The Classic Micro TO Keyer

K8CU again builds a classic 1967 QST semiconductor project
during the 40th anniversary year of the original project publication.

The finished keyer in 2007. The cabinet and front panel are the same ones used in the first project from the late 1960's. The cabinet has been repainted but the front panel is original. The handsome five-pound solid brass key was made by Jerry, K8RA.

Many are interested in vintage radios along with equipment restoration and their operation on the air. Interest in vacuum tube technology is evident too. Another area of growing interest is vintage semiconductors. Yes, some integrated circuits (IC's) are getting that old! Bell Labs was making transistors in 1947 and ICs were soon to follow. Experimenting with these vintage semiconductor technologies has many of the same joys as the older commercial radio equipment except that the cost is of course many times lower. Surprisingly it isn't that difficult to acquire some early components and web sites have sprung up with information and supply sources. This solid-state keyer project uses late 1960's vintage Fairchild uL923 JK flip-flops and a uL914 logic gate of the now obsolete RTL logic family. Some devices are still available "brand new" in the original factory package and are guaranteed to work! The good news is that sometimes the cost after inflation is equal to or lower than the original selling price decades ago.

Introduction QST for August, 1967 described a solid-state version of the popular W9TO keyer. This project by K3CUW also appeared in the ARRL Handbook during the late 60's and early 70's. The original "TO" keyer used vacuum tubes, while the integrated circuit version sported small integrated circuits by Fairchild Semiconductor. These ICs were fabricated with RTL (Resistor-Transistor Logic) and looked much like small signal transistors except for their 8-lead construction and epoxy cases. Two early 1970's Fairchild uL923 JK flip-flops and a uL914 logic gate. Fairchild used the copyrighted name MicroLogic ® and uLogic® for their first ICs.These ICs were some of the first sold commercially. These ICs implemented basic logic functions and contained about 15 transistors with some internal resistors. This is why they were called RTL (Resistor-Transistor Logic). RTL was an early technology and soon gave way to more noise immune and faster devices in the DTL (Diode-Transistor Logic) and TTL (Transistor-Transistor Logic) families. Since I had already successful constructed the W9TO vacuum tube keyer, I decided that a smaller and more efficient IC version of the project was for me. My young ham radio friend, Gary Straub WA8RXQ (later K8BU and now a Silent Key) constructed this keyer too. However, we both had the same problem that spurious dots would sometimes occur in the code pattern. We both carefully checked our construction for assembly errors, but all appeared according to the circuit schematic. After seemingly endless checking we were baffled. I wasn't entirely comfortable debugging the IC logic. This after all was my very first IC digital project and it was a bit intimidating for a youngster. A while later Gary called me with the news that he had found a fix for the circuit after experimenting with bypass capacitors. He said that apparently a glitch was triggering one of the integrated circuit flip-flops and that the addition of a single 1000-pf capacitor to the signal lines between the two IC flip-flops added enough noise immunity to eliminate the false triggering! This was good news indeed. My Micro TO keyer project then worked after I added the glitch suppressor capacitor, and I started using the new keyer on the air.

Building the Circuit in 2007

After I decided to build this keyer project again, it was necessary to locate the three integrated circuits. Since my junk box had none I decided to see what an internet search would turn up. I located an interesting website that offered both classic transistors and integrated circuits. (Note 1). The circuit was assembled on a small piece of proto-board. I used the original schematic as shown in the 1968 ARRL Handbook. The power supply is not regulated and consists of a single 6.3-volt center-tapped filament transformer in a bridge configuration. Through-hole components were used for all parts because they were originally used in 1967. Modern-day version of the Micro TO keyer project uses through-hole components and Fairchild IC's from the late 1960's. Circuit Operation Today I was hoping that this recent version of the project would have the same problems experienced as in the late 1960's. That way I would be able to take advantage of my experience gained since then and be able to analyze the circuit. However the new circuit worked perfectly for me.(Note 2) I suspect that the original problem may have been noise triggering the flip-flops by ground noise. The use of a heavy grounding method in this latest project may have eliminated problems in this area. The original article described the pulse generator as "somewhat novel". It is an interesting circuit and does resemble a unijunction-transistor configuration. I used venerable type 2N2222 and 2N2907 silicon transistors with good results. The pulse rise and fall times are sharp and this helps to insure noise-free triggering of the integrated circuit JK flip-flops. The photograph shows the waveform of this pulse generator. The oscilloscope display reveals that the logic voltage swing is about 1.4-volts which is normal for the RTL logic family. Micro TO keyer pulse generator output The two Fairchild uL923 JK flip-flops act like frequency dividers. Along with the uL914 logic gate they provide for perfectly formed code characters. The circuit features self-completing dot and dashes. This means that once either element is started it is impossible to stop the correct code timing from completing. A characteristic of all keyers with self-completing characters is the possibility for an operator to sometimes miss a dot element if the operator "leads" the keyer (presses the dot paddle before the dash element is done). Note that this keyer is not "iambic", so modern-day "squeeze" keying is not possible. This is not a problem and I can generate perfectly formed CW characters with this vintage design keyer. A single lever paddle works fine, but a dual-lever paddle like the one I have today works also. The photograph shows the output of the Micro-TO keyer's dot flip-flop. Note that some power supply AC ripple is evident on the top of the waveform, but this is not a problem at all. Micro TO keyer "Dot" flip-flop output

IC Semiconductor Alternatives

Conclusion

Notes

1. Get your uL914's and uL923's from The Semiconductor Museum. 2. Circuit revisions: - after construction it was noted that a single dash would not self-complete; the output would stay on forever until another pulse was allowed to generate from the pulse generator. The problem was found to be a missing resistor on the schematic! The circuit as originally drawn in QST and in the 1968 and 1970 Handbooks is missing R12, a 100K resistor from the base of Q2 to ground. Otherwise, the use of a low-beta transistor at Q2 would also work. The addition of the resistor is the preferred modification. The ARRL Handbook for 1972 has a corrected schematic. Other issues of the Handbook have not been checked for accuracy. QST for March, 1968 page 48 has author feedback about this. 3. You may obtain a copy of the original QST article (Aug. 1967 pg. 17) from the ARRL Technical Information Service. Download the Micro-TO schematic here. 4. A useful technical reference is the RTL Cookbook. Try amazon.com, abebooks.com, or eBay for your copy. A complete guide to the understanding and using of Resistor-Transistor Logic (RTL) digital integrated circuits. By Donald E. Lancaster Published 1969 Howard W. Sams & Co. ISBN 0-672-20715-X