XcSoar/XcVario sync problems continued! about xcvario

[hjr]

Hey Graham,

could you please add details to your report?

thx

[GrahamD2]

Helmut I thought I had
From issue #141 using https://github.com/iltis42/XCVario/blob/master/images/sensor-master-21.1230-1736.bin

Eckhard
I am beginning to regret mentioning the ballast sync issue, I rarely, if at all use water as it is!! The MC works great thanks
Bi-directional comms XCV-XCS with ballast has now caused a bigger problem than when it was not working.
With zero litres on both XCV and XCS on power up or when I open the flight setup page on XCS, 42l has already been loaded from XCV, closing the flight setup page writes back 42l to XCV, repeating the process next time 81l is then loaded to XCS which is then written back to XCV and so this continues each time I open the flight setup page,116l then 148l and so on. Pressing the Dump button in XCS produces some very interesting results!!!
It is not possible to zero the ballast on XCS while XCV is running.

Entering a figure on the flight setup in XCS always results in the same number writing to XCV so this is fine. The problem is the 40l or so that is added on startup or each time the flight setup page is open.

With XCS "Sync TO" device ON and Sync FROM device OFF. XCS flight setup updates XSV exactly. Changing the setting in XCV has no effect as you would expect.
With the Sync to and from setting reversed and with the ballast set to zero on both. restarting the XCV will load 42l into XCS flight setup.
I know you suggested I report to XCS but I don't want to find myself having a three way discussion when I don't really know the cause of the problem, would a developer to developer not be a better approach.
For my part I would happy if ballast sync was disabled until resolved, but I am happy to do further testing in the meantime if required.
Happy New Year

[iltis42]

Dear Graham,
the issue reported looks like the reference data is not same in xcsoar and xcvario. Please check the reference wingload in xcvario and compare with the reference weight divided by wingarea in xcsoar, means reference wingload in xcsoar.
If there is still issue with same reference data, please open a new issue with this data on both sides, indicating the settings.
Please note that the issue with ballast transferred in android 7.21 from xcsoar cannot be solved in xcvario, so please kindly open issue there for wrong/cut precision of ballast sent.

[GrahamD2]

I have checked these parameters and they match.
This serious problem only started when you fixed the no ballast response vario to xcs last week, can you unfix it?
I happy to raise a ticket with xcs but from what I can see is XCV is adding 40 litres each time it sends the data to XCS and I think if I tell them that they will bounce me back to you.
I clearly don't understand what is happening here
Turning off the sync from device in xcs (vario to xcs comms) solves the problem but I loose the function I really want, MC.
Graham

[iltis42]

If there is no ballast reported from vario, there will be no update from vario, so this fix must stay. It may behave same as if you switch off sync from device, hence it may just cover another issue, but is not the root cause.
Please add screen shots of pictures from 'Polar Adjust' screen from vario and also 'Plane Details', showing wingarea plus 'Polar: ' dialog showing exact Reference Mass from xcsoar.

[GrahamD2]

Eckhard Link to my OneDrive files showing requested photos and a video showing the issue. Dry Mass of 425 kg on XCS is as per manual to include pilot weight Did you see my Ps using XCS 6.8.11 and Borgelt protocols. Problem unchanged. https://1drv.ms/u/s!AnZWH7uXxHdohpMeRmdx5dOY5PMZ6g?e=DumlCy regards Graham

[iltis42]

The things i see here:
First:
xcsoar 377kg / 11.4m2 equals to 33.07 kg/m2, whereas setting in xcvario is 30.19 kg/m2.
Also the wingarea differs slightly 11.4 in xcsoar and 11.36 in vario.
Note this data must exactly match, otherwise it won't work.
Second:
Also please note that Borgelt Protocol cannot be used, only XCVario protocol has been extended to exchange more precise ballast data. Borgelt devices are in field as is, and cannot change their protocol, so now way to improve something there.
Third:
Tests with 6.8.11 are useless, even XCVario protocol is not using extended precision in this older version.
And i am sorry for that, but it looks like a functioning xcsoar version is currently not avail at the moment with this extension. Please also consider: The discussion about bug in xcsoar is considered as off topic in here.

[GrahamD2]

EckhardThanks for your efforts, I think until it is resolved I will disable part of the two comms exchange and loose the functionality I am interested in, MC and QNH and will check updates from time to time so please consider this closed from me for the time being. I have tried to match the loading but the weight and area steps in the settings on both XCSoar and XcVario don’t allow to get an exact match. Whilst I understand the relationship with ballast and wing loading, I don’t understand why the units don’t simply exchange litres (XCSoar seems to send litres to XCV without problems) and let each sort the wing loading calculations. If you decide to stop exchanging ballast data while the problem exists please let me know.Thanks again for your attention Graham

[iltis42]

Yes, i guess disable this at first shall be okay, hence i want to encourage you to write a bugreport in xcsoar when you encounter issues that are already confirmed, such as ballast precision is got cut in 7.21, otherwise a new version will never happen that fixed this problem.
An exact match for the masses is not needed, but as close as possible or less that 0.2% delta should be okay.
The protocol had been defined by cambridge long time ago and inherited to borgelt and xcvario. Limits in there are well known, and yes i agree there might be benefits from exchanging litres in future, hence its not the root cause of the trouble you see.
This simply comes first from different reference data, and second from rough ballast updates in the older version (solution for that already merged). And once your reference data is different, mostly all other polar calculations won't match, or at least do something that is very hard to explain e.g. different S2F, different netto vario, in both units.
The ballasting things are under work in xcsoar as well, xcvario is a bit advance providing crew weight and empty mass, but this will come this winter i guess.

[GrahamD2]

Eckard I have raise an issue with XCSoar, I hope I am not digging a big hole for myself!!! Regards Graham From: Eckhard Völlm ***@***.***> Sent: 01 January 2022 21:56 To: iltis42/XCVario ***@***.***> Cc: GrahamD2 ***@***.***>; Author ***@***.***> Subject: Re: [iltis42/XCVario] XcSoar/XcVario sync problems continued! (Issue #144) Yes, i guess disable this at first shall be okay, hence i want to encourage you to write a bugreport in xcsoar when you encounter issues that are already confirmed, such as ballast precision is got cut in 7.21, otherwise a new version will never happen that fixed this problem. An exact match for the masses is not needed, but as close as possible or less that 0.2% delta should be okay. The protocol had been defined by cambridge long time ago and inherited to borgelt and xcvario. Limits in there are well known, and yes i agree there might be benefits from exchanging litres in future, hence its not the root cause of the trouble you see. This simply comes first from different reference data, and second from rough ballast updates in the older version (solution for that already merged). And once your reference data is different, mostly all other polar calculations won't match, or at least do something that is very hard to explain e.g. different S2F, different netto vario, in both units. The ballasting things are under work in xcsoar as well, xcvario is a bit advance providing crew weight and empty mass, but this will come this winter i guess. — Reply to this email directly, view it on GitHub <#144 (comment)> , or unsubscribe <https://github.com/notifications/unsubscribe-auth/AXC4ADMLBEZTXPZG2ILSWELUT5Z7XANCNFSM5LBOMSCA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub> . You are receiving this because you authored the thread. <https://github.com/notifications/beacon/AXC4ADOI36ZM7AN633CMMXLUT5Z7XA5CNFSM5LBOMSCKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOHPJBM5Q.gif> Message ID: ***@***.*** ***@***.***> >