On Saturday, 19 November 2022 at 15:42:55 UTC, Davey wrote:
> I have been slowly preparing to install a sundial on a high west-facing
> wall. I conducted several experiments to determine the exact angle that
> it faced, and concluded that it was 1/2 degree South of West.
> Then yesterday, a colleague came by. and said that his smartphone had
> an 'app' that was a compass. He held it up to the wall, and said that it
> was in fact about 25 degrees South of West. I did not believe him, if
> it was true, then the Pole Star was in the wrong part of the sky. I
> went indoors and picked up a plain magnetic map-reader's compass, and it
> showed a reading much closer to mine. The resolution wasn't good, but
> there is a difference between 25 degrees and half a degree. Even
> several feet away from the house, the 'phone still showed the same
> erroneous reading. But at the front of the house, it was fine.
> At the first location, it was next to the house, and then on the
> adjacent patio, and there is no known metal underground, in fact the
> patio has all been dug up and redone within the past few months.
> So the question is: What could be causing this quite large error, even
> if the 'phone is several feet away from the house? There is a ceramic
> sewer in the area, but that should not affect the reading. Should it?
> How does a smartphone detect compass angle?
> If it were mine, I would be wary of using it as a compass while
> walking through the Cairngorm Mountains or similar wild terrain.
> --
> Davey.

Back in the nineties I fitted a large steerable satellite dish. It needed to be pointed to within less than one degree, ideally 0.5꙳. Magnetic not good enough for this, and anyway it drifts, so I used the sun to find due south at zenith. As this is a sundial this would seem a good way to do it, although beware GMT shifts by up to 15 minutes seasonally and better sundials have this marked on them. If your wall faces west, then a sundial on it won't work in the morning or 13:00 BST.

On Sun, 14 May 2023 05:44:46 -0700 (PDT)
"notya...@gmail.com" <notyalckram@gmail.com> wrote:

> On Saturday, 19 November 2022 at 15:42:55 UTC, Davey wrote:
> > I have been slowly preparing to install a sundial on a high
> > west-facing wall. I conducted several experiments to determine the
> > exact angle that it faced, and concluded that it was 1/2 degree
> > South of West. Then yesterday, a colleague came by. and said that
> > his smartphone had an 'app' that was a compass. He held it up to
> > the wall, and said that it was in fact about 25 degrees South of
> > West. I did not believe him, if it was true, then the Pole Star was
> > in the wrong part of the sky. I went indoors and picked up a plain
> > magnetic map-reader's compass, and it showed a reading much closer
> > to mine. The resolution wasn't good, but there is a difference
> > between 25 degrees and half a degree. Even several feet away from
> > the house, the 'phone still showed the same erroneous reading. But
> > at the front of the house, it was fine. At the first location, it
> > was next to the house, and then on the adjacent patio, and there is
> > no known metal underground, in fact the patio has all been dug up
> > and redone within the past few months. So the question is: What
> > could be causing this quite large error, even if the 'phone is
> > several feet away from the house? There is a ceramic sewer in the
> > area, but that should not affect the reading. Should it? How does a
> > smartphone detect compass angle? If it were mine, I would be wary
> > of using it as a compass while walking through the Cairngorm
> > Mountains or similar wild terrain. --
> > Davey.
>
> Back in the nineties I fitted a large steerable satellite dish. It
> needed to be pointed to within less than one degree, ideally 0.5꙳.
> Magnetic not good enough for this, and anyway it drifts, so I used
> the sun to find due south at zenith. As this is a sundial this would
> seem a good way to do it, although beware GMT shifts by up to 15
> minutes seasonally and better sundials have this marked on them. If
> your wall faces west, then a sundial on it won't work in the morning
> or 13:00 BST.

Thanks, but I am way ahead of you. More and more tests, such as the 90
deg. nail and its shadow, tell me that the angle of declination is so
close to due West as to be good enough for due West to be used,
(results varying from 0.13deg. to 0.5deg South of West) but then when I
print out and make a 4' square sundial, it is not accurate up to about
2:30pm GMT deg. From then on, it gets less and less inaccurate.
Maybe it's the Equation of Time that's messing me up.
I do realise the time limitations of a West-facing sundial.
I think I'll join the British Sundial Society and ask them for advice.
--
Davey.

On 14/05/2023 15:11, Davey wrote:
> Thanks, but I am way ahead of you. More and more tests, such as the 90
> deg. nail and its shadow, tell me that the angle of declination is so
> close to due West as to be good enough for due West to be used,
> (results varying from 0.13deg. to 0.5deg South of West) but then when I
> print out and make a 4' square sundial, it is not accurate up to about
> 2:30pm GMT deg. From then on, it gets less and less inaccurate.
> Maybe it's the Equation of Time that's messing me up.
> I do realise the time limitations of a West-facing sundial.
> I think I'll join the British Sundial Society and ask them for advice.

I don't know anything about mobile phone compass function but it is
probably the same as a portable GPS and has to be calibrated every time
used.

I get some funny looks when calibrating the GPS in my camera (and get
dizzy!).

On Sunday, 14 May 2023 at 15:13:33 UTC+1, Davey wrote:
> On Sun, 14 May 2023 05:44:46 -0700 (PDT)
> "notya...@gmail.com" <notya...@gmail.com> wrote:
>
> > On Saturday, 19 November 2022 at 15:42:55 UTC, Davey wrote:
> > > I have been slowly preparing to install a sundial on a high
> > > west-facing wall. I conducted several experiments to determine the
> > > exact angle that it faced, and concluded that it was 1/2 degree
> > > South of West. Then yesterday, a colleague came by. and said that
> > > his smartphone had an 'app' that was a compass. He held it up to
> > > the wall, and said that it was in fact about 25 degrees South of
> > > West. I did not believe him, if it was true, then the Pole Star was
> > > in the wrong part of the sky. I went indoors and picked up a plain
> > > magnetic map-reader's compass, and it showed a reading much closer
> > > to mine. The resolution wasn't good, but there is a difference
> > > between 25 degrees and half a degree. Even several feet away from
> > > the house, the 'phone still showed the same erroneous reading. But
> > > at the front of the house, it was fine. At the first location, it
> > > was next to the house, and then on the adjacent patio, and there is
> > > no known metal underground, in fact the patio has all been dug up
> > > and redone within the past few months. So the question is: What
> > > could be causing this quite large error, even if the 'phone is
> > > several feet away from the house? There is a ceramic sewer in the
> > > area, but that should not affect the reading. Should it? How does a
> > > smartphone detect compass angle? If it were mine, I would be wary
> > > of using it as a compass while walking through the Cairngorm
> > > Mountains or similar wild terrain. --
> > > Davey.
> >
> > Back in the nineties I fitted a large steerable satellite dish. It
> > needed to be pointed to within less than one degree, ideally 0.5꙳.
> > Magnetic not good enough for this, and anyway it drifts, so I used
> > the sun to find due south at zenith. As this is a sundial this would
> > seem a good way to do it, although beware GMT shifts by up to 15
> > minutes seasonally and better sundials have this marked on them. If
> > your wall faces west, then a sundial on it won't work in the morning
> > or 13:00 BST.
> Thanks, but I am way ahead of you. More and more tests, such as the 90
> deg. nail and its shadow, tell me that the angle of declination is so
> close to due West as to be good enough for due West to be used,
> (results varying from 0.13deg. to 0.5deg South of West) but then when I
> print out and make a 4' square sundial, it is not accurate up to about
> 2:30pm GMT deg. From then on, it gets less and less inaccurate.
> Maybe it's the Equation of Time that's messing me up.
> I do realise the time limitations of a West-facing sundial.
> I think I'll join the British Sundial Society and ask them for advice.
> --
> Davey.

Well on reflection, you don't care where true north is, you just set it up so it shows the right time!

On Sun, 14 May 2023 12:14:20 -0700 (PDT)
"notya...@gmail.com" <notyalckram@gmail.com> wrote:

> On Sunday, 14 May 2023 at 15:13:33 UTC+1, Davey wrote:
> > On Sun, 14 May 2023 05:44:46 -0700 (PDT)
> > "notya...@gmail.com" <notya...@gmail.com> wrote:
> >
> > > On Saturday, 19 November 2022 at 15:42:55 UTC, Davey wrote:
> > > > I have been slowly preparing to install a sundial on a high
> > > > west-facing wall. I conducted several experiments to determine
> > > > the exact angle that it faced, and concluded that it was 1/2
> > > > degree South of West. Then yesterday, a colleague came by. and
> > > > said that his smartphone had an 'app' that was a compass. He
> > > > held it up to the wall, and said that it was in fact about 25
> > > > degrees South of West. I did not believe him, if it was true,
> > > > then the Pole Star was in the wrong part of the sky. I went
> > > > indoors and picked up a plain magnetic map-reader's compass,
> > > > and it showed a reading much closer to mine. The resolution
> > > > wasn't good, but there is a difference between 25 degrees and
> > > > half a degree. Even several feet away from the house, the
> > > > 'phone still showed the same erroneous reading. But at the
> > > > front of the house, it was fine. At the first location, it was
> > > > next to the house, and then on the adjacent patio, and there is
> > > > no known metal underground, in fact the patio has all been dug
> > > > up and redone within the past few months. So the question is:
> > > > What could be causing this quite large error, even if the
> > > > 'phone is several feet away from the house? There is a ceramic
> > > > sewer in the area, but that should not affect the reading.
> > > > Should it? How does a smartphone detect compass angle? If it
> > > > were mine, I would be wary of using it as a compass while
> > > > walking through the Cairngorm Mountains or similar wild
> > > > terrain. -- Davey.
> > >
> > > Back in the nineties I fitted a large steerable satellite dish.
> > > It needed to be pointed to within less than one degree, ideally
> > > 0.5꙳. Magnetic not good enough for this, and anyway it drifts, so
> > > I used the sun to find due south at zenith. As this is a sundial
> > > this would seem a good way to do it, although beware GMT shifts
> > > by up to 15 minutes seasonally and better sundials have this
> > > marked on them. If your wall faces west, then a sundial on it
> > > won't work in the morning or 13:00 BST.
> > Thanks, but I am way ahead of you. More and more tests, such as the
> > 90 deg. nail and its shadow, tell me that the angle of declination
> > is so close to due West as to be good enough for due West to be
> > used, (results varying from 0.13deg. to 0.5deg South of West) but
> > then when I print out and make a 4' square sundial, it is not
> > accurate up to about 2:30pm GMT deg. From then on, it gets less and
> > less inaccurate. Maybe it's the Equation of Time that's messing me
> > up. I do realise the time limitations of a West-facing sundial.
> > I think I'll join the British Sundial Society and ask them for
> > advice. --
> > Davey.
>
> Well on reflection, you don't care where true north is, you just set
> it up so it shows the right time!

That is the problem, I think that I know what to change to have it read
right, but it doesn't when I try it! Today's tests were not bad, the
shadow was in advance of where it should have been, and slightly
reducing, from 2:30 to 6 pm GMT. I'm going to work out what the
Equation of Time was possibly doing, and see if that makes any sense.
Each time I make an adjustment, I have to print out a new set of plans,
and make a new gnomon, it's not an easy modification.
--
Davey.

On 14/05/2023 17:00, MB wrote:
> I get some funny looks when calibrating the GPS in my camera (and get
> dizzy!).

GPS doesn't need calibration. Maybe the camera has a compass, in
addition to the GPS. GPS can't give direction unless you are moving.

On 14/05/2023 21:52, David Woolley wrote:
> GPS doesn't need calibration. Maybe the camera has a compass, in
> addition to the GPS. GPS can't give direction unless you are moving.

Sorry I meant the GPS receiver compass.

On Sun, 14 May 2023 20:54:02 +0100
Davey <davey@example.invalid> wrote:

> On Sun, 14 May 2023 12:14:20 -0700 (PDT)
> "notya...@gmail.com" <notyalckram@gmail.com> wrote:
>
> > On Sunday, 14 May 2023 at 15:13:33 UTC+1, Davey wrote:
> > > On Sun, 14 May 2023 05:44:46 -0700 (PDT)
> > > "notya...@gmail.com" <notya...@gmail.com> wrote:
> > >
> > > > On Saturday, 19 November 2022 at 15:42:55 UTC, Davey wrote:
> > > > > I have been slowly preparing to install a sundial on a high
> > > > > west-facing wall. I conducted several experiments to determine
> > > > > the exact angle that it faced, and concluded that it was 1/2
> > > > > degree South of West. Then yesterday, a colleague came by. and
> > > > > said that his smartphone had an 'app' that was a compass. He
> > > > > held it up to the wall, and said that it was in fact about 25
> > > > > degrees South of West. I did not believe him, if it was true,
> > > > > then the Pole Star was in the wrong part of the sky. I went
> > > > > indoors and picked up a plain magnetic map-reader's compass,
> > > > > and it showed a reading much closer to mine. The resolution
> > > > > wasn't good, but there is a difference between 25 degrees and
> > > > > half a degree. Even several feet away from the house, the
> > > > > 'phone still showed the same erroneous reading. But at the
> > > > > front of the house, it was fine. At the first location, it was
> > > > > next to the house, and then on the adjacent patio, and there
> > > > > is no known metal underground, in fact the patio has all been
> > > > > dug up and redone within the past few months. So the question
> > > > > is: What could be causing this quite large error, even if the
> > > > > 'phone is several feet away from the house? There is a ceramic
> > > > > sewer in the area, but that should not affect the reading.
> > > > > Should it? How does a smartphone detect compass angle? If it
> > > > > were mine, I would be wary of using it as a compass while
> > > > > walking through the Cairngorm Mountains or similar wild
> > > > > terrain. -- Davey.
> > > >
> > > > Back in the nineties I fitted a large steerable satellite dish.
> > > > It needed to be pointed to within less than one degree, ideally
> > > > 0.5꙳. Magnetic not good enough for this, and anyway it drifts,
> > > > so I used the sun to find due south at zenith. As this is a
> > > > sundial this would seem a good way to do it, although beware
> > > > GMT shifts by up to 15 minutes seasonally and better sundials
> > > > have this marked on them. If your wall faces west, then a
> > > > sundial on it won't work in the morning or 13:00 BST.
> > > Thanks, but I am way ahead of you. More and more tests, such as
> > > the 90 deg. nail and its shadow, tell me that the angle of
> > > declination is so close to due West as to be good enough for due
> > > West to be used, (results varying from 0.13deg. to 0.5deg South
> > > of West) but then when I print out and make a 4' square sundial,
> > > it is not accurate up to about 2:30pm GMT deg. From then on, it
> > > gets less and less inaccurate. Maybe it's the Equation of Time
> > > that's messing me up. I do realise the time limitations of a
> > > West-facing sundial. I think I'll join the British Sundial
> > > Society and ask them for advice. --
> > > Davey.
> >
> > Well on reflection, you don't care where true north is, you just set
> > it up so it shows the right time!
>
> That is the problem, I think that I know what to change to have it
> read right, but it doesn't when I try it! Today's tests were not bad,
> the shadow was in advance of where it should have been, and slightly
> reducing, from 2:30 to 6 pm GMT. I'm going to work out what the
> Equation of Time was possibly doing, and see if that makes any sense.
> Each time I make an adjustment, I have to print out a new set of
> plans, and make a new gnomon, it's not an easy modification.

And The Equation of Time appears to be the solution. The correction for
my Longitude is -4 minutes, and for May 16th is -3.5 minutes, making a
total correction of -7.5 minutes. Which matches the discrepancy I see
on my mock-up sundial. QED.

--
Davey.