When owners started putting their new Argonaut V rigs on PSK, one of the first things to emerge was reports of receive/transmit frequency offsets that were resulting in what looked like drift in the course of a QSO. As discussion evolved on the subject, a lot of attention became focussed on the fact that this was a PSK software problem and that it was necessary to adjust or tweak the soundcard settings. All that discussion is simple foolishness, based on a lack of understanding about how both the soundcard and the PSK software function. I have no intention of going into that aspect of the subject in this forum, but sound cards do not require tweaking to run PSK! If there is what looks like drift in the course of a QSO, the problem can always be related to frequency control problems in one or both of the transceivers involved in the contact.

The Effect of a Transmit/Receive Offset

Let's start with a basic example of two stations who start a PSK QSO. To simplify discussion about what happens, let's assume the following:
 

• Station A, with respect to RF,  transmits and receives on the same frequency.

• Station B, for whatever reason, transmits at a frequency which is 10 Hz higher (again, RF) than the frequency on which he/she is receiving.

• At the start of the QSO, both stations are showing the other at a frequency of 1000 Hz on both the waterfall and the numerical frequency readout.

• Both stations have enabled AFC (essential to track "drift") and set the "Net" function to "On". The latter assures that both stations should be generating their PSK transmit and PSK receive at the same audio frequency.

(1) Station A calls CQ and shows up at 1000 Hz on Station B's display. <B> clicks on the trace and his frequency now shows 1000 Hz.

(2) Station <B> calls <A> and, if he were on frequency, his trace would show up at 1000 Hz on <A>'s display. However, he is transmitting 10 Hz high relative to <A>, so his signal on the waterfall displays 10 Hz low - at 990 Hz!

(3) <A> may or may not notice the difference, but his software does and shifts to 990 Hz because the AFC function is ON. Now, when <A> replies of the call from <B>, the soundcard will generate a 990 Hz tone because the Net function is ON.

(4) When <B> receives <A>, the AFC will cause <B> to drop to 990 Hz on receive and, then <B> transmits back, the souncard will generate a 990 Hz tone because the Net function is on. Since <B> is transmitting (RF) 10 Hz high, the received tone at <A> will be 10 Hz lower - 980 Hz! <A>'s AFC will move the receive center frequency down to 980 Hz and the Net function will cause <A> to transmit a 980 Hz tone when <A> goes to transmit.

Note that although both stations are absolutely stable in terms of RF, the result of the 10 Hz offset will be that the trace will step down in frequency by 10 Hz on each go-around! It is also worth noting that, while this step-wise movement will eventually become obvious, given the resolution of the waterfall display, such an offset (10 Hz) would be undetectable in virtually any other mode!

Soundcard Problem?

So, is there a problem here with the soundcard software or hardware? The answer is an emphatic NO! The software and hardware are performing exactly as they should and no "tweaking" is warranted!

So Why the Offset?

The effect noted above can be the result of drift by one or both stations or a fixed offset on the part of one or both stations. One of the most common sources of offset is the RIT function. When operating PSK, be absolutely sure that the RIT is OFF at both ends. Note that I'm emphasizing OFF, not just set at zero! You could easily get a 10-20 Hz offset with the RIT control centered, so OFF is what you want.

The Fix?

A am running Version 1.04 of the Argo V firmware and experiments with the HDSSTV digital mode confirm that there is an 8-10 Hz RX/TX offset. I have not yet checked if the offset is presenbt with the Ver. 1.06 software. The HDSSTV AFC function handles the offset in that mode and the fix for PSK is ultra-simple. Leave the NET function ON until you have established contact and then turn it OFF! The two traces may then jump back and forth +/-10 Hz but the trace will not "migrate/step" as the QSO proceeds.