OpenCPN Radar Overlay PlugIn

[bcn]

Hakan,

if VHW describes the vector of velocity through water then this "heading" is not identical to vessel heading. Delta of Yaw angle... Or difference between track and heading.

Hubert

[coupdemistral]

I have cloned the gradar_pi repo and modified the code to use VHW, and OpenCPN now shows perfectly overlapped radar images and charts. Yay!

The magnetic value I get from VHW matches the readings from my binnacle compass almost to a degree. So far I have only tested it on a moored boat. While SOG and COG oscillate randomly, as they always do when a boat is still and the GPS fixes oscillate between points close to the actual position, the VHW readings are steady and accurate, with heading values only changing +- 1 degree (you can see the readings in real time from this public SignalK server, in the navigation.headingMagnetic widget ).

I will repeat the tests while under way, but at this point I am pretty sure that the VHW sentence I get from Raymarine MFD contains the readings from the paddle-wheel speed sensor and from the autopilot compass (which is what I need for the radar).

I will submit a pull request to the gradar_pi github repo, in case this is deemed to be useful to others. I would assume that many folks with the same make/model of navigation system would find this useful.

[Hakan]

Hubert..
I do agree to ".... then this "heading" is not identical to vessel heading" as a general rule. But, have you seen it used as some kind of "leeward angle" or anything else?
And coupdem..., your solution is great for you knowing you've a correct heading in your VHW. I also have some own OCPN adaptions to fit my dated Simrad instruments. But, MHO, for a general implementation it has to be carefully deliberated.
Håkan

[bcn]

Håkan,

I agree that most probably the value shown will be the compass heading. I just couldn't resist to point out the difference - just in case .
Coupdemistral will tell us once he is on the water, I hope.

Hubert

[coupdemistral]

Originally Posted by bcn

Håkan,

Coupdemistral will tell us once he is on the water, I hope.

Hubert

Will do.

BTW - I understand the conceptual difference between the velocity over water vector and the compass and speedometer readings. I just can't think of any sensor (or algorithm) that would enable us to measure (or calculate) the effects of leeward motion when under way, so I am confident that the readings I get are from compass and speedometer.

[bcn]

Originally Posted by coupdemistral

Will do.

I just can't think of any sensor (or algorithm) that would enable us to measure (or calculate) the effects of leeward motion when under way, so I am confident that the readings I get are from compass and speedometer.

INS - Inertial Measurement Sensors - do this. But few of us have those on board.
There exist simple ones for the RasPi, so somebody with OpenPlotter might want to play with those. Problem can be precision here.

Hubert

[petter5]

The Navio2 raspberry pi hat form Emlid has dual IMU's and are pretty accurate - and works fine with openplotter. https://emlid.com/

Originally Posted by bcn

INS - Inertial Measurement Sensors - do this. But few of us have those on board.
There exist simple ones for the RasPi, so somebody with OpenPlotter might want to play with those. Problem can be precision here.

Hubert

[MantaKoala]

Hi. I got a Garmin 18 xHD Random and I'm running opencpn on a Mac Book air with OS 10.11.6 (El Capitan). I soldered a switch at the amphenol (pin 5 to ground as described in a former post), but the Radar does not run. What do I have to do?

[coupdemistral]

is anyone experiencing issues with CPU usage constantly close to 100% when using the gradar_plugin? It seems to steadily grow from 10-15% to 100% within the first couple of minutes of use. Then the rendering of radar scan on the chart slows down and Opencpn becomes sluggish.

Some context:

- OpenCPN 4.6.1, gradar_pi v 1.5
- Ubuntu 16.04 on a Dell latitude with Nvidia NVS 3100 card, latest NVIDIA driver installed
- OpenGPL activated
- nothing looking suspicious in opencpn.log

Changing the NMEA feed from AIS from UDP to serial seems to have improved things a bit, but the system is still so slow to be unusable.

(PS this is redundant with another post I submitted. Sorry for the duplication)

[jrochest]

Does the plugin work with the Garmin Fantom 24? and is it gradar or gxradar?

I can't seem to get it to work.

[merrimac]

Originally Posted by jrochest

Does the plugin work with the Garmin Fantom 24? and is it gradar or gxradar?

I can't seem to get it to work.

I think you are the first to try. It would surprise me if they work the same way as Fantom is a significant upgrade for Garmin.

[rgleason]

Not to hyjack the thread but,
How does Sailoog Shop IMU sensor MPU-9255 $10 compare to
B&G Precision 9 Compass $650 or the all-in-one GPS + IMU Airmar GH2183 $800

in terms of accuracy, use with Radar Overlay, Arpa, Marpa, measurement of heeling angle and as bcn suggests leeway?

Originally Posted by bcn

INS - Inertial Measurement Sensors - do this. But few of us have those on board.
There exist simple ones for the RasPi, so somebody with OpenPlotter might want to play with those. Problem can be precision here.

Hubert

[Andreas29]

Hi,

tried somebody the radar plugin (with radar) on a raspberry?

The plugin will run in emulator mode but what is about the real use (with a radar dome)?

Regards

Andreas

[Hakan]

Originally Posted by Andreas29

Hi,
tried somebody the radar plugin (with radar) on a raspberry?
The plugin will run in emulator mode but what is about the real use (with a radar dome)?
Regards
Andreas

Andreas... Late answer, sorry.
Yes it's running as chart overlay on R_pi 2B with O 4.8 and BR24radar_pi ver 3.0.
All NMEA comes via wlan and the radar traffic via wire to eth0.
Also (M)ARPA works fine! The radar window can be slow while opening but works also fine.
I use Rasbian out of the box and OCPN is locally built.

Håkan

[Andreas29]

Hi,

Hakan, thanks for your post. Glad to hear that the pi can manage this

Do you have a gyro connected to the radar or to your network?

Regards

Andreas