It is time for a few more of us to detach from our moorings here at Emerson and begin our drift apart. I would like to sincerely thank all my friends that I have worked with here. I truly believe the people here is what kept many of us from looking for new careers sooner. I have decided leave engineering and pursue a career in <drum roll>… DEATH DEFIANCE! I have already performed a few shows. In order to increase interest, I am including a few photos from these past events.
Thank you. This has been a: Chris Karle present a Joyce-Karle Production.
|
|||
Original publication date: December 1991 Heurikon Historical Highlights A series of articles for The Horizon Oscar Mayer & Company Going to visit a customer’s site to install a computer? You just haven’t lived until you’ve had to slug your way through an endless moving line of 400-pound hog carcasses that are hanging by their toes in various states of, ah, “disassembly,” while you simultaneously fight the urge to look around out of fear you might actually see something — the other times you did that (look around) your eyes found a room full of organic swine horrors, like dangling pig heads, piles of steamy-fresh stomach parts, and vats of other mysterious entrails. Oh, sorry — were you eating lunch? Let’s back up and start again.
Our second major project was with Oscar Mayer & Company. In the late 1970s, Oscar Mayer was still slaughtering hogs here in Madison. After the hogs were dispatched, the eviscerated carcasses moved past a workstation where two people (called “graders”) keyed in quality values for each carcass. An electronic scales provided the hog’s weight, which was recorded on paper tape along with the grade information. Twice each day, the paper tape was gathered up (sometimes off the floor) and fed into their IBM mainframe computer. Oscar Mayer used the hog data as feedback for their buyers and to compute the payments for the farmer-producers. The hog data was also passed through a selection matrix that would turn on certain indicator lights, thus telling the graders how to mark the carcass. The marks indicated how the meat should be most profitably cut — as in bacon, baloney, or ham. The Hot Carcass Data Acquisition System Mike Werlein, the project engineer at Oscar Mayer, originally just wanted to replace their old mechanical stepper switches with an electronic version. But, when we saw their existing system (a hideous mass of switches, relays, and noisy paper tape punches), it was apparent they needed much more than just a simple stepper replacement. So, we proposed an electronic replacement with a modern eight-inch floppy disk drive (to temporarily store the hog data) and a modem interface (so the data could be sent directly to the mainframe). The first electronic system used our MLP-8080 microcomputer board (as usual) and was installed at the Oscar Mayer plant in Davenport, Iowa. The box was about the size of a large microwave oven, mainly because it included a huge Shugart floppy disk controller board that was as big as the cabinet’s cover! Today, that functionality is on a single chip.
After the Davenport system was operating, they asked us to build three more, but they wanted a backup disk drive, the ability to manually input and modify the grading matrix, and a few other bells and whistles. This time, we proposed a full rack of equipment and a custom software package. The most unusual aspect of the new design was that the stainless steel data entry keyboards and displays had to withstand attack by greasy hog fat and direct hits during the daily steam cleaning.
We had planned to write the application code for the hog scales in BASIC, but first we needed to write a BASIC interpreter for our MLP-8080. That, however, turned out to be a horrendous task. We spent all of our development schedule working on the interpreter, never quite getting it right, and had no time left over to write the hog program itself. After a few months, we convened a meeting at Oscar Mayer, and Chris and I told their project team that we made a big mistake in our planning, wanted to start over on the software, and would they please wait for us and not cancel the order, thank-you, OK? The Oscar Mayer people had a brief private meeting, after which one of them asked: “So, how much more will this cost us?” Whew! That was a relief. We feared they would just kick us out and cancel the project. “Nothing more, same price,” we said, wondering if our original quotation might have been a bit too low. “This was our mistake and we’ll stand by our quote.” “OK, fellows, go do it. And, please get it right this time!” Good Recommendations A few months later, we were pursuing other customers who asked us for some references. Figuring one happy customer would beget another, I called Mike and said, “Mike, we’re wondering if you would be willing to be a business reference for us. May we give your name to our prospects?” We were still behind schedule on the Oscar Mayer project, but they were happy with the work we had done to that point (one of the units had been installed was working fine), and they would certainly want to help us stay in business. “Sure,” Mike said, “I’d be happy to talk with them. Let me get your file.” There was a pause as Mike rolled his chair over to his file cabinet. “Let’s see,” he continued, “where did I put your file? Oh yes, here it is — under ‘d’.” I asked Mike why in the world would he put his Heurikon file under ‘d’. “‘D’ — for disaster, of course!” Well, he must have given us a good reference because we got the other business — and eventually more from Mike, too. More of our fancy hog scale systems were installed at the Oscar Mayer plants in Perry, Iowa, and Beardstown and Momence, Illinois. (Madison never got one.) The Momence unit is still in operation, which might make it the oldest Heurikon system still in use. But, alas, Mike says they plan to retire the hog line there at the end of this year, just as they have already done at the others. NEXT MONTH: O-P-Cue-R-S-T, or how we learned to cue up. [SIDEBAR] Pranksters at Oscar Mayer One day, while installing our equipment at their Beardstown, Illinois, plant, I returned from lunch and found a few dozen hog eyeballs stuck to our keyboard enclosure. They were round, white, and (apparently) sticky — I think they were still warm. I did my best to pretend that nothing unusual was going on — as if I had to work my way around eyeballs like that all the time back at the office. Out of the corner of my eye, however, I could see the burly meat cutters (swinging knives with one hand and wearing steel-mesh gloves on the other) glancing over at me and enjoying a good snicker at the expense of the city dude. |
|||
This is one of my all-time favotire newspaper stories. It's all true. This appeared on the FRONT PAGE of The Capital Times -- September 29, 1983.
|
|||
From the Heurikon Horizon, February 1992
|
|||
Original publication date: November 1991 Heurikon Historical Highlights A series of articles for The Horizon Arrow Sign Company Inc. Our first product development contract came from the Arrow Sign Company Inc., a Chicago advertising firm. Arrow’s Bob Scadron heard about us from a Madison-area relative who knew we were one of the few companies venturing into the new world of micro-computers. Bob drove up from his office one day and asked us to quote on building the computer controls for outdoor advertising message boards, very much like the one American TV has in front of their store today. Bob agreed to provide the lamps, sockets, and sign panels, but we had to figure out how to wire and control over a thousand light bulbs, and make the display show words using special visual effects. We were a low-overhead three-person operation back then, so our quote (only $3,500 for the engineering and $5,782 for the hardware) was attractive. We signed the contract in early 1975, just a few days after moving from Chris’s basement to a “real” office at 700 West Badger Road (another basement).
Do-It-Yourself Purchasing For the control unit, Chris designed an elegant CRT/keyboard terminal, complete with dark walnut side panels. Chris usually handled our purchasing, but when it came time to get a CRT for our terminal, John Burdick simply drove down the road to American TV and bought the cheapest portable black-and-white television set he could find. We removed the case, unhooked the tuner, and mounted the guts in our own enclosure. (In hindsight, maybe we should have left the tuner connected — we would have had the industry’s first combination data terminal and TV set!) The microprocessor was an Intel 8080 running on our MLP-8080 board (described last month). We designed a special 16-line by 32-character video display controller that could also display a duplicate image of what the outdoor sign was showing in a dot matrix area at the top of the CRT screen. As usual in those days, we wire-wrapped our prototype designs (in this case, the display controller and memory expansion boards), so we could easily make changes. Today, we still use the same green wire for our ECOs, but back then most changes were just put onto our boards with the rest of the wires. The electronics at the sign itself consisted of timing and control circuits, latches for the lamp data, and a triac (an electronic switch) for each lamp. We designed a high-speed communication interface so one coaxial cable could transmit data to the sign in real-time, and we figured out how to transmit nine-bit data using an eight-bit serial chip. John came up with a shrewd way of isolating our electronics from the high voltage lamp circuits; we wound our own small pulse transformers.
I wrote the application program in assembly language, one line per CPU instruction. It had a real-time portion that was in charge of computing the lamp states for each new image and serially transmitting the data to the sign at the right instant. A background task allowed the operator to type in and edit the message text and specify transition effects (e.g., travel, wipe, scroll), transition speeds, lamp brightness, and so on. One unique application problem we solved was keeping the messages readable when they were traveling across the sign. As the lamps were turned off, the long-life filaments faded out slowly, so moving letters appeared to smear unless we temporarily reduced the brightness of the lamps. The sign at American TV works that way, too.
During testing, we would often turn on the sign in our shop to check the operation of the lamp drivers. However, if we tried to activate all of the lamps at the same time, the system would consume over 30,000 watts of power and trip the main circuit breaker. As we approached the power limit, the power lines in the wall would hum, and the room temperature would rapidly shoot up past 100 degrees. That was fine during the winter months since we didn’t have any heat in the office (the heat we did get came through the ceiling from the apartments above), but it wasn’t much fun in the summer. When “ROM” Means Remember-Only-Momentarily One of our most nasty technical problems was with the ROMs. Unknowingly, we had been shipped a batch of defective parts from National Semiconductor. The read-only-memory chips either wouldn’t program completely or they’d get a bad case of amnesia and lose their contents after only a few days — and National denied it was their fault. After weeks of frustration, National finally admitted that other customers had complained, too. They checked their records and found they had shipped us parts from the reject pile. We learned that big companies could make big mistakes.
The prototype sign was installed at Jerry’s Fruit Market, not far from Chicago’s O’Hare airport. When we tried the sign at the site, the lamps would randomly flicker or not come on at all — it looked awful. While Bob paced, we huddled. Soon, we figured out what was wrong. We had used the single-phase power circuits in our Badger Road office to test the sign. In Chicago, however, the power was three-phase and parts of the sign were connected to different power feeds. As a result, the brightness control logic had the wrong timing reference. To fix it, we had to rework our timing logic and rewire part of the sign.
Satisfaction and Success It was exciting to watch the sign operate, knowing that Heurikon was responsible for the entire system from keyboard to lamp. The sign was highly visible, and it performed an obvious and valuable service. One otherwise ordinary airplane trip was made memorable when we flew within easy sight of the glowing sign while approaching O’Hare.
Although its original design life was five years, the fruit market sign helped sell apricots and strawberries for almost 13 years. As I look back, this was our “premier” systems project. It came early in our history, and it embodied all of the interesting and good things we’ve ever done. We solved many electronic, mechanical, manufacturing, and business problems while designing and building the signs. I visited the fruit market sign occasionally during the ’80s; each time it was like seeing an old friend. NEXT MONTH: Heurikon goes hog wild. |
|||
 A photo courtesy of Darrel Meyer: "Just a couple of pictures I took flying around one day." |
|||
 A photo courtesy of Darrel Meyer: "Just a couple of pictures I took flying around one day."
|
|||
Original publication date: October 1991 Heurikon Historical Highlights A series of articles for The Horizon Our Beginnings With this issue, The Horizon starts a series of articles about Heurikon’s early years, in particular the middle to late 1970s. During that period, Heurikon went from a basement start-up to being a hopeful “other” in the Multibus I market. The company was born out of Chris Priebe’s association with two electrical engineering students at the University of Wisconsin-Madison. Chris was working on his master’s of fine arts degree and asked the engineers to help with his art projects that involved sound and light. In 1972, the three students decided to start a business. They had intended to build an alarm system to prevent in-store thefts of turntable cartridges and styluses (“The Cartridge Cop”).
Heurikon’s basement beginnings, 1972-1974 at 621 Sheldon Street.
John Burdick looks on as Chris works on a mechanical drawing (the drawing board was in an upstairs bedroom). Another room was used for circuit board fabrication and a third room housed a large copy camera. Instead, Heurikon’s first product was an electronic timer-sequencer called the MTS-2000. It had a small keyboard for entering time values to control the on/off sequencing of 20 output signals. General-purpose microprocessors weren’t available then, so the device used hundreds of simple logic chips and a “brute force” design to implement the functions.
Only five of these instruments were ever sold, but that gave us our start. This was Heurikon’s first product, the MTS-2000, a multichannel timer-sequencer. It sold for $690.00. For the first two years, the operation was a part-time venture, and extraordinary efforts were used to keep costs down. For example, the circuit boards for the MTS-2000 were photographed, etched, and drilled in the basement workshop, and the benches were made from plywood and two-by-fours. In 1974, one of the engineers left the group, so Chris and the remaining founder, John Burdick, advertised for their first employee, an electrical engineer. That would be me. On my initial visit to the basement, I was at first dumbfounded by the dilapidated environment (I had expected to find something more like our current offices), but I was soon impressed with the keen vision and plans that Chris and John had for Heurikon. At that time, they were preparing a booth for an international machine-tool trade show in Chicago, and were about to ship MTS-2000s to NASA, JPL, Cal-Tech, and American Cyanamid — our first customers. Our First Micro About the same time as I joined Heurikon, the technology was changing from discrete logic to microcomputers. Within a few days, we bought our first Intel 8080 chip at a single-piece price of $360.00 (compared with only a few bucks today). My job was to design and build a prototype microcomputer around the chip. I clearly remember wire-wrapping our first board, double- and triple-checking the work, installing all the chips (except the processor!), and doing a smoke test. Finally, after everything else was working, I inserted the 8080 and crossed my fingers. I had the feeling that if that first 8080 went up in smoke, so would Heurikon.
Fortunately, it didn’t smoke — it worked perfectly on the first try. There weren’t any microcomputer programs back then, so the first bytes that we “toggled in” were for a simple routine that just blinked the lights on the front panel. But that was good enough for lots of cheers and many hours of gazing. We began to seek out custom design projects, and we built the business on them. For example, we made computer controls for outdoor advertising light displays, data acquisition systems, and point-of-sale terminals. While we were doing that, the microcomputer industry was being born in Silicon Valley and so, too, were companies such as MITS (who sold the first home computer) and Apple Computer (one of the few originals who are still around). At our beginning, we were not unlike some of those early ventures; I believe the main difference was that we were in Madison, far from the California-based microcomputer mainstream. All things considered, that may have been good; there was no frenzy here. The MLP-8080 Many of our early systems used the “MLP-8080,” our first microcomputer printed circuit board. (“MLP” stood for micro-logic processor. When we switched to the Zilog Z80 chip, we simply changed the P to a Z.) The MLP-8080 was a two-layer printed circuit board and sported 2,048 bytes each of ROM and RAM — not too bad for those days, but no match for the 10-layer boards with megabytes of memory that we make today. At that time there were no standard buses such as Multibus or VMEbus, so we defined our own bus and built a series of other boards and products based on the MLP-8080.
This MLP-8080 was Heurikon’s first microcomputer board. The wide power traces shared the two circuit layers with the data and control signals. The two 40-pin chips are the 8080 processor (center) and a UART for the single serial port (left). Each of the eight ROMs held only 256 bytes of data. Price: $1035.00. The systems we built in the early years were “turnkey.” That is, we took on the responsibility for the entire system working properly, and we wrote the customer’s entire application program. Our customers needed to add very little. On the other hand, we built only a few of any one thing, so we moved quickly from one project to another. In the following issues of The Horizon, I’ll describe some of those projects and the stories behind them. Although small by comparison with our current customers and production levels (today, we ship more every week than we did in all of 1979), those first customers and projects were responsible for our early success and are the foundation for our being here today. Many of the computers we built back in the ’70s are still in operation — you’ll be surprised at where. NEXT MONTH: Our name is put up in lights.
To learn more about the evolution of the home computer and those exciting times, read Fire in The Valley; The Making of the Personal Computer, by Paul Freiberger and Michael Swaine (Berkeley: Osborne/McGraw-Hill, 1984). Their book chronicles the people and companies who were responsible for the birth of the home computer and much of the microcomputer industry. Find out how Apple Computer, Microsoft, Byte magazine, and dozens of small fry got their start. |
|||
I oughta be able to find an older one if I keep digging... |
|||
Heurikon Historical Highlights
Between October 1991 and September 1992, I wrote a series of articles for the Heurikon Horizon, our in-house newsletter. Over the next several months I'll be posting those articles to exHeurikon.com (hopefully, a bit more frequently than when they originally appeared).
One of them describes the world's first portable computer. Yes, Heurikon built it. Stay tuned.
Oh, when you see "today" in the articles, that means circia 1992. Enjoy,
Jeff. |
|||
