gps heading

[dirkvdheuvel]

Hello,
Is it possible to calculate heading from 2 different gps positions in opencpn?
It is for an inland barge and I know a navigation program that can do it, but that is really expensive. The 2 gps antennas are placed 80 meters from each other, so I think it will be accurate. But in opencpn I can only add 1 gps antenna.

[nohal]

You can add as many GPS data sources as you like and OpenCPN will receive data from all of them. But that data will only be used for position and COG.
It is not a complicated trigonometric exercise to calculate the heading from two points, but unless those GPS receivers are high quality and high rate, the results won't be anything spectacularly accurate even with them being 80 meters apart...

[hoolie]

That's true. Multipath errors are a poorly recognised problem with GPS and can be significantly different for two receivers 80 m apart particularly on a moving vehicle.

[dirkvdheuvel]

Originally Posted by nohal

You can add as many GPS data sources as you like and OpenCPN will receive data from all of them. But that data will only be used for position and COG.
It is not a complicated trigonometric exercise to calculate the heading from two points, but unless those GPS receivers are high quality and high rate, the results won't be anything spectacularly accurate even with them being 80 meters apart...

Thank you, is there a way that I can make that calculation in OpenCPN?

[nohal]

Sure, it is open source, so you can modify it in whatever way you want. Or you can implement it in the Javascript plugin.

[champ8242]

Dirk,

With 2 relative cheap GNSS sensors (u-blox F9) you can get easily up to 3-6 cm accuracy and convergence within seconds.

Bart

[nohal]

Yes, the F9 ticks the boxes for this application as it is DGNSS and has up to 25Hz update rate...
Not exatly a GPS puck for 10 bucks, but they would sure generate pretty accurate results for a few hundred, much cheaper than a GNSS compass.

[champ8242]

Originally Posted by nohal

Yes, the F9 ticks the boxes for this application as it is DGNSS and has up to 25Hz update rate...
Not exatly a GPS puck for 10 bucks, but they would sure generate pretty accurate results for a few hundred, much cheaper than a GNSS compass.

If I read:

Originally Posted by dirkvdheuvel

The 2 gps antennas are placed 80 meters from each other.

I don't think they should be using 10 bucks GPS pucks

[nohal]

Yes, but they might be using "marine GPS" with 1Hz update rate, single constellation and not unusual error in the area of 10 meters...

[ian5142]

I did get GPS heading out of a couple of Mikroe GNSS 7 Click boards (uses Ublox M9N) and an Arduino. Creating a 10Hz HDG output. I also did it with a tilt compensated compass board (CMPS14), it was actually more stable then the GPS based heading, also hooked up to an Arduino to create the actual HDG sentence. The Arduino also does some smoothing so that the heading doesn't jump all over the place. I would love to get a couple of Mikroe RTK Click or RTK 2 Click boards but they are a little over my budget for the project.



If you want to see my code, just message me. It should run on almost any arduino.

[Dockhead]

Originally Posted by dirkvdheuvel

Hello,
Is it possible to calculate heading from 2 different gps positions in opencpn?
It is for an inland barge and I know a navigation program that can do it, but that is really expensive. The 2 gps antennas are placed 80 meters from each other, so I think it will be accurate. But in opencpn I can only add 1 gps antenna.

Why don't you just buy a GPS compass? A lot cheaper than the programming time to solve this that way, if you put any value on your time, and much more accurate.


80 meters separation will not give you as much accuracy as you think. The math is a little above my pay grade, but what's the position accuracy of the GPS antennae? At 60 meters (easier to figure), an additive error of 1 meters would be 1 degree, right? So to get within one degree, you would need accuracy of both devices in the centimeters range.


GPS compasses work on a different principle which yields much better results. It was do do with the phase differential between the antennae, and you only need a few centimeters of separation to get accuracy of fractions of a degree.

[ian5142]

Furuno SCX20 and SCX21 are decent GPS compasses but they are ~$1400.


I don't know of any that are cheaper than that.

[Fintry]

The Simrad HS60 and the Garmin MSC-10 are both around $1100.


Jim

[Ben S]

I use a Raymarine AR200 for GPS and heading.
About € 600

[Dockhead]

Originally Posted by ian5142

Furuno SCX20 and SCX21 are decent GPS compasses but they are ~$1400.

I don't know of any that are cheaper than that.

That's what I have.