The Saga of the Serial to FSK interface

So about a week ago I decided I wanted to build two Serial to FSK interfaces and try out the Software Generated FSK feature of Writelog.  After a quick trip to Radio Shack, I sat down and started to work.  I used the circuit descriptions from AA5AU’s RTTY contesting site.  Not knowing what my rig (TS-940) would need, I built up the first circuit on the page – using an NPN general purpose transistor and a 1K ohm resistor.  Plugged this in and all was well – almost.

My FSK was inverted, and the TS-940 has no “menu setting” to change the FSK polarity.  So, back to AA5AU’s website, and at the bottom of the page is a circuit using a PNP transistor to use if the NPN version above transmitts inverted FSK.  So, got the soldering iron hot again and built up the circuit.  Much to my surprise and dismay, nothing!  Ok, maybe I wired something wrong, or had a bad transistor, or got the transistor too hot.  Tried rebuilding it, again to no avail.  So I joined the RTTY reflector and sent out  a message.  Lots of answers, some helpful for their thought provoking, others notso helpful.

I spent a 2nd evening after getting these responces from many folks trying to figure out what I had done wrong.  I tested transistors to make sure the pins were not mis-marked, I tried using a heat sink, I tried connecting pin 5 to the shell of the connector.  Nothing.  Off to bed in frustration.

Third evening at the bench I tested to make sure that something hadn’t happened to the RTTY input of the rig.  Sure enough, shorting the tip to the ring of the rca connector caused the signal to shift.  So that was working OK, but for some reason the circuit wasn’t doing the job.  I could see that the port was changing states as the USB/Serial converter has an LED to show activity.  What was goign on.  In my testing, I just happened to connect the FSK interface to the radio while the rig was keyed – the tone shifted!  In it’s regular state, the circuit was closing the RTTY line and causing the tone to shift.  Very interesting….so I left my “just follow the directions” mode and put my logic hat on.  What I wanted to happen was for the RTTY line to be grounded when the FSK was keying, not when it was not.

In testing the transistor, I remembered that there is contunity (or should I say low resistance) between the collector/base pair and the emmitter/base pair when there is nothing connected to the transistor.  So, if you look at the PNP circuit at the bottom of AA5AU’s page, you’ll see that the setup will ground the RTTY line when it’s connected, because there’s low resistance between the collector (which is connected to the rig) and base which is connected to ground).  I don’t want a connection in that state, I want the connection in the shifted state.  So maybe the collector should go to the rig, and the emmitter should go to ground, and the base should be connected to the signal line from the serial port.  That way when there is voltage present on the base, it will cause there to be a low resistance path from collector to emmitter.

So, I put the hypothesis to work and built up a circuit.  Phew – no smoke AND it worked.  Here’s a schematic of what worked for me to key my TS-940.

I’m no technical guru, but it doesn’t seem to me that the circuit shown on AA5AU’s page will work.  Hopefully I havnen’t done anything that will cause grave harm to my computer or rig, but I built two interfaces using the circuit above and both work just fine.

I forgot to upload the pictures I took of the interface last night, so I’ll be adding some pictures to this post later today or tomorrow.  I think it came out looking pretty good as I was able to build it into a DB9 all metal connector shell, and the panel mount RCA connector fits perfectly where the cable would normally exit the “back” of the shell, making for a neat little package.

Now hopefully I’ll see you all in the ARRL RTTY Roundup in January – I should be able to hear you better as I’ll be able to use my narrow CW filters instead of the wide open SSB filters!

73 de Al, KE1FO