Checking a Compass by the Sun

[Skylark]

A question to those learned forumites in the ways of traditional navigation, please.

The compass can be checked by means of the known bearing of the rising and dipping of the sun, found tabulated in the Almanac, or from Zn, AP3270 Tables derived from a sighting.

How is this done in practice, what are the limitations, expected acuracy and any useful tips from the pros?

For example, if sailing south, how can you acurately sight the suns easterly rise? How do you look through the compass...to add an imaginary lubber line? I've read that some make us of the Portland Plotter from the chart table as an aid?

This would give a single deviation value, net of variation, for that heading. Would you be content with this one check and simply wait until you have a new heading or would you steer on a number of headings, if so how many, until a deviation card could be produced?

Thanks in advance

David

 

[john_morris_uk]

A question to those learned forumites in the ways of traditional navigation, please.

The compass can be checked by means of the known bearing of the rising and dipping of the sun, found tabulated in the Almanac, or from Zn, AP3270 Tables derived from a sighting.

How is this done in practice, what are the limitations, expected acuracy and any useful tips from the pros?

For example, if sailing south, how can you acurately sight the suns easterly rise? How do you look through the compass...to add an imaginary lubber line? I've read that some make us of the Portland Plotter from the chart table as an aid?

This would give a single deviation value, net of variation, for that heading. Would you be content with this one check and simply wait until you have a new heading or would you steer on a number of headings, if so how many, until a deviation card could be produced?

Thanks in advance

David

Traditionally done with a pelorus - although the old style Sestrel Moore compass had a hole in a metal fitting in the centre of the glass in which you could put a pin to give you the shadow of the sun as a direct measurement.

The accuracy is to within less than a degree and is limited by the motion of the boat stopping you read the compass properly in my experience.

 

[maxi77]

Certainly it was standard practice in the RN to check the gyrocompass from the bearing of the rising and setting sun, always remebering that at sunrise and sunset because of refraction in the atmospher you take the bearing when the sun is seen to be a semi diameter above the horizon

 

[ancientsailor]

On passage, we were required to check ship's gyro compasses twice a day and record error in the log book. Used to do it sunrise and sunset.

We used to reckon on accuracy of about a degree or better (as above).

To build up your deviation card, you would have to repeat the exercise with the boat on different headings.

I have seen "dome" type steering compasses with a central pin on the card, for the purpose of checking the error by the sun. Couldn't find one in the catalogue that came in the post this morning however!

If you don't have a pin on the card, it may be possible to use the shadow of something across the compass. You'd ideally need something on the centre-line - such as the single backstay(!) on a sailing boat ... only really practicable in still water.

 

[jdc]

Well I'm certainly not a pro, nor particularly skilled in traditional stuff and am entirely self taught, but I have had to navigate by sextant out of necessity. So the following is a form of admission of ignorance or prejudice or both!

I think that, at least on small sailing yachts made of wood or grp, with magnetic compasses, using the azimuth of the sun at rise or set ('amplitude'?) is pretty much a complete waste of time. Clearly if struck by lightning in a steel boat the deviation can change dramatically, and the technique would come into it's own, but in our normal mid latitude sun-run-sun navigation, one knows pretty well how good the DR is, and so a significant deviation would become apparent, and is not a cumulative error anyway. Gyro compasses drift, so finding the drift would be useful and doubtless explains why the navy used amplitudes (in which case a single measurement would be adequate because the error being measured would then be a simple rotation of the whole compass, not a full quasi-sinusoidal deviation card).

Sources of error are that the compass swings around a degree or 5, one is often well heeled over which makes use of the pelorus hard (and I can't use a shadow with my binnacle). Glancing at reduction tables I see that the azimuth changes by less than 1 degree over at least a full degree of altitude (ie several diameters) when near rising or setting, so this is a small contribution to error, and I expect in perfectly calm conditions one could measure to a degree.

 

[Robin]

I did ours using a Pelorus I bought from West Marine in the USA, I think made by Davies. I also had a programmable calculator (Casio) which I had loaded an Astro program complete with almanac courtesy of PBO centuries ago that made the calculations very quick. You also need the time very accurately and I got this from the GPS time.

You could make a DIY pelorus from a compass rose or an old protractor and make a sight arm for it out of some perspex or similar.

I did all of the readings when in a very steady situation, over time as and when sun sand heading opportunity arose, mostly when in visitor berths or moored fore and aft at different headings. It isn't necessary to do all the headings at the same time.

For those who think their compass is OK as is, we had ours professionally swung and adjusted on that boat and it was wrong after that although beforehand I thought it was probably OK. We had the adjuster back and he repeated the swing but I was still unhappy with the result hence I did it again myself. What I should have done first but didn't was to zero or remove the adjuster magnets before I did my own swing, maybe it would have shown there was little error in the first place, however I did it with the adjusters as set second time round by the 'pro' and had a very useable deviation curve as well as another done one for the electronically swung autopilot compass.

 

[timbartlett]

On RN ships, the bridge compass is designed so that it is relatively easy to take a bearing of the sun at any time, particularly at sunrise/sunset. The same can't be said of most yachts!

The way I prefer is to use the shadow pin that sticks up from the centre of the compass card (if it has one -- they seem to be missing from most modern compasses!)

You just drive the boat round in a "threepenny bit" pattern, i.e. a series of short straight lines, altering course 30 degrees between each straight line, until you have gone through a complete circuit of 360 degrees.

One each straight line, you note the direction of the shadow pin's shadow, and the time.

The shadow gives you the magnetic bearing of the sun, whatever sight reduction process you care to use gives you the true bearing, and the chart (or http://www.ngdc.noaa.gov/geomagmodels/Declination.jsp)
will give you variation.

 

[Lady Campanula]

Compass checks

It has long been the principle, and often the practice, that a conscientious navigator - i.e one who has it in mind to come home again to talk about it - would check the steering compasses ( p. is the plural 'compii' ? ) on regular and frequent occasions, to discover - before any harm is done to reputation or employment - whether something unmentionable has occurred to degrade the accuracy of those usually reliable instruments.....

....such as a bag of ship's plumber's spanners being left in close proximity, and causing all sorts of grossnesses.

On REAL aircraft such as the 'Empire', 'Sunderland', and 'Sandringham', it was an Absolute Requirement that 'compass comparisons' were made and logged both regularly and shortly after each major heading change. A 'compass check' on the gross accuracy would similarly be carried out as part of taking an astro sight, using the aircraft's pendulous reference sextant which had a pelorus, or 'dumb compass', inbuilt for this purpose.

Read the text

Later aircraft used in the maritime role carried various designs of pelorus, for the need was clearly established. Below is one example, adapted and used by the author during his 'Ocean' qualifying trip in the Eastern Atlantic too many years ago. The sight vane, prism and sunshades are clearly visible. The RYA's esteemed and feared 'Ocean viva' examiner - used to multiple 'Zero/Hero' candidates turning up in turn, each with the exact same single suns sight and the exact same errors, was so overcome that he ferreted about in his cupboard and pulled out his own, very similar, ancient device. "No more questions, milad. Tha'll do nicely."

There was a miniature-size pelorus fitted as original equipment to Fleet Air Arm 'Buccaneer' strike aircraft, and used by the observer to help ensure return to their carrier, both day and night.

When those aircraft were transferred to the Royal Air Force, the simple-yet-essential instrument was removed, its purpose ill-understood. The RAF's Buccaneers were originally intended for use in the Land Battle, and their 'different tradition' navigators were expected to read the road signs en route to their targets.

When a maritime strike role was added, in which the likes of the Red Banner Sovyetsky Voyenno-Morskoy Flot 'Kirov'/'Frunze'/'Kalinin'/'Yuri Andropov' were expected targets, it was envisaged that very distant sorties may become necessary, certainly into the Barents Sea, and the challenges to high-latitude non-electronic navigation led to the reintroduction of the miniaturized aircraft pelorus in 'Grid Navigation' techniques ( no unclassified photo is available ). This was used memorably in the RAF's successful 'Capability Demonstration' mock attacks on Soviet 'North Sea Fleet' capital units in the Eastern Barents near Novaya Zemlya, and the southern Kara Sea, long assumed by the Red Fleet command as a 'private reserve'. The multiple aircraft 'packages' came from the east....

And recalling 'a ship's plumber's bag of spanners', it has long been known among RIN yotties that hanging the handbearing compass on a cord adjacent to the back of the bulkhead steering compass is likely to produce significant deviation. This was witnessed and remarked on board a RNSA member's boat a couple of years ago, whereupon the offending HBC was moved to the other side of the companionway. It was noted on a subsequent visit that the vessel's old Zeiss binoculars, held in a hardwood box just inside the cabin bulkhead and immediately below the bulkhead compass, had been replaced with a newer, expensive set incorporating - yes, you've guessed it - a miniature, magnetic bearing compass.....

'There are old navigators, and there are bold navigators. There are no old, bold navigators....'

 

[Twister_Ken]

Flux gate

The same topic but different...

How accurate is a flux gate compass?

Could comparison of flux gate readings with those of a trad compass give deviation for the trad version? Or would both be suffering from the same (or different) deviations?

 

[timbartlett]

The same topic but different...

How accurate is a flux gate compass?

Could comparison of flux gate readings with those of a trad compass give deviation for the trad version? Or would both be suffering from the same (or different) deviations?

A flux gate compass is still a magnetic compass, so it suffers from exactly the same deviation as a swinging card compass.

To this, it adds a "new" error all of its own, often described as "tilting error" because it increases if the fluxgate is tilted away from horizontal. At mid latitudes, tilt error is a couple of degrees per degree of tilt, but it increases dramatically towards the poles. That's why fluxgates are mounted in their own miniature gimbals or oil baths.

Clever electronics take out most of the errors when you go through the automatic calibration process, but they are only masking the symptoms of deviation: they cannot "cure" it.

 

[Twister_Ken]

Thx Tim,

Ah well, back to traditional methods.

 

[whipper_snapper]

There is one fundamental flaw in the whole process. The bally sun always seems to be 1 semi-diameter above the horizon just as I am tucking into a bacon butty and a cup of coffee.

 

[Cruiser2B]

The compass can be checked by means of the known bearing of the rising and dipping of the sun, found tabulated in the Almanac, or from Zn, AP3270 Tables derived from a sighting.

How is this done in practice, what are the limitations, expected acuracy and any useful tips from the pros?

You need a table of amplitudes from Nories or Bowditch, etc - or here: http://164.214.12.45//MSISiteContent/StaticFiles/NAV_PUBS/APN/Tables/T-22.pdf

Enter the table with your latitude and the tabulated dec of the body; this gives you a correction which is added or subtracted (depending on the dec being N/S - this is usually self-evident) from 090/270 to give you the true bearing of the body at rise/set. It's normally done at rising/setting as a) time is calculated for accurate dec and b) it keeps the compass horizontal.


It's fairly accurate, at least good enough for government work.

 

[VO5]

I had an azimuth ring specially made to my design in aluminium.

It is essentially a sundial marked off in degrees which is weighted underneath and suspended within two rings and presented by a vertical fore and aft arc that cradles it.

This arc is cradled at the base in a slot in a block of aluminium that pivots horizontally above another block in two parts that I clamp to the horizontal bar forward of the compass.

The fore and aft line on the "protractor" is lined up with the slot in the mast.

Tilt is very easy to correct.

Its horizontality is ascertained by a very small circular spirit level taken off an old Royal Navy Azimuth Ring. Therefore I am able to set it very accurately.

For stowage, the weight underneath is unscrewed, the pin removed, and the complete assembly stows flat in a special box I made to keep it.

As the ring and compass are independent of each other, it is very easy to take accurate bearings by keeping her on a steady heading and noting where the shadow falls on the protractor.

The stopwatch is kept running and accurate GMT obtained by adding the time run to the corrected chronometer reading.

The Azimuth is very easy to calculate entering arguments Latitude, Hour angle and Declination, the latter two from The Nautical Almanack, the previous from either a fix using sextant and chron or by DSC.

The accuracy is very respectable, approx half to one degree.

Therefore when swinging ship, which I do regularly to keep my hand in, I always swing through a multiple threepenny bit circle in flat water and on a calm day for maximum accuracy, and then keep a deviation card with entries at ten degree intervals.

I prefer ten degree intervals as the interpolation between headings is so easy it is done mentally and applied straight away.

I am now working on using the Azimuth Mirror assembly removed from the RN ring mounted on a new ring to fit the sundial, so that I can use it not only for celestial bearings but for other navigational purposes.

The "art and science" are not dead.

They remain very much alive in my Private Navy, in which, for observations of this nature, SWMBO is the Yeoman..

 

[VO5]

I forgot to add that in order to work out the true bearing, you need to enter ABC Tables (Norie's or Burton's) with the worked values for Lat, HA and Dec.

 

[Jamesuk]

Certainly it was standard practice in the RN to check the gyro-compass from the bearing of the rising and setting sun, always remembering that at sunrise and sunset because of refraction in the atmosphere you take the bearing when the sun is seen to be a semi diameter above the horizon

Can you just run through the proforma for getting the answer please. eg Look up A, Look up B, go to Table X then do etc. Thanks in advance

 

[Cruiser2B]

Can you just run through the proforma for getting the answer please. eg Look up A, Look up B, go to Table X then do etc. Thanks in advance

As per my post #13 calculate time of rising/setting from nautical almanac. For that time determine dec(lination) for the body (sun, etc). Enter an amplitudes table from Nories or Bowditch (http://msi.nga.mil//MSISiteContent/StaticFiles/NAV_PUBS/APN/Tables/T-22.pdf) with the dec and your latitude. This gives you a correction that is applied to 090/270 as applicable - added or subtracted depending on it being N or S dec; this is usually self-evident.

 

[oldbilbo]

You will also find the use of tables in Harry Baker's 'Astro Navigation Tables - Reeds 20***' listed as 'Compass Checking by Amplitude'.

"The integrated table of LOG COSINES and LOG SINES makes it easy to calculate amplitudes by using the formula: SIN AMPLITUDE = SIN DEC ( of the observed body ) divided by COS LAT ( of the observer )"

There's an example and more simple explanation. It is truly easy, and Harry Baker's tables are cheaper and lighter to carry. If you're really stuck, send me a PM with an email addy and I'll p'copy more stuff to you.

This process is part of the RYA Ocean Practical, for good reason. You're probably aware from earlier posts in this thread that most conscientious ocean ( air and sea ) navigators check their compass function quite regularly, for it is known that compass errors large and small arrive unannounced. Those who spot them come home to tell us about it. Those who don't, often don't.

 

[onesea]

I forgot to add that in order to work out the true bearing, you need to enter ABC Tables (Norie's or Burton's) with the worked values for Lat, HA and Dec.

Or a scientific calculator...

 

[Uricanejack]

A question to those learned forumites in the ways of traditional navigation, please.

The compass can be checked by means of the known bearing of the rising and dipping of the sun, found tabulated in the Almanac, or from Zn, AP3270 Tables derived from a sighting.

How is this done in practice, what are the limitations, expected acuracy and any useful tips from the pros?

For example, if sailing south, how can you acurately sight the suns easterly rise? How do you look through the compass...to add an imaginary lubber line? I've read that some make us of the Portland Plotter from the chart table as an aid?

This would give a single deviation value, net of variation, for that heading. Would you be content with this one check and simply wait until you have a new heading or would you steer on a number of headings, if so how many, until a deviation card could be produced?

Thanks in advance

David

In response to the OP question is two separate problems.
The method of calculation and the practicality of taking the amplitude.
Taking an amplitude of the sun is popular due to being relatively accurate, easy to take and calculate.

The reason you are taking it is to check the error of the compass.

A deviation card can be made by solar observation but the process is usually carried out using a terrestrial object after a lay up or refit or once year by a professional compass adjuster.
.
By keeping a compass error log book a compass adjuster can get some significant info on how the compass is affected.

In practice if I were to check on my boat I simply wait for sunset and steer on the sun when it is about 1 semi diameter above the horizon.
This only gives me the error on one particular heading, But even so this is enough to confirm I have not left a hacksaw blade in the vicinity of the compass.

The trouble with taking errors from my boat is the position of the compass makes it virtually impossible to take a direct bearing or azimuth using the main steering compass.

Many years ago when I sailed on an OYC boat we had a large hand bearing compass with an azimuth prism and shades allowing us to take compass errors using the sun.

My current hand bearing compass does no have shades so I do not take observations of the sun using it.
I can however take observations of other objects I.e. The Moon, Planets and stars if they are at low altitudes. Not having a prism I cant take higher objects.

An amplitude can be done with Moon, Planets or Stars. On the rare occasions its clear enough.
For Stars and planets use an approximation of the suns diameter above visible horizon.

The moon setting is possible but due to parallax and refraction of moon being much greater it is on the celestial horizon just as the upper limb sets below the visible horizon. So it is difficult to take.
In practice a low altitude compass error is more practical.

If you want to use with sunglasses or are prepared to look at setting sun through hand bearing compass.
A shadow pin is safer on your eyes if you have compass or pylorus with one..

The method is.
Immediately before or after observation of the sun compare the heading of the steering compass and the vessels heading using the hand bearing compass.

Make note of observed azimuth of setting or rising sun about 1 semi diameter above horizon..
Make note of heading by hand compass.
Make note of heading by steering compass
Make note of time GMT latitude and longitude

Using almanac look up suns declination for day and hour GMT.
Apply d correction. For minutes and seconds.

This will give you declination to use.
Use your DR latitude. Or GPS latitude..

Old Bilbo has already given the formula.

Sin Amplitude = Sine declination x Secant latitude used along with nautical tables.

Converts to use by calculator.

Sine Amplitude = Sine Declination/Cos latitude.

The amplitude is the angle at the horizon between the observers prime vertical rather than the observers meridian. It is named east or west depending upon rising or setting and North or South by declination..

So the amplitude is applied to ether 90 or 270 north or south by declination.

I.e. If suns declination 20 N rising amplitude 30 degrees

Amplitude is applied North of 90 degrees i.e. 90 -30amplitude = azimuth is 060 deg.
If declination is 20 degrees S
amplitude is applied south of 90 degrees i.e. 90+ amplitude 30 = azimuth 120 degrees

There you go now you can take and calculate an amplitude hope this helps..