Anchor Alarm Radius?

[gbanker]

I have an anchor alarm app that I have not had the opportunity to try yet and am wondering what radius circle people have found works best to show a dragging vessel but not go off as the vessel swings at anchor. Thanks

[svmeanderer]

If you're going to swing 180 degrees:

alarm radius = R + 2G + 10 + T

R = rode distance put out
G = distance from GPS position to bow
T = tightness of alarm

The 10 is a good estimate for GPS accuracy if the GPS is outside and mounted. If you're using an internal GPS below deck, you might increase the 10 to 20 or 30.

T has a lot to do with how well you believe the anchor is set and how well it would reset if you swung 180 degrees. A T value of 20 is pretty tight. Experimenting with T take some time and experience.

The 2G is the killer if you have a GPS mounted on a stern rail. If you anchor in 8 feet of water with a 45 foot sailboat in that case at 5:1 scope, you'll put out (8 + 4) * 5 = 60 feet of rode. But that'll be swamped by the 90 feet of GPS error you'll experience with the GPS on the stern. It would feel quite strange with 60 feet of road out seeing the anchor alarm distance read 145 feet. But at 180 degree swing, you wouldn't even be at the end the swing radius yet and the anchor wouldn't have moved an inch.

[noelex 77]

Originally Posted by svmeanderer

If you're going to swing 180 degrees:

alarm radius = R + 2G + 10 + T

It is just G not 2G

R+G + 30 for feet or
R+G+10 for metres

Is a good starting point. The 30 feet or 10 meters is to allow for GPS error. For an SDFix you can reduce this slightly in very tight situations. There is no need, normally, to use trigonometry at reasonable scopes as the base distance (the actual distance from the anchor) is almost the hypotenuse (the rode length)

Most marine GPS units use nautical miles to measure these small distances ( or cables if you like). If you multiply by 200 this is an easy (rough) conversion to metres
0.02nm =40m (approximately)

Originally Posted by svmeanderer

T has a lot to do with how well you believe the anchor is set and how well it would reset if you swung 180 degrees.

Good anchors swing when they reset. If the anchor moves more than a foot or so during a 180 degree swing there is a problem and you want to be alerted.

[svmeanderer]

Originally Posted by noelex 77

It is just G not 2G

I'm sorry but it's not really something that's debatable - it's 2G and that's a fact. Draw it out on graph paper. Mark the position of the GPS and the boat's bow. Move back and count the squares. Then move the boat to the other side and count the squares. It's 2G.

What you're not considering is that the initial position captured by the alarm isn't the boat's bow/anchor position. That initial point capture is the position of the GPS.

Looked at another way, if you've ever captured your track while at anchor over a few days, it's always egg shaped if you can zoom in enough and have been pulled back to the end of the rode. It's an egg because of the extra G distance as you swing toward 180 degrees.

[Suijin]

He has an app. If he takes it to the bow when he first drops the anchor, and hits the set button, he'll simply have to set his alarm distance to R+G+T. That's what I do. Works great.

[noelex 77]

Originally Posted by svmeanderer

What you're not considering is that the initial position captured by the alarm isn't the boat's bow/anchor position. That initial point capture is the position of the GPS.

If you want an accurate GPS alarm it is essential to centre the anchor on the GPS aerial.
This is possible in many (but not all) marine GPS units. You centre the alarm radius over the drop point by using a remote waypoint as the centre.

I do this every time I anchor.

It is also easy to do on tablets and smartphones.

If you do this (as you should) otherwise you will have a very poor alarm.
G rather than 2G is correct.

Originally Posted by svmeanderer

Looked at another way, if you've ever captured your track while at anchor over a few days, it's always egg shaped if you can zoom in enough and have been pulled back to the end of the rode. It's an egg because of the extra G distance as you swing toward 180 degrees.

No it is a circle centred on the anchor.

[Nicholson58]

We use a Watchmate 850. You set the alarm as you drop the hook. In that way, it does not care which way you swing. Wait for the boat to settle at the scope you select. Read the distance and add to that to set the circle based on your comfort and other objects. The Watchmate sets a spec on the screen every 20 minutes or so. You can see where you have been since setting. You can re-set the alarm circle any time. The driver knows exactly where to drive to pick up the anchor.

[noelex 77]

Originally Posted by Nicholson58

We use a Watchmate 850. You set the alarm as you drop the hook. In that way, it does not care which way you swing.

If you want an accurate GPS alarm you cannot do this (unless the GPS aerial is at the bow) it will work fine if the wind direction does not change, but not for a changing wind direction.
You need to centre the alarm radius over the anchor, not over the the GPS aerial which may be 10m back from where the anchor is dropped.

[sabray]

The watchmate corrects for the position of the GPS.

[Cheechako]

I would just set it at the rode length.... that way if the wind shifts substantially, you can get up and see if the anchor is dragging.

[noelex 77]

Originally Posted by sabray

The watchmate corrects for the position of the GPS.

I have never used the watchmate so correct me if I am wrong, but does it have a fluxgate compass? If not the only way it knows where the bow is (to make the correction) is via COG which would work well for AIS, but would not work when anchored, or when stationary dropping the anchor.
Anyway happy to be corrected, but I am struggling to understand how it can make a correction when anchored unless it has an inbuilt fluxgate compass.

I am looking to add AIS shortly and the Watchmate is on the short list so it would be great if had this feature.

[sabray]

You are correct it needs the input from a heading sensor.

[noelex 77]

Originally Posted by sabray

You are correct it needs the input from a heading sensor.

Cool feature

[svmeanderer]

Originally Posted by noelex 77

If you want an accurate GPS alarm it is essential to centre the anchor on the GPS aerial.

...
If you do this (as you should) otherwise you will have a very poor alarm.
G rather than 2G is correct.

No it isn't. In that case, the distance from the bow to the GPS is zero. It's still 2G but in the case of G=0, 2G is the same as G.

And whoever said that the track picture would be a circle, that's right and I explained it very poorly. If you charted the distance that the anchor alarm shows you to be back from the anchor, it would be an ellipse. The typical chartplotter track will be a circle. It's just that along that circle, the anchor alarm can be showing vastly different values even when always pulled back to the limit of the rode. This is quite easy to prove by drawing it.

Vesper's WatchMate does do it correctly - that's true. They're able to take out most of the error because they compensate to eliminate G. And as above, 2G = G = 0 in that case.

[noelex 77]

Here is a diagram showing if the anchor alarm is centred on the anchor (as it should be) the swing is a circle the radius is:

R+G

There needs to be a factor added for GPS error so the final formula is

R+G+ GPS error (say 10 for m)