Inexpensive heading sensor?

[nohal]

Eric...
Not good for radar - the NMEA feed is too slow (and inaccurate). A high speed heading sensor is usually used in such a set up.

Pavel

[SailFastTri]

Originally Posted by Turbofly

Is anyone having any luck with the cheaper heading sensors? If so what models is everyone using?

Also I couldn't find much direction on how they connect and are configured in opencpn. Am I missing something in the manual? I plan on connecting it directly to the pc's serial port. Is it really as easy as power and ground and connecting the db9 pin 2 and 3 to the nema?

You might already have it. Do you have an autopilot? If yes, what brand/model, and does it have a course computer with NMEA-out?

[Dockhead]

If you use heading data for autopilot and/or radar, it is best not to use "cheap" ones. The quality of the data is dramatically different with different types. The best one on the market for less than $1000 is the Airmar H1283 -- a beauty, with three-axis stabilization.

If you just want electronic compass data just to know which way you're pointed, then any fluxgate will do. They're cheap and, yes, power, ground, data, and Bob's your uncle. Get a used Raymarine fluxgate off Fleabay for probably $50 or less.

[SailFastTri]

Originally Posted by Dockhead

If you use heading data for autopilot and/or radar, it is best not to use "cheap" ones. The quality of the data is dramatically different with different types. The best one on the market for less than $1000 is the Airmar H1283 -- a beauty, with three-axis stabilization.

If you just want electronic compass data just to know which way you're pointed, then any fluxgate will do. They're cheap and, yes, power, ground, data, and Bob's your uncle. Get a used Raymarine fluxgate off Fleabay for probably $50 or less.

Won't a Raymarine fluxgate output SeaTalk? If he has a Raymarine course computer already, chances are it has an NMEA-out port that will put out "fast heading" info on that port (using the info from the fluxgate already installed for the AP) and if the course computer has a gyro that info will be stabilized.

[rebel heart]

Originally Posted by nohal

Eric...
Not good for radar - the NMEA feed is too slow (and inaccurate). A high speed heading sensor is usually used in such a set up.

Pavel

Ah, interesting. This is for course correction or some real-time type stuff, where a near constant ~5 second delay would be pretty problematic.

That's what I get for having all my old school stuff onboard.

[nohal]

Originally Posted by rebel heart

Ah, interesting. This is for course correction or some real-time type stuff, where a near constant ~5 second delay would be pretty problematic.

That's what I get for having all my old school stuff onboard.

Yeah, he is talking about stabilizing the radar overlay - the radome being fixed to the boat, you need a fast and accurate source of the heading, not course over ground, to be able to sync the chart rotation with it reasonably...

Pavel

[bcn]

Originally Posted by nohal

you need a fast and accurate source of the heading, not course over ground, to be able to sync the chart rotation with it reasonably...

Pavel

Important: heading, not COG - and this data stable and compensated from pitch, roll, yaw and heaving if you want something perfect. GPS - at least the common sensors - can't do this. Inertial Navigation Systems do this. And they will give you both, HDG(T and M) plus COG. But prepare some "boat units"..

Hubert

[rontom]

I use a dedicated Desktop for Navigation, I have used it for Years with no problems, It has its own closet with lots of Ventilation. I like the G radar, with OpenCpn,i incorporated it with the Garmin 18 HD. The problem being that at low speed 3 Knots or less, Lots of Gyration, unstable target definition. At Higher speed 5 to 7 knots, works fine, the targets stay put.

I have a Wagner Mk4 Auto Pilot with Standard Magnetic compass and no Gyrocompass so that is out of the Question.

I thought it would be interesting to build one, but i will probably just go ahead and buy a good one.

Thanks everyone.

[nohal]

The cheapo's with soldering irons between us would probably start at Selfmade Tilt compensated Compass Pitch, Roll, Yaw - Arduino Forum

Pavel

[svlamorocha]

Originally Posted by rebel heart

Doesn't a normal GPS with nmea have the cog?

OP is concerned about heading, which is not the same thing as COG. If you want heading from GPS data (uncommon) you need two GPS antennas and probably two GPS receivers. If you want COG from GPS data (very common) over some time your GPS will just calculate arc tangent of (delta position / delta time)

C

[bcn]

Two simple GPS will not do the trick:

for random noise of a GPS count with 2m error. Using a very optimistic 1m error you will get:
distance between GPS 20m, error 2* 1m --> almost 6 degrees jumping/noise
With distance of 10m (33ft) we would have to count with 11 degrees.

Tell this to your autopilot

Hubert

[cagney]

Originally Posted by bcn

Two simple GPS will not do the trick:

for random noise of a GPS count with 2m error. Using a very optimistic 1m error you will get:
distance between GPS 20m, error 2* 1m --> almost 6 degrees jumping/noise
With distance of 10m (33ft) we would have to count with 11 degrees.

Tell this to your autopilot

Hubert

...unless they are exposed to the same "random noise"..
Check GPS Compasses, for example. A quote

If you have two GPS antennas onboard one boat, you can calculate your heading by comparing the position of one with the other. The fact that they are in more or less the same place and working at exactly the same time means that most of the errors that affect GPS as a positioning system become irrelevant; it doesn’t matter, for instance, that a satellite may be a couple of feet out of position or that its signals have been distorted by the earth’s atmosphere because the effect of these errors is exactly the same at both antennas.

Tim Bartlett

Thomas

[bcn]

Originally Posted by cagney

...unless they are exposed to the same "random noise"..
Check GPS Compasses, for example. A quote


Thomas

This is correct for atmospheric influences and the reason why the absolute error is changing slowly over time. This one is aprox. 1.8m for a receiver without differential corrections.
SBAS/WAAS/EGNOS can reduce this to 0.9m depending on the GPS chipset used.

The remaining error of each GPS unit is independent from each other. There the 1m noise is coming into effect.

If you have a base station with L1/L2 receivers for differential corrections and you are using GPS receiver with topograhical performance (not only doppler for position but carrier-phase too - this is what we don't have in the commercial receivers we use generally - L1 Doppler) you will be able to go down to 1 or 2cm precision. More difficult for systems that are moving.

Bandwidth is the next problem with all "GPS-only" solutions. 15hz is about the best you can get.

The article cited mentions the combined solutions (accelerometers, gyros plus GPS) . Also known as INS - Inertial Navigation Systems.


Good inexpensive solutions are difficult. The idea Pavel was pointing to might work.

Hubert

[dacust]

Originally Posted by bcn

...

The article cited mentions the combined solutions (accelerometers, gyros plus GPS) . Also known as INS - Inertial Navigation Systems.

...

Just to clarify, inertial navigation systems would not include a GPS. Only gyros and accelerometers.

-dan

[bcn]

Some modern do combine both - to get the best from the two worlds:
- high precision and bandwidth from gyros and accelerometers
- no drift and precise timing from GPS

for example:
| Coda Products

non marine aplications (from the manufacturer of the Coda stuff):
OxTS

Hubert