Personal Home Page | Business Home Page | Resume
Main Projects Page | Computer Projects Page | Telesaver Projects Page | Project Chronology Page

or... wandering around the house, finding stuff that any normal person would have thrown out decades ago. Take a walk down memory lane with me! Many of these photos were taken in 2010 -- so the subjects don't look quite as nice as when they were built (:>).

My eighth grade science teacher Joe Oursler W3VLZ (now a Silent Key, RIP) administered my Novice exam when I was in ninth grade, in the spring of 1969. So I became WN3MSW for a couple of months or so. By July or August I had gone down to the FCC office in the Custom House and passed the General and Advanced class tests. When that ticket finally arrived in the mail (August 14, 1969) I was transformed into WA3MSW. At the time, there was an Extra class requirement to have been licensed for two years; so I had to wait until 1971 to take that exam (along with the 20 wpm cw proficiency test). I took my last sojourn to the FCC office which had moved to the Federal Building by then for a very long day. Passed the Extra, no sweat. Then progressed through the Third Class, Second Class, and finally the coveted First Class Radiotelephone Operator's licenses. I asked Mrs. ___ (someday I'll retrieve the proper name of the long-tenured and well-known "FCC Lady" in the Baltimore office) what other exams I could take for the same money as long as I was there. She gave me the Radar Endorsement test. Hadn't studied for it, but what did I have to lose? I almost passed! (Can't win everything.)

In late 1976 the FCC began offering hams the opportunity to change their callsigns to preferred formats (predominately 1x2). I believe I was in the second batch offered (Extras by seniority); I was granted my second choice on April 19, 1977 -- and became N3IC forevermore. The FCC had opened up the prefixes (there had only been "W" and "K" ham callsigns) and offered "N" for the first time. I must have been the first "N" call to show up locally, because when I got on the 67 repeater using my newly printed N3IC call, people thought I was a bootlegger. Or maybe I was using a military callsign. Or an airplane moniker. It's nothing out of the ordinary now.

Amateur Radio serves up a marvelous plate with the opportunity to construct all sorts of electronic apparatus. Here are some of mine.

I didn't start out small -- this is the first transmitter I built after getting my Novice license in May(?) 1969. This transmitter was in one of the magazines, covered all five bands, and sported dual 6146B's in the output stage. I think it ran a quarter kilowatt input power. I was finished except I didn't yet have the 6146 tubes, so I wired a capacitor between the driver stage and the output tank circuit. And made my first ham radio contact, unwittingly QRP -- on 40 meters with WN1JYO, in West Haven, Connecticut (July 29, 1969). I might have already passed my Advanced, but hadn't received it yet and only had Novice band crystals anyway. This transmitter is currently in a sad state of affairs. You can see the fine strong construction, built with flimsy printing press sheet metal! Do you like the keylock switch I used for main power? (I took my newly acquired FCC license seriously, haha.) The power transformers (the most difficult parts to procure) and much of the other components came from discarded TV sets I found in the neighborhood. I specifically remember that all the books I had read (I didn't know any hams and had never seen a transmitter in the flesh) said that when the tuning cap was turned to resonance, the plate current would dip. I had a hard time believing that changing a capacitance (an AC action) would actually cause the DC plate current to dip! I was really fearful that after all my efforts, I would tune that cap and the meter would show no change -- and I didn't know what I'd do then. But, lo and behold, the plate current did dip! Class C amplifiers did work. I was very, very, relieved!

My next transmitter was a used Heathkit DX-60B (for receivers, I had an HR-10B and then a more sensitive HR-20B from Lee Nackman WA3IRQ) which has a single 6146B in the output. I added a second one in the final cage and added another power transformer; that doubled both the voltage and current, keeping plate impedance the same -- so I didn't have to modify the output tuning circuitry. It was probably the hottest (thermally) DX-60B around! It had AM modulation with a carbon mike I configured -- rarely used that mode, but one time magically I called in a pileup and got South Africa on AM. Eventually I panned the transmitter off to Steve Kennick WA3SOR. I also picked up Lee's homebrew VFO which stopped drifting a few days after power-on, though certainly better built than I could muster. [A few years later Lee got me a summer job at Electronic Center (where he was the guru); I always managed to repair the equipment (TV's, stereos, 8-track players...), but after re-assembly my jobs seemed to have parts left over. Mostly nuts and bolts and other extraneous hardware. The boss didn't appreciate that, but I figured the manufacturer just had put more hardware into the gear than it needed.]

Somewhere along the line (1971?) I modified the WWII ARC-5 transmitter (above left) according to a magazine article to convert it to SSB. It used the phased audio/rf method, and used an internal mixer to provide outputs on 80 and 20 meters. There's a separate power supply chassis. It worked somewhat, but don't recall the audio quality being all that usable.

No, I didn't build the Galaxy V transceiver (I bought it from W3HOP) above right. But I did build the receiver preamp in the copper-clad box on top of it (the Galaxy's receiver was a bit weak on the higher bands). And the combination VOX/QSK box on top. After my previous dual 6146B experiences, I actually liked and preferred the sweep tubes the Galaxy used (most people had the opposite sentiment).

I first used a bug, then moved up to keyers. The contact paper clad keyer at left (early 1970's) was all CMOS and had an internal battery; I built-in a paddle. When the AccuMemory Keyer was published in August 1975 QST, I built that (but later purchased a commercial paddle for it). It was great, and is still in use. You program four messages into RAM for playback, especially during contests. Its main drawback is nonvolatility, i.e. if you turn off the power you lose the stored messages. But once I tried iambic keying, there was no turning back!

After all my previous keyers, I designed my own with an 8748 microcontroller (October 1979):

I called it the CW King; it was both a keyer and a station controller. (Actually it used an 8035/2716 combination instead of an 8748, which was much cheaper at the time.) I constructed a homemade paddle for iambic operation (the copper clad paddles needed periodic burnishing). As a keyer, there were four preprogrammed (uP EPROM) messages and four programmable messages (CPU RAM). As a controller, it could be connected to a rig; it would listen to the audio, decode it into cw, and reply to any calls. For example, if it heard "N3IC de WA3EPT", it would respond appropriately (WA3EPT de N3IC K). It would also store received messages, and accept commands to turn on/off an AC outlet on the back of the box. Connected to my HW-8 QRP transceiver at the home QTH, I called it a couple of times from the JHU radio club station; it worked, and I left it at that. (FCC regulations, etc.)

I operated mostly cw, got WAS (worked all states), some contesting, blah-blah-blah. See the RARC homepage for Field Day. Did cw NTS traffic-handling, where I met Bob Collidge WA3MJF (now N2AW) who joined our FD team. One (really two) notable little "expedition" was:

Bill was not that easily dissuaded, however, and he drove us to the top of Sugarloaf Mountain (you have to trek up a ways by foot from the last road access to the top):

I made some contacts, but don't think I succeeded in making at least one on each band.

I believe it was the following year (1975) that Bill had moved to Atlanta and I made the Chesapeake Bay Bridge Walk by myself:

Gordon Schunick WA3OIA had bought a Drake TR-22C 2m FM transceiver, and was goading me into getting on two meters and the local repeaters. My first entry into the field of VHF communications was the Pip-Squawk 2-meter receiver in the July 1971 issue of QST. It converted the band down to the AM broadcast radio frequencies, so it used a garden variety AM receiver as its IF. The vernier dial connects to the receiver's tuning cap, where I've marked the corresponding repeater frequencies, etc. The project also added an FM discriminator to the AM receiver in order to demodulate the 2m FM signals. It worked okay, so I added the Pip-Squeak transmitter from the March 1971 QST issue. That has been removed from the box somewhere along the line. I made my first VHF contact with the combination on December 19, 1973 (with WA3OIA). Do you like my naugahyde covering the blemished chassis? The radio worked, and for a little while I used it at home and in the car (mounting hardware on the sides and a carrying handle for those purposes). But I wanted something better, so I replaced it with more powerful equipment.

I had two meter and 440 MHz gear in my 1972 Chevy Nova: there's a 5/8 2m antenna on the trunk and a 440 antenna on the roof. Rigs trunk-mounted. Probably 1974 pics. When I got the 1981 Toyota Tercel I put 2m, 220 MHz, and 440 MHz antennas in a line on the roof. Inside there are Kenwood TR-8300 (440 MHz) and Clegg FM-21 (220 MHz) crystal controlled transceivers, and an ICOM IC-255A 2m synthesized unit. Looks like Hooked on Classics and Olivia Newton-John cassettes. Probably 1984 pics; didn't have my ham tags on the car as at the time it was registered in the company (Telesaver) name.

In the early 1970's all VHF/UHF gear was crystal controlled. I had converted tube-type strip receivers and transmitters at the home QTH (above left), and an RCA Super Carphone on 2m and a GE Prog line on 440 MHz in my car (the Nova). GLB came out with a frequency synthesizer that you plugged into a crystal spot in the transceiver. Someone obtained the artwork for the circuit boards, and "cloned" GLB's were very popular in the Baltimore area (notwithstanding the ethics). Mine is at the right. I moved it between the home station and car. I spent some time helping others to get their synthesizers to lock. Can't recall what most of the difficulties were, maybe in the PLL filter section. In 1974 I added a little circuitry to turn the GLB into a two-channel scanner (between the two thumbwheel switch frequencies). June 1978 CQ Magazine published my modifications. Working with the GLB was an eye-opener for me; it was the source of many test questions I made up for a digital electronics course I was teaching at the time. It also was the germ of my idea to create a synthesized dot clock for variable video resolution on a CRT terminal.

In the 1975/1976 timeframe I was active on the Oscar 6 and Oscar 7 satellites. For Mode A (2 meter uplink/10 meter downlink) I transmitted with my FM gear modified for cw operation and received on the Galaxy. For Mode B (432 MHz uplink/2 meter downlink) I transmitted with the UHF gear modified for cw operation; I used a down converter to translate 2 meter signals to HF frequencies (I think 10 meters) so I could listen on the Galaxy. For both modes I used omnidirectional antennas which worked well enough. We had a more sophisticated satellite setup at the JHU radio club.

I became enamored with Motorola HT-200's and HT-220's. I acquired many, many broken ones (the ones that worked were very expensive). Bought the service manuals (made many trips to the Motorola center on Sulphur Spring Rd. to purchase replacement components) and managed to repair most of them and move them from the commercial frequencies into the ham bands. I had these walkie talkies operational on 10 meters, 6 meters, 2 meters, 220 MHz, and 440 MHz. Built a rapid charger (top left of commercial multi-charger) and a simpler trickle charger (to its right) for the nicad batteries. I should have sold the HT-220's while they still had value (hundreds each -- Motorola's new price was about $1100, I believe).

I did build a lot of VHF/UHF gear. While I was virtually always successful with analog and digital circuitry, and mostly with HF RF designs, my success ratio in the VHF/UHF arena was less so. Modifications of commercial equipment to move the frequency range to the ham bands was mostly always successful -- probably my HT-220 on the 220 MHz band was the crowning achievement in that area. But my trials with complete from scratch VHF/UHF construction cause me to admire those who developed a better track record!

Local repeaters had "autopatches" to make telephone calls. It was necessary to have DTMF pads to make calls and to access repeater control systems as well. I think I used the black one at right at home and the silver boxed one in the car. Little did I know that in a few short years at Telesaver we would be purchasing functionally identical devices commercially made to enable customers to dial telephone long distance access codes.

I can't find any photos of the unit itself. At the JHU Amateur Radio Club (WA3EPT, also see this) we had our own building (it really was a shack). The shack was a WWII munitions building adjacent to the Newsletter's building -- it was demolished to provide area for the BMA's sculpture garden. There was an 80 foot tower, and along with the HF beam we had an azimuth/elevation rotor system (hard to see above the HF beam) for the VHF/UHF Oscar satellites that we used with transverters for satellite operation. In May 1975 I found yet another application for the venerable 2102 1024x1 bit memory chip: satellite tracker. This was a "dumb" box which worked by dead reckoning. The controller would pulse the az/el rotors at programmed times with a fixed pulse width. Fellow undergrad student Alec Steingart (can't remember his call back then, now he's N2RX) wrote a FORTRAN (I think) program which produced a voluminous printout for all possible Oscar orbits. The user consulted those pages and selected the crossing of interest; it provided the data (1's and 0's) to preprogram the Orbit Memory with, and a time to push the Start button. It was about five minutes worth of data entry. The device then tracked the satellite for the duration of the pass (typically 15-20 minutes). It worked great. Here's the documentation.

I was selected to provide communications for the 1978 Preakness Balloon race. Here is an article I wrote describing it in the July/August issue of the Modulator. Here are the newspaper articles providing an account. Sad to report that my communications companion for the first attempt, Vern Chapin K3VC (RIP), became a Silent Key in July 2009.

Many projects were related to the Baltimore Amateur Radio Club (BARC) repeaters. The main installation was at the old WBAL TV tower on Violet Ave., where we had the 146.07/146.67 and 444.35/449.35 repeaters. Above is a histogram I made of 67's usage; it is for a one week period (Thursday to Wednesday) beginning September 14, 1978. We had counters on the repeaters: over a one year period around that time, 07/67 showed an average transmit time of 12.05 hours/day (851 cycles/day); 449.35 had an average transmit time of 1.4 hours/day.

There were a number of people involved in installing and maintaining the repeaters. Carroll Van Ness K3HZU (RIP, right) did a lot of the early work, and he built the first autopatch and control function hardware. It used traditional (reliable) Western Electric subcircuits and relay logic (which he was quite familiar with as a longtime Bell System employee). He was very encouraging towards me when it became time to modernize those components.

Larry D'Anna WA3KOK (center) did a lot of the RF work. But in the 1970's/early 1980's timeframe, the person predominantly responsible for the overall repeater networks was Jack Biggs WA3SEY (left, now K3SP). I spent a lot of time with Jack -- and I learned a tremendous amount from him. Ham radio is indeed a "hobby" and it is in fact called "amateur" radio -- but repeater users did not behave that way. They wanted their 1 watt walkie-talkies to work anywhere it seems within a 30-mile radius or so. And they expected "their" repeaters to be up 100% of the time. No, the local hams really didn't want an "amateur" repeater.

I learned from Jack that even a hobby can be handled in a professional fashion. There was little guesswork with him -- he had learned from his job responsibilities running the State Police communications systems what would work. And he applied it to our repeaters. He also was very supportive of my meddling with experimental circuits on a system which otherwise was fairly similar (and reliable) to commercial repeaters of the era. The BARC repeaters offered a rich platform in which to develop and test electronics designs.

[One might "blame" Jack K3SP, Vern K3VC, and Bruce W3YVV for getting me hooked on motorcycles. Together we toured to many hamfests in Maryland, Pennsylvania, and Virginia on our Honda Goldwing, CB-750, and CB-550 bikes.]

The two meter repeater's transmitter was at the Violet Ave. location; initially we had one remote receiver and then added another. A "voter" circuit is needed to select the preferable audio source for the transmitter from among its inputs. Above left is the 2-channel voter circuit I designed for that purpose in February 1974. It worked by sampling the noise levels on the two link receivers 140 ms after the squelch opened, and picked the lower one for the duration of the received signal. Not sure how I came up with that scheme. It worked, but not nearly as well as the 6-channel commercial GE system (new cost thousands of $$$) we later installed. Here's the schematic and an explanatory letter.

All repeaters have cw identifiers that operate every few minutes. I was bored listening to the same announcement repeatedly. When the Intel 2102 1024x1 bit memory chip came out, I could find no end of tasks for it. Above right is one chore I assigned it in March 1975: a reprogrammable IDer. You had to write the 1's and 0's to create the dits and dahs for the identification. But at least you could change it at will. Here's my April 1976 73 Magazine article on it.

In the summer of 1975 I designed a 4004 controller for the two BARC repeaters at Violet Ave. (see below). This was a large project, and entailed system architecture, circuit design (back then you even had to build transistorized clock circuitry), printed circuit board layout and etching, control software, and test. This was a followup to my very successful JHU 4004 project the previous summer. On the one hand, it was a success: I wrote a test program which exercised all aspects of the hardware, and it worked perfectly. But on the other hand, when I tried to write the actual repeater application control program... I failed. I just could not get the 4004 to do everything I needed simultaneously. My hardware design was such that it put everything under software control for maximum flexibility. That meant that the uP was responsible for picking up the carrier-operated-squelch (COS) signal from two repeaters, routing the audio itself to the proper output, and activating the transmitter push-to-talk (PTT) line. Keep track of the on-time for both repeaters, generate cw identifications for each when necessary, and by the way, look for DTMF signal detection and parse the digits for the autopatch (telephone calls) and control functions. Theoretically possible. I just couldn't do it. Probably partly due to my software proficiency level, perhaps -- but largely due to the fact that the 4004 was simply not that powerful a processor (the first invented by man). Still, compared even to the slow CPU speeds of pMOS, the speed required to handle the real-time repeater control wasn't all that great. The problem was that the 4004 did not have an interrupt input. Yes, Intel cleverly labeled one of its pins Interrupt, but that didn't make it so! It really was just a general purpose input pin. No interrupts means no real-time clock interrupt which means it's really hard to schedule multiple tasks. I might know enough now to construct a loop which interrogates external clock hardware, but then again, that might not be sufficient because the individual tasks could not be interrupted. After all, this chip was designed to make a calculator. Of course it's a moot point, but this project to this day gets my ire up! Truthfully, I learned an awful lot from this failed project.

These double-sided circuit boards are certainly low density. I made them with drafting tape and stick-on pad footprints on clear acetate film, double-sized. Then did the photographic work in the darkroom to produce 1:1 Kodalith film. Prepared the copper-clad with photoresist, exposed, developed, and etched the boards. Drilled the holes. (Same process as the JHU 4004 project.) The Clock board above includes an address decoder. The CPU board includes four RAM chips (320 bits each, if I'm remembering right), plus some buffers. The ROM board (center shot) holds up to eight 1702A EPROMs for a maximum of 2Kx8 program memory. The Input Ports board supplies 16 input lines (there were two of these boards), and the Output Ports board provides 16 output lines (there were two of these boards as well).

The Input Buffers board (above left) takes analog audio inputs from the receivers; the two Switches boards contain analog multiplexers (controlled by output port bits) which route the signals from the Input Buffers board to the Output Drivers board -- which contains several summing op-amps and drives the audio inputs on the transmitters. The fourth board above is the Hi/Lo board; it contains two analog comparators for each of two receivers. These connect to the receiver discriminators, and permits the detection of the user being high or low in frequency. At the time, one add-on some repeaters were installing was hardware to send a tone or cw message to inform the user (remember, everybody used crystal controlled radios which needed to be frequency trimmed) if they were off-frequency. The Hi/Lo board permits the controller to perform this function, too. The rightmost board above is the Interface board, which contains relays, an audio transformer, and high power transistors to drive the transmitters. These last two boards were only single-sided, the traces are underneath.

There was a separate power supply, don't know where that went. The interconnections on the card cage were made via wire-wrapping the 44-pin edge socket connectors. Here is some of the design documentation.

In November 1977 I installed my newly designed 8080 controller for the repeaters. After my failure with the 4004 controller, I threw hardware into it to be certain it could do its task. Probably more than necessary. It used an external Western Electric DTMF decoder and audio amplifiers (silver protrusions). In the above pic it sits on my basement floor, ignominiously pulled from service after a decade or more of continuous operation.

Our 67 repeater was fairly well-known on the East Coast due to its wide area of usage for somewhat hilly terrain and for our multiple receive sites driving the voting system. The innovative, visible features which the 8080 controller brought increased its figurative footprint; but when I had a series of articles published in 73 Magazine describing it, the buzz surrounding it grew even more. This was a pretty long series of articles (38 pages in all) and you'd think that with its relative complexity, while perhaps hams would enjoy reading about it, nobody would actually bother to duplicate it. In fact I received an awful lot of correspondence regarding it (some totally off topic, haha), and I've lost track of how many club's repeaters did end up with copies. Duplication was accelerated because for a small fee I was customizing the access and control codes in the software for individual groups, and burning 2708 EPROM's for them. So all they had to do was build the hardware.

This was an interesting time period because the publisher, Wayne Green W1NSD, was very interested in microprocessors and was pushing them in 73. He was receiving a lot of blowback from old-fashioned hams objecting to "computer" stuff being placed into a "ham" magazine. Wayne was thrilled to receive my manuscript because it perfectly showed how this new-fangled "computer stuff" could actually improve the operation of "ham" equipment. So he broke the article down into three (which morphed into four) parts and put each into his new "I/O" section of the magazine. It was his way of saying "so there" to the disgruntled hams. Wayne had just created the first big hobby computer magazine, Byte -- which he subsequently lost to an ex-wife in divorce proceedings. I still have every Byte issue ever published (September 1975 through July 1998) except for a few that weren't returned by borrowers. Here are the 73 8080 repeater controller articles:

I redesigned the bulk of the 8080 repeater controller for the simpler 146.34/146.94 repeater -- this time with newer chips: the 8085/8755/8155 chip set. I brought its predecessor's 57 IC's down to a total of 8. Partly there was less to control, partly the newer Intel chips had higher integration, and partly I relegated more functions to software (I had overdone the hardware for the 8080 controller because of my 4004 controller all-software method disappointment). This system still required an external DTMF decoder. I installed it in November 1979. I wrote up an article for 73 Magazine and they paid for it. Unfortunately, they never got around to printing it. This was a shame, because with all of the interest the 8080 controller article had brought forth, this one was so much simpler to build that I'm sure it would have been of far greater interest to repeater groups. John Williams K8JW (my Intel benefactor) built a duplicate and installed it on his Anne Arundel Radio Club's 220 MHz repeater. I don't recall either of us encountering any difficulties at all.

My final repeater controller design incorporated an integrated DTMF receiver chip. It was very exciting to have the ability to decode those tones with a self-contained reasonably priced part for the first time. I used it in this controller and then in the Message Converter shortly thereafter. Robin Becker WA2NYE (just above, now N3IX) and I (with controller at Charles St. location when I had hair) built a 222.64/224.24 MHz split site repeater for the BARC locations. We used a band downconverter with a good two meter receiver, and the receiver boards from a couple of HT-200 walkie-talkies for the control and link receivers along with an HT-200 transmitter board for the link transmitter. It made for a very small and low power system (except for the transmitter). The receive site was at the Charles St. condo location and the transmit site was the TABCO building in Towson (Robin working on antenna at that spot above). I got a kick out of Robin's labeling of the Towson control panel (at right): instead of marking a toggle switch with the more traditional Enable/Disable or On/Off terms, he put Energize/Deflect on it. This controller was derived from the 8085 design built for the 34/94 repeater. Installed in November 1980. Worked fine. I have a vague recollection of building a duplicate controller for a temporary self-contained backup repeater as well.

After we ended up with 6 remote receivers feeding the voter it became difficult to troubleshoot reception problems. Just by listening to the repeater, you couldn't tell which receiver(s) was/were being used. I actually spent many hours sitting on the floor at Violet Ave., just watching the voter lights. (I still maintain that there is no substitute for actual observation.) In October 1978 I designed and installed a telemetry encoder on a second UHF transmitter, and made up a kit for about 15 of us to build telemetry decoders. At left is my prototype decoder and at right is my personal UHF telemetry receiver (HT-200). Below is my finished decoder from the kit. Robin Becker WA2NYE (now N3IX) made up circuit boards. The bottom yellow LEDs show which receivers are picking up a signal, and one of the top green LEDs illuminates to show which one the voter has selected. In addition to sending voter information, by entering commands from the 8080 control system, we could temporarily change the telemetry signal to indicate plate voltage and current for each of the two quarter kilowatt repeater power amplifiers. That's what the meter is for. It could also be commanded to measure ambient room temperature.

The encoder board at the repeater location is shown at the right.

It was entertaining to watch the voter in real-time -- it might switch several times in a second. This is similar to the cellphone tower handoff prevalent today, although that works at a slower rate. When troublesome receivers were identified with the telemetry system, we could easily use the 8080 controller to disable individual receivers until the problem was resolved. Here's the documentation. I never published this information, because it was unclear what the FCC's opinion on the legality of the telemetry transmitter would be under Amateur Radio Rules. Too bad, because it was extremely useful and could easily have been adapted to meter other digital and analog signals remotely.