VAC Doppler Shift

USB headsets to digital audio workstation software...
W6LVP
Posts: 14
Joined: Wed Aug 22, 2018 11:47 pm

VAC Doppler Shift

Postby W6LVP » Sun Sep 04, 2022 9:13 pm

I own two great ANAN radios with the latest P2 firmware that run Thetis V2.9.0. Most all of my operation is digital modes with recent focus on VarAC and its underlying VARA modem.

As part of trying to optimize Thetis VAC, I think I have discovered an interesting anomaly.

WARNING: What follows may be completely bogus so take it for what it's worth.

The objective the VAC audio channels is to deliver a steady stream of audio in both directions between Thetis and some digital mode software like VARA, FLDIGI, or WSJT-X.

The concept seems to be to use buffers with separate input and output pointers so that irregular input to the buffer can be compensated to deliver steady output from the buffer. The ideal location of the output pointer is in the middle of the buffer so that instability at the input can be adjusted either direction up to half the length of the buffer. The automatic adjustment of the buffer output pointer is a bit like a mechanical servo. The ideal buffer midpoint is adjusted by feedback in an attempt to keep the pointer in the middle of the buffer.

When VAC is started or when adjustments to Thetis VAC are made, the servo is destabilized and goes into a dampened phugoid oscillation. The movement of the pointer in the buffer is a bit like moving the tap on a delay line. Movement of the pointer generates effectively a Doppler Shift FM modulation of the buffer output audio stream. For speech audio, the frequency shifts are small enough not to be perceptible. However for digital modulation, the shifts are well beyond the ability of the digital software decoder.

WSJT-X has a mode called FreqCal where you can point the radio to 1,500 Hz below a reference frequency source such as 10 MHz WWV and then track the resulting audio output that should ideally be 1,500 Hz. The image below shows what that looks like for the output of Thetis VAC just after it has been destabilized.

Seems like there are two solutions:

1. Wait until the phugoid reaches stability after several minutes.

2. Click the Force tick boxes to lock the pointer and accept a few errors.

I hope someone who understands this much better than I do can shed some light.


Doppler.jpg
Doppler.jpg (182.98 KiB) Viewed 2552 times
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: VAC Doppler Shift

Postby w-u-2-o » Sun Sep 04, 2022 11:53 pm

This topic has been done to death.

Short answer:

If you make the right adjustments it should work perfectly, and better than ticking the "force" boxes. Adjustments are a little different for everyone. If that's too much trouble and you get the performance you want (stability and number of audible artifacts/glitches) with the force boxes checked, do that.

Long answer:

Read this topic first: https://community.apache-labs.com/viewtopic.php?f=6&t=3078

Then read this topic, particularly pages 4 and 5: https://community.apache-labs.com/viewtopic.php?f=13&t=3901

Recommend you continue the discussion, if necessary, in one of those existing topics, rather than creating a third.

Return to “Digital ("Virtual") Audio”