Deviation check

[pugwash]

Bought a 30-year-old wooden boat a couple of years ago and on gentle shakedown cruises close to home noticed by simple checks (transits by eye) etc that there didn't seem to be much in way of deviation. (Yes, I KNOW I should have checked properly but there you go, I didn't get round to it).

When I picked up a hand-held GPS to carry as an auxiliary to the main system I used it to check the compass. I steered a close course for two or three minutes to give the satellites time to see where I was tracking, then checked the compass against the GPS; it was slack water with no tide or current discernible. I boxed the compass in 15-deg increments and found at most two degrees of deviation, alll under power, so that leaves possible differences when heeling. For practical purposes the deviation was small enough to be disregarded. Great.

The other day, however, I found an old deviation graph for the boat made by a previous owner and the figures were substantially higher, about ten degrees on some courses. A new engine has been fitted since then but of course it's in the same place.

I can check again in the recomended way using a hand-bearing compass away from the engine and also lining up on transits, but I'd like to know whether my check was in fact water-tight. Did I make a basic error?

 

[tome]

Using GPS for bearing checks isn't a great idea as bearings are based on cmg over the ground. Also there are quite a few ways of filtering cmg in GPS receivers. You'd have to hold a rock steady course with no tide or wind to obtain useful results

Was the GPS set to true or magnetic headings? Most GPS receivers can be set to use either and use an internal model to apply variation based on position and date for mag bearings.

If it was set to true, did you apply variation?

 

[pugwash]

No tide, no wind, so any kind of lateral drift was unlikely. GPS was set to Magnetic. I thought my little trial, though imperfect, would be good enough on a "rule of thumb" basis but the difference with the old deviation card I mentioned was startling.

 

[muchy_]

I tried using my GPS to check for deviation but found that the GPS fluctuates to much to be of any use. The transits are probably your best bet or even better still is a fluxgate compass which is not affected by deviation, thats what I'm going for.

 

[bedouin]

Deviation is caused by external magnetic influences on the compass. As such fluxgate compasses are just as prone to it as conventional magnetic compasses. Some fluxgate compasses can auto-correct the deviation - but that process is rather simplistic

 

[paulineb]

Colin Scott is an expert on deviation

Pxx

 

[ccscott49]

You've been and gone and done it again! Just because I said I was a deviatore! er! Anyway, I do know a thing or two about deviation, which particular deviation were you interested in? (we call it declination in downhole surveying circles)

 

[david_e]

Without knowing how the previous owner did his checks the card might be useless, 10 degrees sounds alot. I would check his extremes against your own findings again just to verify that you are correct. Additionally, much can change over time so I would stick with your own findings until you have evidence to make you believe differently.

 

[ccscott49]

I agree. Until you can either swing the compass yourself or have it done for you, I would go with your own observations, you have done enough already to prove the old deviation card wrong. I would suggest you periodically check against your handheld, when the oppurtunity presents itself, (calm day, no tide, excellent transit etc.)

 

[hillyarder]

This is probably a silly thing to say. But is it the same compass as was checked before. Hasnt been a newer one since. Its the kind of mistake i would make, but then again i am rather half soaked.

 

[ccscott49]

Now that is a bluidy good point!

 

[pugwash]

Hail to the Deviatore

Thanks to the deviatore and all you other Scuttling deviants! Yes, it was the same compass as before, a lovely Sestrel brass job that wouldn't know a flux if it came through the gate. The main drift of my question, perhaps not clearly enough put, was to determine how trustworthy a GPS is for checking (magnetic) deviation...er... declination. You have given me some useful pointers. I'll do a swing properly in the spring and see how it turns out.

 

[sailbadthesinner]

best way is to follow a known transit course so you know your true
you can then take a note of your compass course
you know your variation so you can calc;
you know your true
+/- variation
gives magnetic
diff between magentic and compass is your deviaiton.


has to be done in flat water with no wind and no tide

can be used as a back up with gps.


Come on brain.get this over and i can go back to killing you with beer

 

[peterb]

Sinbad is right, of course. The best method is to follow a known transit under calm conditions with no tidal stream. At least, he's right in theory. The trouble is that even a very small stream will affect the result. A stream of 0.1 kn is almost imperceptible (about 2 inches per second) but if you are doing 6 knots down your transit then a 0.1 knot cross tide will give you a one degree error.

What's more, sailing down a transit is unlikely to give you an accurately readable heading. Most small craft compasses are only calibrated at five degree intervals, and are difficult to read to better than one degree unless the heading exactly lines up with one of the calibrations.

We've found that on a wood or GRP yacht comparison with a carefully calibrated hand bearing compass gives the best results.

 

[kdf]

On a similiar vein - I'm about to install a cockpit VHF speaker and when I place it in position it pulls the compass out by about 10 degrees. I can't pick any other place to install the speaker so my question is if I do install it will a good compass adjuster be able to pull back the 10 degrees.

thanks

 

[sailbadthesinner]

I am always right in THEORY
in THEORY i know how to get a boat off a pontoon surrounded by other boats with an awkward wind and v little space
in practice i surround the thing with fenders and people and pray that my theory stands up.

Come on brain.get this over and i can go back to killing you with beer

 

[vyv_cox]

GPS compass checks

It depends on the GPS. My old monochrome Garmin GPSMAP and even older Garmin 75 had a function that allowed changes in the averaging period to cater for speed differences between say, walking and flying an aeroplane. Although the instructions recommended against it, it was possible to adjust this period to give a sensible average heading figure when sailing, although the actual heading might be varying a lot. But this would not be much use for checking the compass.

When I bought my colour GPSMAP 220 I found that this feature was no longer provided. I spoke to Garmin to verify this. Now the heading updates every 2 seconds or so, which on a broad reach in a bit of sea gives useless information. So on a quiet day the indicated figure would be pretty accurate and could presumably be used to verify the compass heading.

There's an article on the subject of compass checking in last month's YM (or was it PBO?) They recommend the more usual method, verify your sighting compass against landmarks, then check the ship's compass against that.