Flight track errors

[Oblivian]

Hi Mike,

Charky just pointed out some live examples of what he has been seeing over his airspace with AC going a little crazy.

Not your usual timestamp forward/back but some very bad plot information.

I have seen similar here in 2 cases, but did not think anything of it. Any ideas on the cause if you look at the incomming data around this time?

I can only but think there is some bad GPS data being seen and not filtered out by the server realising such a big change?

ploterror.jpg

[Oblivian]

Lyon (that general area) appears to be ryfe with them. 3 more at Present in the area

[Mike]

This happens when incorrect position data comes into database. The reason could be transponder error, reception error or receiver error.

Some aircraft have problems with their transponder and transmits incorrect data. Could be wrong position, random position, wrong speed, wrong heading.
Sometimes there may be a problem when receiving the signal. This is handled in most cases by checking if the check sum is correct, but sometimes error can come through.
There are also problems with some receivers, especially microADSB. We have seen that feeders with this receiver almost daily upload random incorrect data. We don't know why this happens to this receiver but that is why we don't recommend feeders to buy this receiver.

When using the new feeding software this problem should be reduced as the software can only upload data that is reporting a position about 500 km from the receiver. Other positions are ignored.

[Oblivian]

It would be interesting to know if the feeders in that area causing it are running a Beast with the new FW with the spurious HEX codes fix when they were received in the error data/noise decodes.

It may now be filtering these false HEX, but letting bad position data through.

There is only NZCH and myself sharing to FR covering the 200nm radius around here. And I witnessed this phenomenon myself a few nights ago at about 120nm. So was possibly me uploading the bad data with the new feeder.

I have to keep close eye on my local area and find where it is being caused in the logs if I see it again.

[Oblivian]

Have been checking differences between my setup and another user seeing many of these zig-zag ones.

The only common feature is the planeplotter version, New uploader and fact they are from a Beast.

I have trust the uploader does not manipulate the data before sending, so focus will be going on planeplotter which provides the data to it.

SBS, new PP and uploader going to be tested shortly to see if it continues. (rule out beast)

I thought 1.4x FW may have introduced it. But the other 2 units experiencing it are on 1.36 and 1.2. So perhaps the error correcting was adjusted in new PP, as one change was done to correct HEX truncating with certain characters in callsigns.

[Charky]

Hi
I am guilty. Its seems to be my receiver doing that after some time of running. There are not only the hooks there are also around 80km south of my current position planes grouped with wrong positions. After a PC restart the planes are shown at their correct positions some 100 km away (Cache problem with W7 ?). Sometimes the same plane is shown at 2 positions at the same time.
Im using the Beast plus PP plus fr24feed on W7 netbook running currently testwize as T-EDFY1.
The firmware of the beast is the "old" 1.32. PP soft is the latest version 6.2.6.2.(fJ)
This hooks are not visible at the PP screen, nor at the secondary upload to Planefinder. Switching off the 2.upload has no effect to the hooks.
Uploading to Net1 (PP plus 2.upload) doesnt show this problem.
Using a SBS1er with BS and old and new upload doesnt show this problem.
If somebody has an idea please let me know.
I could simply switch it off but that will not help to findout where the problem is.
regards

[Charky]

Hi,
I made a logfile. what I found so far, sometimes old timestamps are visible.
The file enclosed is a textfile put into excel with no sorting, so should be represent the correct order the messages have been received.
If somebody has some time to have a deeper look into it that would be very nice helping to find the error.
thank you

sorry excel is too big, only textfile

[Mike]

We will check the document.

[Charky]

sorry, to nerve you Mike
here an example of double plane at the same time. upper circle = wrong position, lower circle = correct

[Charky]

fyi
uploading currently using Beast, PP (2.upload to PF), ModeS (LSGG)
no hooks and no dancing planes. running now for a few h.
(only the jump from udp to http)

[Mike]

What firmware do you have on the Beast?

[Charky]

the 1 Im using currently is 1.3x (not updated to the latest 1.40/1.41)

[Oblivian]

And I'm on 1.41(beta) and have had an instance of a zig-zag here

[Oblivian]

Had one again today my time.

And only me this end of the country on FR24feed.exe

trackoops.jpg

[Charky]

yesterday I was running ModeS again for upload.
There were very few hooks and Im not sure if they were coming from me. They were a bit too far away.
The "hanging, double" planes were not visible.
today back on fr24feed