View Single Post
  #28  
Old September 6th 17, 10:24 PM posted to rec.aviation.soaring
[email protected]
external usenet poster
 
Posts: 580
Default XCSoar With LNAV?

Actually, I don't know for sure that there aren't dropouts or interruptions in the GPS/baro sentences data streams, just that there are no hesitations or flickering. The "missing" FLARM target data appear for one or several seconds, then disappear. The climb rates are more often missing than displayed but when they do display, they're also there for one or more seconds. The same with the IDs on the targets.

I
I'll try "monitor device" this weekend. What am I looking for: a FLARM target sentence that only appears once in a while even though there's a target within range continuously? Can I test that without having other FLARM-equipped aircraft in range? I can say when testing on the ground that power traffic targets move smoothly across the map display without any erratic behavior. I can't recall if the erratic targets are more likely when there are a lot of them vs. just one or two.

NOTE: found this in the FLARM Data Port Spec doc:
Data on other proximate aircraft, intended for connected devices with sufficient CPU performance. This sentence should be treated with utmost flexibility and tolerance on a best effort base. Individual parameters may be empty.. The sentence is only sent when port baud rate is 19.2k or higher. In case of serial port congestion or high CPU load, this sentence may be omitted for several objects independent of the alarm level.

So if I set the baud rate higher on the FLARM and the Top Hat device, would that cause any problems?

Chip Bearden

On Wednesday, September 6, 2017 at 3:27:50 PM UTC-4, wrote:
Interesting situation... Are you really sure that the Kobo is getting all of the GPS and baromatric data? Or, is it possible that it is getting 50% - 75% of it but the data is transmitted so often (every second) that you don't discern any data packet loss from those streams?

Like Rob, it is rather puzzling that one data stream is erratic and the other two not, which is why I ask the above question. Maybe try the monitor device feature for a few minutes and watch the data flow to see if you can detect any gaps. Maybe the USB conversion HW isn't quite up to snuff.