(Do)TDoA for FTx: Fun with a very slightly modified jt9 build.

[Adrian Boyko]

Hi everybody,

Just thought I'd share the results of some hobby experiments that are based on a very minor modification of WSJT's jt9 program. To be exact, I modified decoder.f90 so that "DT" values are output as f7.3 instead of f5.1. This exposes the full resolution of the DT values calculated inside jt9.

In my case, I used the modified jt9 to simultaneously collect decodes from a handful of web-based SDRs and then used that data to perform a variation of TDoA that I call DoTDoA. The extra "difference of" eliminates the need for various time synchronizations.

Here are a few images that illustrate how well DoTDoA performed in the case of LZ2KV at KN33.  The first, below, shows the expected LoPs (lines of position) in bright green along with the observed fuzzy LoPs in orange.

The next shows the fuzzy LoPs combined into a hotspot:

The final image is just a zoomed-in version of the previous image so that smaller details are clearer. The dark green ellipse is based on the geometry of the locations of the receivers and the transmitter. The red contour is a particular level of the hotspot. The red dot is the hotspot's max. The black rectangle is KN33.

If you're interested in knowing more, check out my latest post on the topic. Perhaps this minor modification to jt9 will prove useful to other HamSCI readers who could somehow benefit from better FT8 time data. If so, maybe we can talk the WSJT-X project into adding a command line option for jt9 which switches on the f7.3 formatting of DT. Such an option wouldn't be useful for WSJT-X itself but could be useful in other contexts where jt9 is used by software like ft8modem or OpenWebRX.

Adrian Boyko / kr0dak