Deviation Table from GPS

[barnakiel]

Originally Posted by osirissail

(...) there are a few marine surveyors who will cite you for not having a current deviation table if you have a magnetic compass. So Yes, you do need to make one(...)

Wrong.

I believe that IF a surveyor cites you for not having one THEN they will not accept a home-brew, GPS-generated one. You WILL have to hire a duly certified pro to have one made.

I derive it from the simple fact that if something is required in an X country then sure like hell the only way to get it is by buying it from a certified source / supplier, etc..

You cannot make your life jackets. You cannot self-service your life raft. What makes you think you can make your own table?

;-)))

Cheers,
b.

[estarzinger]

Originally Posted by unbusted67

So why is a GPS's COG not an accurate way of measuring a true coarse?

Just FYI . . . Typical yachting GPS accuracy - from a furuno spec sheet:

Accuracy

GPS: 10 m (95%)
DGPS: 5 m (95%)
WAAS: 3 m (95%), limited coverage
SOG: ±0.2 kt (SOG ≦ 10 kt)

COG: ±3° (SOG 1-17 kt), ±1° (SOG > 17 kt)

Source: http://www.furunousa.com/ProductDocu...2011-17-05.pdf

Originally Posted by Ziggy

A GPS receiver determines COG by differencing two positions taken a short time apart (usually a few minutes). Let's say the averaging interval is 6 minutes. On a sailboat moving at 5 knots the position changes 0.5 nautical miles (926 meters) every 6 minutes. Given that GPS WAAS position accuracy is about 10 meters, the COG error based on two position measurements taken 6 minutes apart may be as much as 1.2 degrees (arcsin(20/926)). Typically, the error will be less, however.

Good explanation.

See above, Furuno says their COG accuracy spec is ±3°. This is higher than your 1.2 degrees because the IMO COG spec calls for 30 second averaging. With the 6 minute averaging you are proposing the gps readout response too slowly (at least for the IMO) to course changes.

But if you can turn the gps averaging up to 6 minutes, and you can hold a steady magnetic compass course for that long, and have a track the long without any significant current or wind, then I agree its possible to achieve your 1.2 degree accuracy. That's still marginal accuracy if you are trying to create a proper deviation table - deviation errors on a (fiberglass) cruising boat are often in the 2-3 degree range, so your gps COG source is creating 30-50% noise.

Again, it just depends on what the OP is trying to accomplish. If he is trying to create a professional quality deviation table then using the COG is not the way to do it. But if he is just trying to check that his compass is not way way out of whack, then the COG will do the job. Since the OP seems to have disappeared we may never know

[David M]

Just do it the old fashioned way. It's easy and one more navigation merit badge that you can add to your sash that the purely GPS cub scouts will never have the chance to earn.

[osirissail]

Originally Posted by barnakiel

Wrong.
I believe that IF a surveyor cites you for not having one THEN they will not accept a home-brew, GPS-generated one. You WILL have to hire a duly certified pro to have one made.
I derive it from the simple fact that if something is required in an X country then sure like hell the only way to get it is by buying it from a certified source / supplier, etc.. . . .
Cheers, b.

Wrong, right back at yah - First hand experience - All the marine surveyor in Florida wanted to see is the little card with numbers on it showing the amount of deviation at each cardinal. The existence of the card is all he was concerned with. How you got that card is of no interest to the surveyor. All he wants to do is put a tick mark on his checklist that says a current deviation card exists. Other countries of the world and even other political regions may have different requirements - but I doubt it.

[Pblais]

the reason I said they are more accurate is that ( a) they are decoded to a higher resolution but more importantly they are capable of easily having their deviation Offsets calulated and such offsets are often computed for 0. 5 degree changes sometimes even more.

They do compute their own deviation as part of the calibration but they tend to slosh a lot more than a good old fashioned mineral oil dampened compass as far as a pure display. You also can't tell when a fluxgate is corrupted nor can it compute the declination. The GPS will tell where you are going (more like where you have been) but not where you are pointed. They both suffer in they tell you a compound solution and it can be confused as a real time course display by your brain or lack thereof.

In unsettled water the magnetic compass to me is easier to hand steer. It resolves the motion of the boat faster against where you are pointed. The fluxgate over time works exceptionally well in adverse conditions when combined with a course computer to compensate and drive an autopilot. In that context I'm most trusting to the fluxgate. It's just not tuned to be a real time readout at all.

A GPS is a tad easier but steering a boat by watching the GPS COG will drive you nuts if not aground in close quarters. Adding a chart plotter just feeds that bad idea quicker.

If you also can see a GPS course to steer it is a plus. If you can use the fluxgate to steer the auto pilot with a course computer it too is a very huge plus. I'm not arguing that I would not want all three. As a display I find the fluxgate to be the poorest and the GPS inadequate.

In a strong cross current about equal to the boat speed both can be deadly. Computing set and drift is nothing either can perform well unless you did it manually and aimed the boat. Watching the GPS COG will seem totally illogical and the fluxgate will confirm the same result and both be totally wrong. The GPS VMG against a way point would in that case be the best display given you are not pointed to where you are going and and not moving in the direction you are pointed while moving as best as possible to where you mean to be. The GPS VMG will seem wrong over short periods of observation. You end up with no real time display of your problem and must have faith in the set and drift computation.

[Ziggy]

Originally Posted by estarzinger

Just FYI . . . Typical yachting GPS accuracy - from a furuno spec sheet:

Accuracy

GPS: 10 m (95%)
DGPS: 5 m (95%)
WAAS: 3 m (95%), limited coverage
SOG: ±0.2 kt (SOG ≦ 10 kt)

COG: ±3° (SOG 1-17 kt), ±1° (SOG > 17 kt)

Source: http://www.furunousa.com/ProductDocu...2011-17-05.pdf



Good explanation.

See above, Furuno says their COG accuracy spec is ±3°. This is higher than your 1.2 degrees because the IMO COG spec calls for 30 second averaging. With the 6 minute averaging you are proposing the gps readout response too slowly (at least for the IMO) to course changes.

But if you can turn the gps averaging up to 6 minutes, and you can hold a steady magnetic compass course for that long, and have a track the long without any significant current or wind, then I agree its possible to achieve your 1.2 degree accuracy. That's still marginal accuracy if you are trying to create a proper deviation table - deviation errors on a (fiberglass) cruising boat are often in the 2-3 degree range, so your gps COG source is creating 30-50% noise.

Again, it just depends on what the OP is trying to accomplish. If he is trying to create a professional quality deviation table then using the COG is not the way to do it. But if he is just trying to check that his compass is not way way out of whack, then the COG will do the job. Since the OP seems to have disappeared we may never know

Evans,
Thank you for supplying some "real" specs here. I used 6 minutes in my example only because it made the math simpler--I agree 6 minute averaging would be too long an interval for trying to adjust a compass in anything but flat calm conditions. Even so, there would still be current to contend with.

By the way, in my experience, Furuno's 3m/95% WAAS spec seems a bit optimistic. My Garmin certainly does not seem to do as well.

[osirissail]

The original question was about building a Deviation Card for your magnetic compass using a GPS. Early on in GPS's history of when it became available to the "masses" it was thought, "Wow, GPS will do this very nicely." It was assumed by everybody - just like personal computers were assumed - that they were "accurate" and did not make mistakes. That however, over years of actual usage was found to be quite incorrect - there are a lot of errors real and potential in using these wonderful machines.
- - The poster who suggest physical positioning of the boat and then sighting (making "gunsights" out of stays and masts, etc.) and then coordinating with a good nautical map to find the observed difference between the magnetic compass and the calculated heading from the map - is probably the most reliable and accurate method available. The problem comes in fixing the boat in a static position with no movement causing the magnetic compass to swing one way or the other.
- - In a marina slip you can probably get two readings by securing the boat tightly to the dock facing first "in" and then "out". Then you have to find someplace in the marina that is 90 degrees to the direction your slip faces and do the same thing there to get the total of 4 points of the compass. The 4 points do not really need to be exactly north/south and east/west - they just need to be as close to 90 degrees from each other as possible. Of course a trailer-sailor boat would be the easiest as you can turn the trailer. In a boatyard with a large boat this would be expensive as you would need to have the travelift reposition the boat and jackstands.
- - So for economy, doing it in the water is most desirable. Just devise a way to "lock" as much as possible the boat in one heading and take readings when the water and other factors are the most stable. Here is where your primary navigation training with taking and resolving sight positions onto a chart comes in handy.
- - Bottom line is the whole process is "not simple" although the concept may seem simple. As can be read from the other posts there are so many sources of error from all the environmental and electronic conditions in which your boat "sits." I suggested actually attempting to build a deviation table for your boat as an exercise in learning in real life about all the things and conditions that directly and indirectly influence and affect your magnetic compass and all the other navigation equipment on board. And then you will find out that nautical navigation has more to do with using your brain to average and compensate for all the "inaccurate" physical and electronic aids to navigation you have on your boat.
- - This is where "humans" are far superior to the machines/computers we use. And since it is your butt sitting in your boat as it is heading for the "rocks" even when the GPS or whatever says there is clear water ahead - your learning how to integrate and use the various navigational tools available is quite personally important.
- - Here is a table I found on GPS averaging and errors in plot versus time. You can see that under an hour or two the errors are most pronounced.

[goboatingnow]

Furuno quote CoG accuracy on their GPS navigators as +-3% under 17 knots and +- 1% for over 17knots, thats upto 6 degree error on a sailboat!!

Dave

[GordMay]

Originally Posted by goboatingnow

Furuno quote CoG accuracy on their GPS navigators as +-3% under 17 knots and +- 1% for over 17knots, thats upto 6 degree error on a sailboat!!
Dave

Wouldn't ± 3% of 360 degrees be up to 10.8 degrees either side (±) of true?

[goboatingnow]

read degress, typed percentage doh!

Dave

[cal40john]

Originally Posted by estarzinger

Just FYI . . . Typical yachting GPS accuracy - from a furuno spec sheet:


Accuracy

GPS: 10 m (95%)

DGPS: 5 m (95%)
WAAS: 3 m (95%), limited coverage
SOG: ±0.2 kt (SOG

≦ 10 kt)
COG: ±3° (SOG 1-17 kt), ±1° (SOG > 17 kt)

Source: http://www.furunousa.com/ProductDocu...2011-17-05.pdf



Good explanation.

See above, Furuno says their COG accuracy spec is ±3°. This is higher than your 1.2 degrees because the IMO COG spec calls for 30 second averaging. With the 6 minute averaging you are proposing the gps readout response too slowly (at least for the IMO) to course changes.

But if you can turn the gps averaging up to 6 minutes, and you can hold a steady magnetic compass course for that long, and have a track the long without any significant current or wind, then I agree its possible to achieve your 1.2 degree accuracy. That's still marginal accuracy if you are trying to create a proper deviation table - deviation errors on a (fiberglass) cruising boat are often in the 2-3 degree range, so your gps COG source is creating 30-50% noise.

Again, it just depends on what the OP is trying to accomplish. If he is trying to create a professional quality deviation table then using the COG is not the way to do it. But if he is just trying to check that his compass is not way way out of whack, then the COG will do the job. Since the OP seems to have disappeared we may never know

What does typical mean? I seem to remember that when I was making the deviation table that it took less than 10 seconds for the course on the GPS to stabilize, and I assumed some of that time was due to the autopilot settling out. After that the course displayed by the GPS either did not change or varied occassionally by 1 degree (not +/-, just one number different). I ran maybe 30 seconds to a minute in each direction as I wanted to make sure the course was stable, then was noting down both steering compasses and the autopilot's fluxgate. To me this says that the averagiing is doing better than typical as I would expect to see the course vary both up and down in this time frame. If the averaging time were too long I should see a slow change in one direction. If I'm getting a consistent number but it is not in the correct direction then it seems that it should be a different problem than noise.

Most times the field that says my position is +/- so many feet shows a number that is better than what typical is supposed to be. Depending on how you define typical, it doesn't necessarily mean always must be this bad.

John

[goboatingnow]

lack of accuracy doesnt mean the display is jumping around, ( thats more correctly called jitter). Its just that the steady CoG figure you are seeing ( steady as its heavily damped), is not the correct CoG, ie the boat isnt going where the Cog says it is.

Dave

[Troubadour52993]

The USN ships had two columns, one to use when the anti-magnetic signature degaussing coils were turned on, and one when they were off.

If the card is really going to be used, instead of just shown to a surveyor, there will likely need to be similar columns in a boat table, one with the engine on and off. There isn't enough physical separation available on my boat for the compass, the fluxgate anyway, to be unaffected by the electrical currents from cables and I know that when switching "from sail to steam" under autopilot I need to adjust otto by a few degrees when the diesel is fired up or shut down.

We always used a physical range or sun azimuth for determining compass error. If you use it to determine the compass card, do you assume compass error is 0 until the card is written, and then the next time you get a difference it goes into the compass error column?

[cal40john]

Originally Posted by goboatingnow

lack of accuracy doesnt mean the display is jumping around, ( thats more correctly called jitter). Its just that the steady CoG figure you are seeing ( steady as its heavily damped), is not the correct CoG, ie the boat isnt going where the Cog says it is.

Dave

If the COG is heavily damped then when you change course COG is going to lag and then slowly creep onto the new course. With enough damping it could be minutes, but the lag is going to be huge at that point. It isn't going to hold steady on a wrong course due to damping.

If the accuracy (got the right position) is off, but the precision (repeatable results, but not necessarily accurate) is high, then COG is going to be correct.

[Wotname]

Originally Posted by barnakiel

........

You cannot make your life jackets. You cannot self-service your life raft. What makes you think you can make your own table?

;-)))

Cheers,
b.

Maybe be off topic but why do you think you can't make your own lifejackets?
Bottom end jackets aren't exactly rocket science; making one's own table is even easier.