In 1983, Rádiótechnika magazine published an article about a clock built around the TMS1122 IC. I redesigned the circuit and the PCB somewhat, then built it. It served continuously for 27 years, essentially without being turned off. Due to lack of space, I recently scrapped it, though it was still in working condition.
The Clock
The TMS1122-based clock displayed time on a four-digit, seven-segment display and showed the days of the week on an LED bar. It could also control four outputs. I used one for the internal alarm and brought out the other three to connectors, using reed relays instead of the open-collector outputs shown in the Rádiótechnika schematic. One of these outputs turned on my radio—I used to wake up to it during my student days.
I also added a precision time base built with an SAJ300T IC to the board. The rather large keypad was used to set the time and program the clock. The DCF77 time standard transmitter near Frankfurt was already operational back then, but in Hungary, you couldn’t buy radio-controlled clocks yet, and building one was quite a hassle. Low-power microcontrollers were not available yet—you would have needed an Intel 8085 or a Z80A, practically requiring a power plant. And you’d also have had to build a longwave receiver for DCF77 reception.
The TMS1122 was a good solution at the time. You can still occasionally find it on auction sites today, but it’s hardly worth the trouble now. Its LED display is multiplexed but still consumes a lot of power, so it needs mains power. The TMS1122 itself is relatively low power and can run on a NiCd backup battery during power outages, but its consumption is too high for a few long-life lithium button cells to suffice. Naturally, the display doesn’t light up in battery mode. Maintaining NiCd (or NiMH) batteries is also a hassle.
When the backup battery is depleted, the clock doesn’t just stop—the entire program is lost because the IC stores the schedule in RAM. When power returns, you have to reset not only the time but also reprogram all the timers. This makes using the TMS1122 somewhat inconvenient.
About the TMS1122
The TMS1122 is part of Texas Instruments’ TMS1000 4-bit microprocessor family. The first chip in the series, the TMS1802NC, was introduced in 1971 and contained a complete calculator application. At the time, it was remarkable because it integrated the processor and memory (1024×8-bit ROM and 64×4-bit RAM) in a single package. It was used in the Texas SR-16 scientific calculator and the Big Trak programmable toy vehicle.
The TMS1122 clock IC hit the market in 1974 alongside other family members. For example, the TMS1000 NL 3228 was a musical doorbell with 14 melodies, and the TMS1000 NL 3227 was a Mastermind-like logic game. Built with PMOS technology, the chip consumed about 6 mA. Its logic levels and I/O pins were TTL-compatible at +5 V, which was standard back then (and thus compatible with NMOS and CMOS if powered at 5 V).
The CPU ran at 0.3 MHz—painfully slow by today’s standards—executing one instruction in 10–15 μs. The base version came in a 40-pin package with 23 I/O ports. Customized versions omitted unused pins and came in 28-pin packages. The ROM program could not be changed after manufacturing; it was hard-coded into the IC during production. To modify the program, you had to redesign the lithographic masks used in manufacturing.
Modern Alternatives
Nowadays, you’re better off building such a device with a PIC or AVR microcontroller, which stores the program in EEPROM so it survives power outages. These controllers consume so little power that you can easily run them on batteries, or you can attach a real-time clock (RTC) module running on a button cell that keeps time even when the processor has no power. DCF77 receiver modules are also readily available, making it easy to build a radio-controlled clock with near-atomic accuracy.
Inside the Built Clock
In the picture: the TMS1122 is in the center, to its right the SAJ300T time base and quartz. The ICs on the far right formed a gentle, soft-start alarm tone generator. On the left, the 75492 IC controlled the display cathodes, while the resistor-transistor network on top drove the anodes. The display consisted of a single-row 3 mm LED and two East German-made green VQE 2-digit LED displays. The blue packages at the bottom left are the reed relays.
Off to the trash
Eventually, I replaced my trusty alarm clock with a smaller, radio-controlled projection clock with a built-in radio receiver. The old device languished in a box until I decided it was electronic waste. A used TMS1122 might fetch 100 forints—hardly worth desoldering.
A little sad, but that’s life. Electronic devices usually become obsolete long before they fail. According to research by the European Environmental Agency (EEA), an average EU citizen buys 20 kg of electronic devices per year, which on average serve 2.3 years less than their planned lifespan. For example, televisions are designed for 25 years but are used for only 7.3 years on average—18 years less than intended. They don’t break down; people just get bored and replace them. Similarly, vacuum cleaners and washing machines designed for 10 years usually last eight in households. Smartphones are designed for a 5-year lifespan but typically used for just 2.
Environmentalists campaign for longer-lasting products and better recyclability. Eternal whiners, on the other hand, rant online that manufacturers are ripping them off with increasingly flimsy products, while “the old stuff was built to last.” The truth is that products either become outdated, or users ruin them.
For example, the legendary Hajdú Energomat washing machine, popular among retro fans, still works if refurbished—true. But it uses so much water and power that the savings from two years of lower utility bills would cover a brand-new washer. It’s heavy as a tank, jumps around like crazy, and will turn your wool sweater into a doormat.
Most smartphones don’t become unusable after two years because of structural failures but because their screens crack or their storage gets clogged with apps and photos. If they didn’t break, they could be restored—but users don’t know how. TVs? People just get tired of them: sticky remotes, scratched screens, and besides, they want a bigger one for the World Cup because the neighbor just got a new set.
So yes, there are cases—like my old clock or many computers—where obsolescence is inevitable because the device can’t meet expectations. But in most cases, users are responsible for rapid (and often physical) obsolescence. Don’t blame the manufacturers for making “short-lived” products. They could make longer-lasting ones and still turn a profit—but why bother? People are fine with things as they are… We’re only destroying the Earth and the future of today’s youth—no big deal!
PS: The original article about building the clock can be found in the February 1983 issue of Rádiótechnika, on pages 50–53. An article on adding a seconds display appeared in the February 1988 issue, pages 58–61. Additional ideas for the clock were published in the July 1988 issue, on pages 357–358.
UPDATE: Long after writing this article, I discovered that the Mechatronika Cooperative in Gyöngyös manufactured this clock sometime in the 1980s. Here are some photos of the factory-made version. The company most likely evolved into DIGITERM Technical Development, Manufacturing and Trading Ltd. in 1991.
