Satellite Compasses?

[Dockhead]

When I did my new electronics two years ago, there were two types of data I really wanted to improve -- speed through the water, and heading.

I spent a fair amount of money on what I thought were state of the art sensors for these -- an Airmar CS4500 ultrasonic speed transducer, and an Airmar H2183 three-axis gyro-stabilized heading sensor.

Both are improvements on what went before, but not nearly as good as I had hoped.

The heading sensor, in particular, suffers from one big problem -- it is extremely sensitive to metal objects even several meters away, and it can be really hard to remove those influences.

Pilot performance is great, since the heading data is at least very consistent -- no complaints there -- but distortions of heading data ruin radar overlay (which I use a lot), ruin the Zeus' nifty calculation of tide set and drift, and ruin the quality of bearings taken by radar. I really want to improve this.

So I am looking again at satellite compasses.

Simrad makes one -- the HS70.

There are also two from ComNav -- Vector G1 and G2.

And Furuno make the SC-30.

All of these seem to be somewhat old models, and do not seem to receive GLONASS, Galileo, or EGNOS, which seems to me a huge disadvantage.

And of course they are pretty expensive (around $2000), but I would be willing to pay that much for perfect heading data.

I know that our own Dave (Goboatingnow) has actually built satellite compasses, so I hope he'll weigh in. But I'm interested in anyone's experience and knowledge of such systems.

[seandepagnier]

it is possible to get heading from two gps sensors if they are spaced far enough apart, but the accuracy is questionable unless you have a very specialized receiver which reads from both inputs and is designed to determine attitude from gps. In this case it could be highly accurate.

A good digital compass will give heading accuracy at least as good as a floating compass, so if metal affects it, there isnt much you can do. If the floating compass is ok but the digital one is bad, then they screwed up their software and didn't properly determine the 30 (or possibly more) calibration coefficients needed to calibrate acceleromters, gyroscopes and magnetometer sensors. It's a difficult enough problem that it's likely they just didn't figure the math out well enough to deal with changing magnetic distortions and called it good when it worked without any metal.

Finally, it is possible to use other sensors to determine attitude, for example if you can determine side slip through the water, then you could figure out your heading just from that and the gps.

[noelex 77]

The Seapilot Vector Compass looks like a promising new addition. It is cheaper than the alternatives at $999

Sail-World.com : Vector Compact, the smallest GPS compass in the world

There is a review by Panbo here:
http://www.panbo.com/archives/2014/1...o_the_usa.html

The Trimble Bx982 is another option that may be adaptable to marine use:
http://www.trimble.com/gnss-inertial...aspx?dtID=feat

[Garrettw]

Originally Posted by Dockhead

Pilot performance is great, since the heading data is at least very consistent -- no complaints there -- but distortions of heading data ruin radar overlay (which I use a lot), ruin the Zeus' nifty calculation of tide set and drift, and ruin the quality of bearings taken by radar. I really want to improve this.

A satellite compass is the cream of the crop and I've been trying to talk myself into "needing" one as well. One thought on your existing gyrocompass. Does your system allow you to enter a correcting offset? I think that, on my Raymarine setup, I can enter a manual correction if I wanted.

[Ziggy]

Originally Posted by Dockhead

The heading sensor, in particular, suffers from one big problem -- it is extremely sensitive to metal objects even several meters away, and it can be really hard to remove those influences.

Did you go through the autocompensate procedure for your compass? Turn through ~3 slow complete circles as described in the installation manual? Unless you have a lot of ferromagnetic material that is not fixed (loose cannon on deck?), you should be able to compensate well enough to get better than ~2 deg accuracy. Since a typical small boat radar beam is about 5 deg wide, that should be more than sufficient for radar display alignment.

[Dockhead]

Originally Posted by Garrettw

A satellite compass is the cream of the crop and I've been trying to talk myself into "needing" one as well. One thought on your existing gyrocompass. Does your system allow you to enter a correcting offset? I think that, on my Raymarine setup, I can enter a manual correction if I wanted.

Yes, naturally, I can enter an offset, but the problem is that I can't seem to control the proximity of metal objects. The compass is mounted at the roll and pitch center of the yacht, which is the ideal spot, just behind the mast, under the salon settee. Once I had an electric drill in the next cabin (!), probably 3 meters away, and it threw off the compass by 10 degrees. I'm tired of it.

Oddly, the old fluxgate didn't seem to be so sensitive. ALthough maybe it was just so inaccurate that I didn't notice it.

[Dockhead]

Originally Posted by Ziggy

Did you go through the autocompensate procedure for your compass? Turn through ~3 slow complete circles as described in the installation manual? Unless you have a lot of ferromagnetic material that is not fixed (loose cannon on deck?), you should be able to compensate well enough to get better than ~2 deg accuracy. Since a typical small boat radar beam is about 5 deg wide, that should be more than sufficient for radar display alignment.

Yes, I've done the procedure. The "loose cannon" is metal stuff that people bring in and out of the salon and other cabins. 2 degrees is a bit crude. The unit is rated for better than 1 degree RMS, and you can really feel the difference in those degrees of accuracy. The satellite ones are good to a half degree or so, and obviously are impervious to influence of metal.

[Dockhead]

Originally Posted by noelex 77

The Seapilot Vector Compass looks like a promising new addition. It is cheaper than the alternatives at $999

Sail-World.com : Vector Compact, the smallest GPS compass in the world

There is a review by Panbo here:
Panbo: The Marine Electronics Hub: Seapilot Vector Compact GPS Compass & True Heading comes to the USA

The Trimble Bx982 is another option that may be adaptable to marine use:
Trimble - Precision GNSS + Inertial - BX982

Thanks, hot tips!

The Vector Compact looks very attractive in every way except accuracy, which is less than my existing compass. Maybe it doesn't matter with the stability and freedom from interference of this type of compass.

The Trimble, at $4000, does not actually seem expensive for what it is! Centimeter position accuracy!! Heading accuracy better than 0.05 degrees! Wow!! 50 hz output!!!

I would be just half tempted to buy something like that, just for fun. It would be some trouble to integrate with an N2K system, but maybe not too bad. It can output NMEA0183 over RS232, so I guess there's some Actisense box or another you could hook it up to.

You also have to buy and install separate antennae for it, so it's far from plug and play. Very intriguing, however.

[David M]

How you calibrate your fluxgate compass for deviation (making an electronic internal deviation table) makes a difference. You want a very large true circle over the ground using an electronic chart to do this, a nice smooth curve and to end up at the exact place you started, usually using a day marker as your exact point to start from and to return to. The electronic deviation table generated removes all the influence of iron on the boat.

There is no need to spend tons of money on other types of compasses if you have a well made fluxgate. I have a Furuno fluxgate that has been very reliable and accurate. It is about 3 meters away from two very large diesel engines.

Product Detail - PG500R

The survey grade GPS's rely on being able to receive a radio signal from a shore based GPS that is transmitting the current position error. Or the position error can be determined by post processing, which does not help you in real time. I have a survey grade Trimble GPS onboard for doing science. The survey grade GPS has to be put ashore 24 hours ahead of time so by averaging it gets an exact position. In other words, this system is worthless to recreational boaters.

The GPS heading compasses work by placing two GPS receivers onboard and measuring the difference between the positions. Processing then determines the boats heading.

[SailRedemption]

Why not get the air more heading unit with the built-in GPS? I think with it being above deck and with GPS it would have a lesser chance of having anything alter heading.

I looked at the specifications on the GH2183 brochure and it says "perfect for metal hulled boats" so maybe it's not as sensitive to metal or its GPS compensated. Worth a look..

Also, could you just put it, say in the overhead so there's no chance of anyone putting anything near it? I don't think the 5-6ft would affect pitch and roll that much, especially if your getting better readings.

- Ronnie...on the geaux

[Dockhead]

Originally Posted by SailRedemption

Why not get the air more heading unit with the built-in GPS? I think with it being above deck and with GPS it would have a lesser chance of having anything alter heading.

I looked at the specifications on the GH2183 brochure and it says "perfect for metal hulled boats" so maybe it's not as sensitive to metal or its GPS compensated. Worth a look..

Also, could you just put it, say in the overhead so there's no chance of anyone putting anything near it? I don't think the 5-6ft would affect pitch and roll that much, especially if your getting better readings.

- Ronnie...on the geaux

Thanks. The GH2183 is exactly like my H2183 except it also has a GPS receiver. The GPS receiver is not used for heading data, which is produced by a normal fluxgate compass with gyro stabilization. But it's true that a different position might help. With the stabilization, this one might not be as sensitive to boat motion as a plain dumb fluxgate.

[Dockhead]

Originally Posted by David M

How you calibrate your fluxgate compass for deviation (making an electronic internal deviation table) makes a difference. You want a very large true circle over the ground, a nice smooth curve and to end up at the exact place you started. You can remove all the influence of iron on the boat. No need to spend tons of money on other types of compasses if you have a well made fluxgate. I have a Furuno fluxgate that has been very reliable and accurate. It is about 3 meters away from two very large diesel engines.

Product Detail - PG500R

OK, that's a good tip. Maybe I'll try the calibration again.

The H2183 was supposed to be the best fluxgate money can buy, and it was not cheap, nearly as expensive as the Vector Compact satellite compass.

[David M]

Yes, give your fluxgate another try at generating an internal deviation table. Follow the tips that I described. Good luck with it.

When you are done with the calibration for deviation, range off a few things out on the water to see how accurate the deviation table is.

[Dockhead]

I'm rather fascinated with this Trimble BX982. I think it's a pipe dream, but still a compelling one.

It outputs NMEA 0183 sentences, half of which I can't find anywhere. Over RS422.

But RS422 per se doesn't seem to be a problem -- that seems to be electrically compatible with NMEA 0183, so you just hook it up to, say, an Actisense NGW-1 and voila.

Unfortunately it doesn't seem that 0183 has any sentences for roll & pitch (I think the N2K PGN is "attitude"), which is a shame, because that's data you will want to get out of this unit.

How cool would it be to have centimeter position accuracy and 0.05 degree heading accuracy. This unit receives every kind of constellation, and not just GPS. This is a total different level of accuracy from the regular satellite compasses and at only twice the price.

A bit expensive -- just a bit, but still tempting.

[noelex 77]

Originally Posted by Dockhead

Unfortunately it doesn't seem that 0183 has any sentences for roll & pitch (I think the N2K PGN is "attitude"), which is a shame, because that's data you will want to get out of this unit.

I cannot see how a satellite compass with just two GPS units could output Roll information.

The SC30 has a 3-axis gyro in addition to GPS units. This would have to be used for the roll information as there is no way to determine this from the GPS signal (if there are 2 GPS units mounted on the longitudinal axis). I would also expect the information from the gyro is also used for pitch since a GPS unit's estimation of elevation is much worse than its horizontal accuracy.

The Trimble would provide exceptional heading and position information, but I don't think it has a gyro so it could not provide roll information.