Dashboard bug with Signalk

[jlsail]

Hi
If using signalk for data and value of environment.depth.belowTransducer is an even number
without decimal. depth shown in Dassboard is 0.0m.
In attached picture 5 gives 0.0m. If value in server are 4.9 gives 4.9m and 5.1 gives 5.1 meters.

Also data given in Dashboard is from transducer not surface.If I take data direct in to Opencpn from my B&G system I get depth from surface.

Same behavior in "normal" version and flatpack version.

Regards
Johan

[Hakan]

Johan..
re: SignalK depth = 5 m
Yes, strange! We didn't expected SK to send a value without decimals so when we make it to a "double" from the JSON string it fails. The fix is obvious so thanks for the catch.

The depth instrument in O are rather a complex situation.
Historical we used DBT ("Depth below transducer") and DPT where the main data is "Depth below transducer" but optional also included "offset from transducer", (positive for distance to water line or negative for distance to keel)
When Dashboard receive DPT with an offset it's used to correct the value before presentation. As you've noticed.

So when we added SignalK to the Dashboard we selected to use "environment.depth.belowTransducer".


The value without decimals from SK can't be corrected until next release or beta.
You can always correct your own offset in Dashboard settings. (Right click in a Dashboard->Preferences->Appearance->"Depth offset")
Håkan

[jlsail]

Originally Posted by Hakan

Johan..

The depth instrument in O are rather a complex situation.
Historical we used DBT ("Depth below transducer") and DPT where the main data is "Depth below transducer" but optional also included "offset from transducer", (positive for distance to water line or negative for distance to keel)
When Dashboard receive DPT with an offset it's used to correct the value before presentation. As you've noticed.

So when we added SignalK to the Dashboard we selected to use "environment.depth.belowTransducer".


The value without decimals from SK can't be corrected until next release or beta.
You can always correct your own offset in Dashboard settings. (Right click in a Dashboard->Preferences->Appearance->"Depth offset")
Håkan

Hi Håkan
Thanks for your explanation.

Personal I think it is dangerous and confusing that opencpn handles depth diffrent if data comes from Signalk or from nmea/tcp.How will you know your depth If you have both a connection to signalk and to nmea with different priority for backup?
Changing in Dashboard preferences will make the value not from signalk wrong (the one that is actually right).

I dont know anyone talking about depth under transducer during navigation. People normaly use actual depth or some times dept under keel.

cheers
Johan

[Hakan]

Johan.
I can't say we handle it different. The NMEA0183 DBT ("Depth below transducer") and DPT where the main data is "Depth below transducer" is the same as SK: "environment.depth.belowTransducer"
If you decide to send an offset from an instrument using DPT you're supposed to be aware of that. That's the logic I suppose. (No offset data are factory default from the instrument.)
Håkan

[Hakan]

Johan.. @All..

OK. we handle it different in one way. We could have used the offset also from SK, as from DPT, if available?
Would that be a solution? (That is Depth from surface)
Håkan

[jlsail]

Originally Posted by Hakan

Johan.
I can't say we handle it different. The NMEA0183 DBT ("Depth below transducer") and DPT where the main data is "Depth below transducer" is the same as SK: "environment.depth.belowTransducer"
If you decide to send an offset from an instrument using DPT you're supposed to be aware of that. That's the logic I suppose. (No offset data are factory default from the instrument.)
Håkan

Yes,when you install your system you put in the distance to the surface and often to the keel also.
Opencpn before have always displayed depth+correction. with signalk it doesnt


Regards
Johan

[jlsail]

Didnt see your last reply.
Yes for me that would be a good solution!
Thanks
Johan

[SaltyPaws1]

Rpi4, openplotter v2, opencpn 5.2.4, nmea 0183 serial to signalK. Odd reading in dashboard plug-in. Rather odd true wind direction....

[Hakan]

SaltyPaws..
The details and what's wrong is not clear for me. What did you expect to read for the true wind angle?
- What NMEA wind messages was loaded to SignalK? Any example available?
- What are the resulting SignalK path? Check "Data Browser" for "self". See my example shot.
- And corresponding OCPN Dashboard presentation for the same data?

[SaltyPaws1]

Sorry I wasn't clear. The true wind angle should never be expressed as 371° - I believe that 11° would be acceptable. Looks like SignalK is getting wind data from NMEA sentence MWV.

[Hakan]

Salty...
OK, now I see. The shot wasn't enough visible for my bad? eyes. Sorry.

If you look at the graphical wind instrument the true wind angle from ship's bow is 123°. That's from the SK path: "environment.wind.angle.TrueWater".

The O Dashboard instrument "True wind direction" is reporting true wind with respect to north, also used in the wind history graph. That data in O Dashboard is from the SK path: "environment.wind.directionTrue" originating from the NMEA0183 MWD message.
Not very useful in a sail boat under way if you ask me. It's more suitable for a weather station or similar.
You better use the instrument called "True wind angle". That is with respect to ships bow.


Håkan

[Hakan]

Salty..
BTW: If neither the SK path: "environment.wind.directionTrue" nor NMEA MWD are present O Dashboard actually calculates the wind angle with respect to north if HDT is present. That calculation is there to be able to produce the wind history graph where we can admit some generalization for a graphical overview.

[Hakan]

Originally Posted by SaltyPaws1

Sorry I wasn't clear. The true wind angle should never be expressed as 371° - I believe that 11° would be acceptable.

You're of course correct. forgot that one. I agree it's a bug.

[Hakan]

SaltyPaws1..
Do you know if you have the SK path "environment.wind.directionTrue" present. The calculated value would not exseed 360°
Håkan

[Hakan]

SaltyPaws1..
Sorry for the delay.

The bug (+/- typo) is obvious. Your example reading wind angle with respect to north to 371° would have shown 349°. I'll PR a change for next OCPN release.
Thanks for the report.
Håkan