A Simple CW Keying Interface*

Functional, Easy, and Cheap
All Good Ham Radio Characteristics

When I discovered that I could make my logging program, ACLog, read the frequency and mode from the big rig, I also saw that I could cause the rig to jump to the frequency of a DX spot just by clicking on the spot. Scott, N3FJP, has a very capable, intuitive, and flexible logging program here. When I mentioned this to a fellow club member, and darned good op, Dale Stoy NSōD, he said ďWelcome to the 21rst century.Ē Of course heíd been controlling his rig like this for a long time. When building buddy and super op Steve Miller NōSM, mentioned that he had his logging program keying CW on his rig, I felt like I was still in the dark ages, and had better hurry and catch up.

Why Do It?

Although I really enjoy sending manually, Iím neither fast, nor good. But it is very enjoyable, and akin to ďbeing one with the radioĒ. Anyone else ever have that feeling? So why should I want to make my logging program key the rig? Only one reason:

Itís fast, efficient, and accurate transmitting when in a contest and running a frequency (it sometimes happens, even at QRP).
Operating a Special event Station.
Quick, easy, and accurate logging (and MUCH less work than paper)
Easy generation of an ADIF file (emaillable doncha know)
OK, so thatís thatís a few more than one, but you get the idea.

So I set about working up a interface circuit with these design goals.
1. Simple (or it would be well over my head)
2. Functional (it canít mess up characters, I use my paddle for that)
3. Easy to build into a connector shell (I already have a ratís nest behind the computer)
4. Cheap (meaning donít buy nothin)
5. Reliable (I donít ever want to have to crawl under the desk to unplug it)
6. Complete isolation between the rig and the computer (no ground loops)

Road Map to Success

Below is a description of the fun that can be had on a simple project.. It was a great way to spend part of an afternoon, therapeutic even. You may wish to scroll or jump down to the test circuit schematics and a picture of the cable connector at the bottom of the document.

Circuit #1

The first circuit I cobbled together was a 4N32 opto (isolation remember?). A 270 ohm resistor, 1N914 diode, and an audio oscillator on the output to monitor. You know the circuit, itís on WM2Uís web site.. A resistor in series with the optoís pin 1 and the diode across the input. ACLog wants the DB25 pin 1 to be gnd and the signal pin to be 17. It worked great, called CQ for 30 minutes at 25 wpm, using an audio oscillator, not on the air. I thought that 30 minutes would demonstrate initial reliability. It worked well the whole time. But the circuit had too many parts. Remember I wanted to build the entire interface into the connector shell. So in the interest of simplicity I removed some parts and tried again.

Circuit #2

I couldn't figure out why the reverse diode had to be across the input. The parallel port sigs go from zero (or close) to +5. So why a diode? Cant imagine any spikes to shunt, the diode in the opto should turn on and off just fine on signal excursions, no reverse voltages to contend with, and no need for need any bias. So out it comes! Now it's down to just the resistor, and the opto.... 30 more minutes calling CQ and all was still fine.

Circuit #3

Next I wondered if the input resistor was necessary. A little current measurement shows the opto diode drawing 24 ma on 5 volts with no series resistor. Hmmmm wonder if the port will stand that? I read somewhere that early parallel ports were kinda puny but that later ones were more robust (I like that word). So out comes the resistor Ė still works great! Now it was down to 1 part and itís looking elegant. One of the Four State QRP Groupís prime movers and OzarkCon organizer Joe Porter, WōMQY, sent a note advising that parallel port lines were rated at only 20ma. Whew! I dodged a bullet that time. I was actually going to ďfreeze the designĒ at that point until I heard from him, thank you Joe!
Circuit #4

Time to get rid of the noisemaker and in the process see if the opto's output transistor can hack it. So, I plugged the input resistor back in, rigged up a battery and a resistor, and away the little opto goes with a 25 ma load.. Half an hour later itís still perkin along just fine, and cool too. Hey Hey, maybe I've got it!! Or maybe not, the output transistor doesnít turn on very hard. It has about .8 volts from collector to ground. This may be ok, but I didnít plug it into the big rig, since the manual says it wants a half volt or less when keyed. I thought I had it here for a minute.

Circuit #5, or whatís next?

I've become fond of the little 2N7000's so I wondered if I could use one of those little thingys. Well why not try it. I did and it worked just fine, all by itself!! It turns on hard (millivolt range) and ran for 30 minutes solid at a 25ma load and didnít even get warm. At 13 cents apiece, thatís a really cheap one part interface.! Of course any small switching transistor could be used, the ubiquitous 2N2222A for example. Except a current limiting base resistor would probably be needed Ė back up to 2 parts..

The single transistor approach doesnít meet one of my design goals, that of complete isolation. I most cases isolation isnít a problem but itís seems like a good design practice here. I used an opto in my homebrew digital interface for the same reason.

Wrap up and build ( or quit messing around and get it done)

Bottom line: I decided to use the opto and a current limiting input resistor, keying the line directly with the optoís output transistor. This is circuit #2 with pin #5 connecting to a mono phone plug tip and pin 4 being rig ground. This circuit will key modern, non tube rigs, most of which have a positive line going to ground when keyed. Not only does it key the big rig, it also keys the ATS3 nicely as well.

In the course of ďspearmintinĒ, I had tried using a pnp transistor on the optoís output to key the rig, Circuit #6. But I noticed that the keyed C-E voltage drop was over a volt. So I scrapped that idea and looked at the optoís saturated voltage again. I was surprised to see it drop to about .1 volt immediately and then rise and stabilize at the .8 volts I had seen earlier. I donít know why that happens, but it keys the rig nicely, so Iíll stay with it.

The two parts fit inside the connector shell with loads of room to spare. The resistor needed a little clearance, so I pulled the cable out a bit before globbing on a wad of hot melt glue to secure it for strain relief.


Did I meet the original objectives? I think so, minimal parts count, it works well, all parts were in the junque box, a snap to stuff inside the connector shell, and provides complete isolation. Only time will tell the story on itís long term reliability, stay tuned.

Now if I could just learn to type.

Terry, WAōITP
16 April 2007

*Or, it doesnít take much to entertain an olí ham..

Scroll Back UP

Scroll Back UP
Return to home page