any compass adjusters near Hull ?
Nigel
to be very thin on the ground.. Anybody know of one ?
Thanks
Hugo 'NOx' Tyson
martin <mar...@wanadoo.nl> writes:
> By a Raymarine flux gate compass, it's self calibrating, use the
> fluxgate compass to correct your magnetic compass.
Does it have a GPS in it then?
- Huge
Hugo 'NOx' Tyson
martin <mar...@wanadoo.nl> writes:
> On 23 Apr 2003 20:41:09 +0100, Hugo 'NOx' Tyson
> <hm...@surfingsuggestionx.co.ukx> wrote:
> >martin <mar...@wanadoo.nl> writes:
> >> On Wed, 23 Apr 2003 09:07:40 +0100, "Nigel"
> >> <nigelREMOVE_THIS_...@hotmail.com> wrote:
> >>
> >> >I require the services of a compass adjuster in the Hull area but they seem
> >> >to be very thin on the ground.. Anybody know of one ?
> >>
> >> fluxgate compass to correct your magnetic compass.
> >
> >Does it have a GPS in it then?
>
(Actually, I wasn't concentrating; my reponse would have been "so how does
it know the local variation then?" which isn't the question asked, but
nontheless...)
Seriously, so how can it know the deviation curve for your boat then?
== How can it know more than a magnetic needle type compass?
- Huge
MP
B.Cooke & Son Ltd Compass Adjusters
58-59 Market Place Hull HU1 1RH
Tel: 01482 223454 / 224412
Mike Starr Hull Marina
"Nigel" <nigelREMOVE_THIS_...@hotmail.com> wrote in message
news:UUrpa.250$1n2....@newsfep1-win.server.ntli.net...
Ronald Raygun
>
> martin <mar...@wanadoo.nl> writes:
>> On 23 Apr 2003 20:41:09 +0100, Hugo 'NOx' Tyson
>> <hm...@surfingsuggestionx.co.ukx> wrote:
>> >martin <mar...@wanadoo.nl> writes:
>> >> Buy a Raymarine flux gate compass, it's self calibrating, use the
>> >> fluxgate compass to correct your magnetic compass.
>> >
>> >Does it have a GPS in it then?
>>
>> no it has a fluxgate in it.
>
> (Actually, I wasn't concentrating; my reponse would have been "so how does
> it know the local variation then?" which isn't the question asked, but
> nontheless...)
>
> Seriously, so how can it know the deviation curve for your boat then?
> == How can it know more than a magnetic needle type compass?
Self-adjusting compasses work by getting you to motor the boat
round in calm conditions in a long slow circle. The key
assumption made is that the real rate of change of course is
constant during this manoeuvre. If the sensed compass course
change rate is not constant during this exercise, the assumption
is that the difference is due to deviation.
Ric
"martin" <mar...@wanadoo.nl> wrote in message
news:tvpdavssob738q3sr...@4ax.com...
> fluxgate compass to correct your magnetic compass.
Adjusting a compass is not necessarily an easy job, even if you have a
reasonably accurate fluxgate compass to calibrate to (unless you just want
to make do with a deviation table - but I think the OP wants to adjust his).
Here in France we have to carry a little certificate on the boat showing
that the compass has been properly adjusted.
And relying on your fluxgate compass is not necessarily the best idea in
the world - I'm sure even the yank fellow will tell you that no worthy sea
salt would ever rely totally on a flux compass on his boat blah blah yawn.
Nigel
> round in calm conditions in a long slow circle. The key
> assumption made is that the real rate of change of course is
> constant during this manoeuvre. If the sensed compass course
> change rate is not constant during this exercise, the assumption
> is that the difference is due to deviation.
A compass can only react to the magnet field it finds it's self in, ie a
combination of the earths and the boats magnetic field. I don't understand
how a compass, fluxgate or otherwise, could deal with this. In order to use
the change in coarse for calibration, it would need to know what error was
present when it started. I think perhaps you are talking about variation,
not deviation.
Ric
"Ronald Raygun" <no....@localhost.localdomain> wrote in message
news:MkFpa.1788>
> Self-adjusting compasses work by getting you to motor the boat
> round in calm conditions in a long slow circle. The key
> assumption made is that the real rate of change of course is
> constant during this manoeuvre. If the sensed compass course
> change rate is not constant during this exercise, the assumption
> is that the difference is due to deviation.
>
Doesn't it also require a GPS input? Otherwise, how does it know where
magnetic north points? I thought the self-calibration worked by measuring
the flux-gate bearing against a GPS bearing. I suppose it might also
fine-tune that by measuring rate of change as you say, but surely it needs
at least one hard fix.
Ronald Raygun
On the contrary. Variation would be constant and therefore undetectable.
Deviation varies with heading, and so if you motor round in a slow and
steady circle, your true heading will change at a constant rate, but
your compass heading will not, because it will be affected by deviation.
Therefore the clever compass can deduce the shape of its deviation
curve, and is now able to transform from uneven to even. It still
needs one further piece of input, a constant correction, because all
it has been able to do is work out the curve's shape, not its absolute
offset.
This constant might as well take in variation as well, so that the
compass displays true heading all the time, but that's really only
useful if, like our friend in Florida, you never venture far enough
afield for variation to change.
Nigel
> Doesn't it also require a GPS input? Otherwise, how does it know where
> magnetic north points? I thought the self-calibration worked by measuring
> the flux-gate bearing against a GPS bearing. I suppose it might also
> fine-tune that by measuring rate of change as you say, but surely it needs
> at least one hard fix.
Why would a fluxgate compass require a GPS? aren't they two differenet
things
A GPS only knows where it is and where it was, it has no idea where it is
heading.
Ronald Raygun
Indeed it does, though I'm not sure GPS is necessarily the best way
of telling it. Of course, in the case of autopilots, they don't
actually need to know absolute heading, to them everything is relative.
Ric
It does if it measures the bearing between where it is and where it was.
Nigel
is
> > heading.
>
> It does if it measures the bearing between where it is and where it was.
>
Then it would know where it had been heading at some time in the past, which
is not necessarily where it is heading now
Ric
You are correct that it can'tmeasure the rate of change of heading, but for
most transport puropses (particularly sailing) rate of change of heading is
small and a "historical" heading is accurate enough for most purposes, I'd
guess. That is why, I suppose, when autocalibrating a flux gate compass
linked to GPS, you have to do a big wide circle. Whatever, when I did mine,
it kept bleating at me if I overtightened the calibration circle.
Hugo 'NOx' Tyson
Ronald Raygun <no....@localhost.localdomain> writes:
Ah, now I understand. It "knows stuff" that it "can't know" intrinsically
because you perform a special action, and tell it you're doing so, which
gives it additional information, ie. to expect a constant rate of change.
Clever stuff!
I suppose one could do the same by hand with a magnetic compass, by
circling slowly and noting the indicated bearing every 30 seconds or
whatever, and then plotting them in comparison to a presumed or predicted
constant rate of change. Much easier than dicking about with sightlines to
fixed points and large protractors on deck and so on. Interesting idea, I
might try it out.
Thanks, that makes perfect sense.
- Huge
Simon Brooke
> Nigel wrote:
>
> >> Self-adjusting compasses work by getting you to motor the boat
> >> round in calm conditions in a long slow circle. The key
> >> assumption made is that the real rate of change of course is
> >> constant during this manoeuvre. If the sensed compass course
> >> change rate is not constant during this exercise, the assumption
> >> is that the difference is due to deviation.
> >
> > A compass can only react to the magnet field it finds it's self in, ie a
> > combination of the earths and the boats magnetic field. I don't understand
> > how a compass, fluxgate or otherwise, could deal with this. In order to
> > use the change in coarse for calibration, it would need to know what error
> > was present when it started. I think perhaps you are talking about
> > variation, not deviation.
>
> On the contrary. Variation would be constant and therefore undetectable.
> Deviation varies with heading, and so if you motor round in a slow and
> steady circle, your true heading will change at a constant rate, but
> your compass heading will not, because it will be affected by deviation.
I find this all very dubious. Except in conditions of slack tide and
no wind, few boats are capable of achieving a constant rate of
turn. If your bow tends to blow off (or vice versa) you'll get a
considerable inaccuracy if there's any wind at all.
--
si...@jasmine.org.uk (Simon Brooke) http://www.jasmine.org.uk/~simon/
Hobbit ringleader gives Sauron One in the Eye.
Paul Cooper
<hm...@surfingsuggestionx.co.ukx> wrote:
Only one point that hasn't been raised - this maneuvre must be carried
out in a location where the background magnetic field is reasonably
constant. After all, it sounds as if you will be covering a few miles
of sea doing it, and local geology can cause small changes in the
earth's magnetic field over distances smaller than that. The magnetic
field gradient near geological boundaries such as the edge of a
granite mass can be quite large - a potential source of error in the
West Country or the West Coast of Scotland. More likely to be a
problem, though, would be man-made magnetic objects - wrecks, power
lines, pipe-lines or whatever. All can be avoided in deep water.
Paul
Paul Cooper
>>
>>Only one point that hasn't been raised - this maneuvre must be carried
>>out in a location where the background magnetic field is reasonably
>>constant. After all, it sounds as if you will be covering a few miles
>>of sea doing it,
>
>no AFAIR the recommended turning circle is 200 metres
>
>
>--
>martin
Well, that probably makes it immune to geological magnetic field
effects (except in very extreme circumstances not likely to be found
in the UK), but not to cables, pipelines, wrecks etc. I suppose that
good seamanship would pre-suppose that you would be doing this in deep
water, which would minimize the problem.
Paul
Ric
> existed before artificial satellites.
>
How can it sense where magnetic north really is if the fluxgate compass is
in a deviating field, which exist on all boats? I know exactly how fluxgate
compasses work - they are used extensively on aeroplanes. But you still have
to calibrate them by telling it at some point where magnetic north is. You
then need to apply a deviation correction to it at other points of the
compass.
The Raymarine fluxgate compass that you mention is linked to the autohelm
and the GPS. You have to self-calibrate its deviation by driving the boat in
a circle. I suspect the software approximates the circle into a large
polygon, with the bearing of each side of the polygon measured by the GPS.
The software then calculates the difference between this bearing and the
bearing from the fluxgate compass to apply a deviation correction to the
fluxgate reading. It bleats at you for overthightening the circle as in that
case the change in heading over one leg of the polygon is too great for the
calibration to work accurately.
The fluxgate reading is then used as the primary heading measure, because as
somebody else pointed out, GPS can only measure heading when moving in
relation to the sats, and there are obviously lots of situations where this
is not good enough.
Ric
> I don't rely on it. It's a back up to the magnetic compass.
On my boat the fluxgate compass (Raytheon) is more accurate than my compass
on the points I have checked against known bearings. This is because the
reading on the Autohelm is automatically corrected for deviation, whereas on
my compass I have to apply manually a deviation card, which at its worst is
about 3 degrees out. Not too bad. But I tend to use the compass as backup to
the fluxgate reading, rather than the other way around.
Mike L
"MP" <cou...@courier.karoo.co.uk> wrote in message
news:vadv6dj...@corp.supernews.com...
Failing that try the MCA office in Beverley, thier Surveyors should be able
to do it (for a fee!). Failing that find your self a "Master Mariner" or
similarly certified sea going bod.
M
Ric
In that case, yours will not be much more accurate than a conventional
compass at European latitudes, as it will be subjected to more or less the
same deviating fields on the boat, unless you have a huge boat and are able
to put the two compasses well away from each other.
The fluxgate would probably get a bit more accurate if you sail way up north
or south, as they can more easily deal with dip in earths magnetic field, so
rolling of the boat will not make it swing about so much.
Paul Cooper
You are mistaking the effects of local magnetic fields (compass
errors) and the global deviation from true north. A boat compass is
expected to point to Magnetic North, and marine charts show the local
difference between magnetic north and true north. A marine compass
corrected to show true North would be of little use when the vessel
sailed more than a hundred miles or so. The procedures documented are
entirely capable of finding the local compass errors, allowing the
compass reading to be corrected to give magnetic north (the full
circle allows the mean reading to be determined, which has to be
Magnetic North - the errors necessarily average out over the full 360
degrees). The only reason a fluxgate compass is used for this
maneouvre is that is reads instantly, and provides an electronic
output that can be input directly into correction algorithms. In
principle the same procedure would work with a magnetic compass, but
would be more difficult to carry out. Indeed, before electronic
fluxgate compasses were invented, this was the method used to
determine corrections, but was usually done at anchor so that the
vessel could be swung under control. "Swinging the compass" was a
usual routine before embarking on a long voyage in sailing ship days.
Paul
Paul Cooper
>
>
>I did mine in 2-3 meters of water in the IJsselmeer.
>--
>martin
Well, I guess that was the deepest water you could find ;-)
Depth of water isn't important - only the absence of external
peturbations. I guess these would be well charted in the Ijsselmeer.
Paul
Ric
"Paul Cooper" <a.paul....@NOntlworldSPAM.com> wrote in message >
> You are mistaking the effects of local magnetic fields (compass
> errors) and the global deviation from true north.
No I am not. What you call "global deviation" is referred to as magnetic
variation by the rest of us. Swinging your compass is to correct for
deviation of the ships field, and is nothing to do with variation.
Correction for that is applied separately. True virgins make dull company
etc.
Ric
No it wasn't because variation is already taken into account by the GPS,
well on my set at least. Where I mostly sail it is 1W - when I sail to
another region, I look on the charts, add the new variation into the GPS,
and my fluxgate is then spot on again. As I said, the primary use of the
fluxgate data on my boat is to give instantaneous and accurate magnetic
(hence true) heading of the boat which is not the same as the boat's track
when you are under sail, or in currents. It just avoids having a photodiode
reading the ships compass which was the case on older electronic pilots. The
reason you motor around in a circle in an area where there is no wind or
current is so that the fluxgate can calibrate itself to magnetic north from
the GPS (taking the manually inputted local variation into account) and then
automatically work out a deviation table for the boat.
Ronald Raygun
> You are mistaking the effects of local magnetic fields (compass
> errors) and the global deviation from true north.
Please don't confuse things by deviating from standard terminology.
"Local magnetic fields" (local to the vessel) cause compass errors
known as deviation. "Global deviation from true North" (global to
the vessel but local to, er, well, your locality) causes errors
known as variation.
> A boat compass is
> expected to point to Magnetic North, and marine charts show the local
> difference between magnetic north and true north. A marine compass
> corrected to show true North would be of little use when the vessel
> sailed more than a hundred miles or so.
Agreed.
> The procedures documented are
> entirely capable of finding the local compass errors, allowing the
> compass reading to be corrected to give magnetic north (the full
> circle allows the mean reading to be determined, which has to be
> Magnetic North - the errors necessarily average out over the full 360
> degrees).
Not true. The errors do not necessarily average out to zero.
Therefore the "cruise around" method of calibration can only
determine the shape of the deviation curve, not its absolute
position.
It's the job of a compass adjuster to tweak the correction magnets
in such as way as to (a) make the deviation curve average out to as
near zero as he can, and (b) to make its excursions from the average
as small as he can get them.
Ronald Raygun
> Ronald Raygun <no....@localhost.localdomain> writes:
>> On the contrary. Variation would be constant and therefore undetectable.
>> Deviation varies with heading, and so if you motor round in a slow and
>> steady circle, your true heading will change at a constant rate, but
>> your compass heading will not, because it will be affected by deviation.
>
> I find this all very dubious. Except in conditions of slack tide and
> no wind, few boats are capable of achieving a constant rate of
> turn. If your bow tends to blow off (or vice versa) you'll get a
> considerable inaccuracy if there's any wind at all.
I think you'll find that conditions are adequate more often than you
think. Of course it helps to be in a monohull. :-)
Ronald Raygun
> Ah, now I understand. It "knows stuff" that it "can't know" intrinsically
> because you perform a special action, and tell it you're doing so, which
> gives it additional information, ie. to expect a constant rate of change.
> Clever stuff!
>
> I suppose one could do the same by hand with a magnetic compass, by
> circling slowly and noting the indicated bearing every 30 seconds or
> whatever, and then plotting them in comparison to a presumed or predicted
> constant rate of change.
I think this is unlikely to be possible in practice, because unlike
a fluxgate compass which gives an instant reading, a "normal" compass
will be subject to a friction lag and also to inertia, so you couldn't
get good enough data from it while turning.
But there is an alternative to the old-fashioned method of lying to
three anchors and adjusting lines, locking the boat on a fixed heading,
and then rowing out in the dinghy with a handbearing compass.
That is, just to clamp the handbearing compass to the main vessel in
an interference-free position. Then motor on a steady compass
course and get your assistant to read the HBC for comparison, then
change course and repeat.
Ronald Raygun
> The Raymarine fluxgate compass that you mention is linked to the autohelm
> and the GPS. You have to self-calibrate its deviation by driving the boat
> in a circle. I suspect the software approximates the circle into a large
> polygon, with the bearing of each side of the polygon measured by the GPS.
Just because the compass and GPS are "linked" doesn't necessarily mean
the compass listens to course information put out by the GPS. Is it not
more likely that the purpose of the linking is to be able to display the
compass heading on the GPS screen, especially if the compass's display
is in the cockpit but the GPS screen is at the nav table?
In other words, are you just assuming they can do this, or do the
manuals actually say this is what happens?
In any case, GPS units won't give an instant course, but the fluxgate
will, and so it would make the two very difficult to relate to each
other in order to extract any useful information from the difference.
Iain Hibbert
> So long as it's moving, a GPS does know where it's heading.
A GPS knows nothing but where its been.. the "heading" you are arguing about
is in fact "course over ground". I think the heading is the direction the
boat is pointed, which as we all know can be quite different..
--
iain
Ric
"Ronald Raygun" <no....@localhost.localdomain> wrote in message
zero.
> Therefore the "cruise around" method of calibration can only
> determine the shape of the deviation curve, not its absolute
> position.
Absolutely true. I think your intial point about rate of change determining
the shape of the deviation curve is correct, but I bet that the absolute
position of magnetic north is calculated into the calibration from the GPS.
So for the likes of martin, whose Raytheon fluxgate compass is not connected
to a GPS, I suspect that he has deviation mostly dialled out, but his
compass does not necessarily point to magnetic north (lets take variation
out of the equation and assume sailing in an area of zero variation). I'd be
interested to hear from martin - when you sailed in your circle, did your
autohelm beep twice at completion of 360 and proffer up a small figure (eg a
2 or a threee)? If so, then it was able to eradicate deviation in the way
that RR suggested (the figure being the max deviation detected). If it
didn't, then sailing around in a circle did nothing in your case. If the
fluxgate has no means of correcting itself to accurate magnetic north
(either from GPS or manually inputted) it is no different from a compass
adjuster clambering on your boat without a second calibration compass (or at
least known headland bearing) but assuming that the compass already reads
accurately north and working out adjustments and deviation card based on the
not necessarlily accurate initial north reading.
Ric
"Ronald Raygun" <no....@localhost.localdomain> wrote in message
> Ric wrote:
>
> > The Raymarine fluxgate compass that you mention is linked to the
autohelm
> > and the GPS. You have to self-calibrate its deviation by driving the
boat
> > in a circle. I suspect the software approximates the circle into a large
> > polygon, with the bearing of each side of the polygon measured by the
GPS.
>
> Just because the compass and GPS are "linked" doesn't necessarily mean
> the compass listens to course information put out by the GPS. Is it not
> more likely that the purpose of the linking is to be able to display the
> compass heading on the GPS screen, especially if the compass's display
> is in the cockpit but the GPS screen is at the nav table?
I think this is merely a side-product. I suspect that a heading based on GPS
magnetic north must be fed into the system during the cruise-around
correction, otherewise as you yourself point out how does the fluxgate know
where magnetic north really is?
Incidentally, on my gps, the boat heading is not fed back to it - only a
magnetic bearing of the boat track is shown on the screen. There is no NMEA
"in" or Seatalk to the GPS, so it has no way of knowing what the fluxgate is
up to . But the actual magnetic heading of the boat is only shown on the
autohelm, which of course comes from the fluxgate (the autohelm also
receives magnetic track from the GPS which allows it to make the inital
calibration). Other systems may differ of course.
>
> In other words, are you just assuming they can do this, or do the
> manuals actually say this is what happens?
You are correct - I assumed that this is what happens. That is not to say
that I don't think you are also correct in your comment about rate of change
determining the deviation. But we both agree that somehow you have to feed
into the fluxgate a true magnetic north direction. So if you do it once from
the GPS, the logical extension is to do it lots of times to improve
accuracy - ie the "polygon" I described. This incidentally is also how
fluxgate compasses on aircraft are corrected for deviation (though they are
slaved to a gyroscope instead of a gps, and an inital true magnetic bearing
is fed to them manually to kick off the process) so it is a reasonable
assumption that the same methodology is used by Raytheon for their boat
fluxgate compass given that they also make fluxgates for aircraft and
missiles etc.
If no GPS connected, then I suspect the Raytheon at least does it the way
you initally described - with the caveat that only deviation is known, not
correction to true magnetic north. That is why I asked martin elsewhere how
his raytheon fluxgate responded when he did his little circle.
I suspect the software on the Raytheon at least is clever enough to detect
whether or not a gps is connected, then calibrate itself accordingly.
Ric
> Paul's explanation is the correct one. As I said if you don't
> understand it use google. You said you did understand it, but still
> assume that GPS is involved.
If it not involved, it can't possibly know where true magnetic north is.
Somehow, you have to calibrate a fluxgate to a true magnetic north
direction. Neither Paul, nor you, nor anybody else has yet explained how
this is done.
If you don't calibrate it to true magnetic north, it is the equivalent of a
compass adjuster adjusting your boat's compass based on the initial
(possibly erroneous) original north pointing direction. He can get rid of
deviation more or less, but the compass will not be accurate as compass
north was never adjusted to true magnetic north. It might only be a few
degrees out, and this is probably good enough for most boaties, but it will
not be accurate.
And you *can* get a true magnetic bearing from a GPS. If you motor in a
straightline (or with only very small rate of turn in the case of the
calibration circle), without wind or currents, track and heading are the
same as there is no wind leeway or tidal drift - adjust for local variation
and bingo you have a very accurate magnetic bearing, independent of the
ships deviating fields. Now if Raytheon didn't think to incorporate that
into their boat fluxgate calibration software I'll be amazed, given that
they use a similar process to calibrate fluxgates compasses on their
aeroplane compasses.
Obviously if you motor around in a tight circle, the gps cannot give a
magnetic heading as you say, because gps measures historical track bearing
which will be changing to rapidly for the accuracy of gps. That is one of
several reasons you need a fluxgate compass on an autohelm system.
Ric
"martin" <mar...@wanadoo.nl> wrote in message > Paul's explanation is the
correct one. As I said if you don't
> understand it use google. You said you did understand it, but still
> assume that GPS is involved.
On further reflection, I think your confusion may arise from an incorrect
assumption that magnetic devaition of a boat is fixed at all headings. It is
not - it varies slightly as you go around the circle. The magnetic field
detected by the compass is the resultant of the earth's field and the boat's
field. These are not necessarily in the same horizontal plane (in fact
almost never in the same plane). The resultant field therefore varies in its
elevation wrt the ships plane as you go around the circle. Both fluxgate and
normal compasses (through differing mechanisms) measure the projection of
the component of this field in the horizontal plane. If you think about the
geometry a bit, you will realise that this causes slight changes in the
deviation as you go around the circle (a sinusoidal variation curve - just
like most compass deviation curves, oddly!)
Thus deviation can be considered to be of two components, an absolute
deviation which is does not change as you go around the circle, due to
installation and instruement mechanical error, and a secondary component
which does vary due to the interaction of the different planes of the earth
and ships fields.
When you calibrate your fluxgate compass (without gps or other manual
input) by going in a circle, it can dial out the varying component,
according to the principle described at the start of the thread by RR. But
it cannot dial out the absolute component - as also pointed out by RR. For
that you need somehow to calibrate it to an accurately known magnetic
bearing.
On the Raytheon fluxgate that I have, there is no way to manually do this
(eg from a star or known transit bearing). I therefore presume that it takes
an accurate magnetic fix from the GPS, based on the argument that I have
already given. As the autohelm can get this information from Seatalk, and
Raytheon generally think of these sort of improvements and logic, and given
that these sort of ideas have been around a long time in aircraft slaved
gyro/fluxgate/inertial systems, I'd be astonished if Raytheon had not
incorporated it yet into their autohelm software.
But as you say, sailing in a circle to calibrate a fluxgate does still have
a purpose - though it can only dial out the sinusoidally varying component,
which is probably what you have acheived on your boat. I'd expect that
Raytheon have logic built into their software which calibrates with a
mixture of both methods, and so can cope with systems with no gps input, but
does a better job when it does detect gps data on Seatalk. As I have always
said, this is just a guess. But if you can think of a good reason why
Raytheon would ignore such an obvious improvement in the accuracy of their
autohelm for the cost of a few lines of software, please let me have it!
Nigel
> I am talking about deviation. The fluxgate compass sensor is mounted
> as far away as possible from large lumps of metal, unlike most small
> boat magnetic compasses that are mounted close to the engine.
>
> --
> martin
Unfortunately the boat is a large lump of metal, perhaps I could mount the
sensor on a wooden pole
Ronald Raygun
>
> "Ronald Raygun" <no....@localhost.localdomain> wrote
>> In other words, are you just assuming they can do this, or do the
>> manuals actually say this is what happens?
>
> You are correct - I assumed that this is what happens. That is not to say
> that I don't think you are also correct in your comment about rate of
> change determining the deviation. But we both agree that somehow you have
> to feed into the fluxgate a true magnetic north direction. So if you do it
> once from the GPS, the logical extension is to do it lots of times to
> improve accuracy - ie the "polygon" I described.
If the FGC takes info from the GPS, I suspect it will *not* do so
during the calibration circle (the purpose of which is to learn
the shape of the deviation curve), but at some other time, possibly
immediately afterwards, when going in a straight line.
The most likely scenario is that it just assumes (as you (or
someone else) did) that the average is zero.
Jens K
> that the compass has been properly adjusted.
>
Are you serious? Do you also need any documents certifying that your
boat isn't overloaded with certifying documents? I think I will never
sail to France, what a pity...really...
Jens
Tee
Bollocks, RTFM or do a bit of googling to educate yourself.
Ronald Raygun
> On Thu, 24 Apr 2003 23:37:25 +0100, Iain Hibbert <plu...@mail.com>
> wrote:
>
>>martin <mar...@wanadoo.nl> wrote:
>>> So long as it's moving, a GPS does know where it's heading.
>>
>>A GPS knows nothing but where its been..
>
> Isn't this true in all situations, whatever the measurement system?
> It's the meaning of life. "You ask me what of the future and I say let
> us look to the past" Beyond the Fringe circa 1962.
Hey, grasshopper, you sure it wasn't Confucius who said that?
Ronald Raygun
> please don't confuse the fluxgate compass in the St1000 autohelm with
> the ST 30 fluxgate compass that I said could be used to adjust a
> compass. As I posted elsewhere today, the fluxgate compass in the
> Autohelm is wildly inaccurate because it is adjacent to a strong
> magnetic field, namely an MD2B diesel engine. The ST30 fluxgate is
> mounted well away from any ferrous metals and gives a very accurate
> reading after self calibration.
If it's mounted so far away from deviation-inducing influences,
how does it it benefit from self-calibration at all, and what
does this self-cal actually do to achieve that benefit? Or is
there some inherent problem with fluxgates *other than* deviation
which a spin on the merry-go-round fixes?
Nigel
> I am not confused!
> --
> martin
I am...
but I think the answer seems to be 'No there arn't any compass adjusters
near Hull' :-)
Iain Hibbert
> On Sat, 26 Apr 2003 08:34:46 +0100, "Nigel"
> <nigelREMOVE_THIS_...@hotmail.com> wrote:
>>> I am not confused!
>>
>>I am...
>>but I think the answer seems to be 'No there arn't any compass adjusters
>>near Hull' :-)
> The simple obvious answer, which avoids discussing the meaning of life
> :-)
I know a chap who a few years ago paid good money to learn to be a
compass adjuster as he is getting on in years and thought it would be
a good way to earn a little money while out cruising.
It turns out that hardly anybody is interested in getting their
compasses adjusted these days, since they have GPS and just ignore
the minimal deviations on the compass that you get on fibreglass
boats.
He is a long way from Hull though, sorry..
iain
Iain Hibbert
> I think this is merely a side-product. I suspect that a heading based on GPS
> magnetic north must be fed into the system during the cruise-around
> correction, otherewise as you yourself point out how does the fluxgate know
> where magnetic north really is?
"GPS magnetic north" does not exist.
The GPS knows where the Geographic North Pole is, and it knows where the
Magnetic North Pole is, and it knows where you are, so can calculate the
local magnetic variation, but mine at least(*) has no inbuilt compass or
any way of orienting itself other than how the boat was moving during
the last few seconds and I think the errors are too significant to rely
on in that case.
> Incidentally, on my gps, the boat heading is not fed back to it - only a
> magnetic bearing of the boat track is shown on the screen. There is no NMEA
> "in" or Seatalk to the GPS, so it has no way of knowing what the fluxgate is
> up to . But the actual magnetic heading of the boat is only shown on the
> autohelm, which of course comes from the fluxgate (the autohelm also
> receives magnetic track from the GPS which allows it to make the inital
> calibration). Other systems may differ of course.
The GPS can't send "magnetic track", only geographical track..
I think also that comparing aeroplanes and boats instrumentation is
pushing it a bit - the GPS figures on an aeroplane are likely to be
much more accurate since they are only ever averages and you are
moving much greater distances in the same time interval, and thus
the GPS inaccuracies (inherent and induced) are nowhere near as
significant.
Why do you assume that Raytheon would take the easy way when writing
software? barely anybody else does in the commercial world - they get
different programmers to write different parts with different specs,
and as soon as it appears to work, send it out the door..
iain
(*) some GPS units these days do have a compass built in I believe..
Ronald Raygun
> I know a chap who a few years ago paid good money to learn to be a
> compass adjuster as he is getting on in years and thought it would be
> a good way to earn a little money while out cruising.
>
> It turns out that hardly anybody is interested in getting their
> compasses adjusted these days, since they have GPS and just ignore
> the minimal deviations on the compass that you get on fibreglass
> boats.
>
> He is a long way from Hull though, sorry..
Gone to France, no doubt, where the wine is cheap, the fags are
strong, and the living is easy, if you believe what someone here
said that French boats require adjustment certificates by law.
Iain Hibbert
>> if you believe what someone here
>>said that French boats require adjustment certificates by law.
> The question is "do they enforce the law?"
and indeed, do they actually need to be adjusted or is the
certificate itself all that is required?
iain
Ronald Raygun
Heh, heh. The certificate is all that's required, but the owner
will be jailed if he forges it, and the adjuster will be jailed
if she issues one without doing the biz.