than my antennas could comfortably handle. The amp was useful on 10 meters to boost the Drake's anemic output to something reasonable.
I had an Eico 753 HF transceiver in my car for a brief time before it was stolen. I worked east coast on 75 and Japan on 20 mobile with the Eico.
I dabbled in VLF monitoring to detect solar flares using a strip chart recorder.
In 1995 I moved from the houseboat to Caddyshack. I miss the houseboat's metal roof which made a good counterpoise for verticals. I now monitor a number of VLF stations using Linux. I haven't been on the bands as much as before but my 30 year old Astron power supply and Butternut HF6V soldier on. My main rig is now an ICOM 756 Pro which recently blew its driver transistors. I also have a ladder fed multiband dipole and a Thunderbird TH5MK2 up 56 feet.
In April 2008 I fired up the 30 year old Micronet hardware, got a copy of Sidereal's old PDP-11 Unix running in my Linux server, and updated CTU for a Sidereal Reunion.
New (to me) Tristao CZ-454-FS self-supporting crank-up Tower.
Base section is 18 inches edge to edge, 21 feet long.
The tower came with pivot base and raising fixture.
The winches were replaced before the tower was raised.
Shack in Houseboat 1980s
Left to right: 4012 storage tube terminal, ASR33 Teletype printer, custom 6800 computer,
Dovetron HF RTTY modem, Heathkit panadaptor, Drake (4) C-line with general coverage synth,
Dentron MLA2500 and Nye Viking Matchbox
Basement of Caddyshack
showing 4016 storage tube terminal, Hameg digital scope, OMEN.COM server, test PC, and 4012.
Caddyshack operating position
with Supervisor.
756 Pro, 706mkiig, MFJ-993B tuner, Bendix MicroMatch, CDE rotor with Idiom controller, MFJ paddle, Yaesu HTs, office computer, Thunderbolt GPSDO.
Homebrew audio interface can be seen under the 756 Pro.
Astron RM-35M,
RF Concepts rfc 2-317 2 meter amp/preamp.
Not seen: ICOM R-9000 with replacement green CRT.
Below: Classic Astatic D-104 microphone bought at the Rickreall hamfest for $20.
The original element was dead so I replaced it with a $2.99 Radio Shack electret element (270-0090).
My junkbox contributed a few parts to interface the electret to the 756 Pro.
I didn't have a spare four conductor mike cable,
so I used an old S-video cable.
Some pictures of hy ham tower
ICBM N 45 35.689 W 122 52.420 Z 319'
160 Meter RFI from Panasonic Plasma TV as displayed by WSPR.
The receiving antenna was about 130 feet from the TV.
Signal to noise ratio improved 5 to 6 db
when the Panasonic was turned off at 0558Z.
This was taken on a summer evening when the natural noise level is high from
numerous thunderstorms.
The plasma TV would make more of a difference in winter
when atmospheric noise
is much lower.
