Watch-chronometers

[oldbilbo]

Those interested in the dustier corners of the 'navigation art' will be aware of the challenge of deriving GMT/UTC with accuracy sufficient for astro-navigation at sea. They ( we? ) will also be committed to the reduction of the numerous 'errors' to a minimum. 'Timng error' has always been one of these.

The answer has long seemed to be an expensive and cossetted chronometer - or the nearest to that to which funds could stretch.

No more.

Gary LaPook at NavList@fer3.com has conducted an experiment using 3 cheap quartz watches here. His study suggests that this economic approach can give a residual timing error well less than 1' arc, which is of the order of 0.25nm.

Most of us already possess one or more cheap quartz watches..... This 'array of watches' is a low-cost but high-tech and effective solution, appealing to the Jester spirit.

 

[jwilson]

Those interested in the dustier corners of the 'navigation art' will be aware of the challenge of deriving GMT/UTC with accuracy sufficient for astro-navigation at sea. They ( we? ) will also be committed to the reduction of the numerous 'errors' to a minimum. 'Timng error' has always been one of these.

The answer has long seemed to be an expensive and cossetted chronometer - or the nearest to that to which funds could stretch.

No more.

Gary LaPook at NavList@fer3.com has conducted an experiment using 3 cheap quartz watches here. His study suggests that this economic approach can give a residual timing error well less than 1' arc, which is of the order of 0.25nm.

Most of us already possess one or more cheap quartz watches..... This 'array of watches' is a low-cost but high-tech and effective solution, appealing to the Jester spirit.

I don't think anyone has seriously used "chronometers" for years - cheap Casio waterproof digitals are the usual choice - carry two and rate them before voyage.

 

[sarabande]

establishing the rate, and verifying the consistency of the rate has always been part of the maths of nav. That work you referenced seems to have been carefully carried out, and is very encouraging. Cheap digital watches at £10 each, and a sextant at £200 - bullet (and solar flare) proof


I'd like to see some digital oscillators, as used in decent digital watches, linked up to a big display with a touch function for freezing / recording the time of sight, perhaps by means of a touch screen.

Isn't Angus McDoon into these sort of technologies, or is it someone else on board the forum ?

http://www.wharton.co.uk/digital-wall-clock/LED-digital-clock-specifications.htm

 

[prv]

I don't think anyone has seriously used "chronometers" for years - cheap Casio waterproof digitals are the usual choice

Agree - Bilbo's post may have been revolutionary in the 70s, but it's business as usual these days as far as I'm aware.

carry two and rate them before voyage.

Personally I'd follow the same advice as for compasses - "carry one or carry three; two will drive you mad". I.e., if you have two and they disagree, which one is right? At least with only one you'll be in happy ignorance if it's wrong . So three watches would seem like the right number.

Pete

 

[snooks]

Couldn't you just take the time off the GPS?

 

[maxi77]

Establishing the rate is the same whether it is a chronometer or a digil watch and is required for both. The significance of the chronometer was the stability of it's rate rather than as often assumed it's lack of a rate. Digital watches tend to be quite stable if they use properly matured and calibrated crystals, and are an excellent tool. I would shy away from the cheapest because the quality of the crystal is one of the places where they save cost, but by the time you get up to something like the Casio Seapathfinder they seem pretty stable. I lost one a year ago and found it a couple of days ago and it was only a couple of minutes out in time, pretty much in line with the rate it showed when in use.

My worst sight error ever was the result of a time error, the deck watch had stopped and the person responsible for it had simply rewound it and set it approximately. It put us of Angola when we knew we were just of Beira in Mzambique. Correcting the time error put us back where we were.

 

[maxi77]

Couldn't you just take the time off the GPS?

I have heard questions on the latency of the display, but I am not sure. Latency itself is not a problem though as long as it is fixed and the value known.

 

[onesea]

I have heard questions on the latency of the display, but I am not sure. Latency itself is not a problem though as long as it is fixed and the value known.

So you would need a GPS time error log

 

[sarabande]

assuming the GPS is still working two days before Armageddon....

 

[snooks]

assuming the GPS is still working two days before Armageddon....

If you knew it was two days before armageddon, would you care where you were?

 

[oldbilbo]

Agree - Bilbo's post may have been revolutionary in the 70s, but it's business as usual these days as far as I'm aware.

Personally I'd follow the same advice as for compasses - "carry one or carry three; two will drive you mad". I.e., if you have two and they disagree, which one is right? At least with only one you'll be in happy ignorance if it's wrong . So three watches would seem like the right number.

Sure. I was simply intrigued by the good Gary's results - which one may think have intuitive validity - confirming what one might anticipate, or hope for.

And there are those on 'ere to whom astro is a new-ish interest.

So this raises the prospect of not only doing 'line best-fitting' for a run of sights, but also the calculus for an MPT - or Most Probable Time. Then, having worked up this observation, one sharpens pencil again to construct a Most Probable Position.

As long as nothing else needs doing on deck....

 

[maxi77]

So you would need a GPS time error log

If there is latency I would suspect it is fixed for practical purposes so just a straight forward correction.

 

[Twister_Ken]

You wouldn't think the latency was a result of the processing power of the unit's chip?

I've two DAB radios. The big one puts out the time pips about a second before the small one (and both are behind an analogue radio). I assumed the big one is 'fast' and the small one is 'slow' because of differences in the speed of the components.

 

[Sailfree]

Couldn't you just take the time off the GPS?

Great idea and then ..take your position from GPS. Use time saved not doing the maths to polish the brass on the sextant.

No it would never catch on!!

 

[jwilson]

Sure. I was simply intrigued by the good Gary's results - which one may think have intuitive validity - confirming what one might anticipate, or hope for.

And there are those on 'ere to whom astro is a new-ish interest.

So this raises the prospect of not only doing 'line best-fitting' for a run of sights, but also the calculus for an MPT - or Most Probable Time. Then, having worked up this observation, one sharpens pencil again to construct a Most Probable Position.

As long as nothing else needs doing on deck....

I have quite often done a graph of six or eight sun sights, particularly when it is a bit lumpy on a small boat and you are not certain your horizon is true. If they are mostly more or less in a line with just 1 or 2 off then use one that is on the line, and if they are really mixed up draw a line down the centre of the path on the graph and pick a spot on that line and work from that. The graph exercise shows up approximately how much error can get into taking a single individual sight, and it's interesting that you can often "know" in advance which of the series might be off the line.

It gives you something to do on a sunny day with no land in sight.....

 

[oldbilbo]

It gives you something to do on a sunny day with no land in sight.....

Try this?

A. Take your 7 or 8 sun shots and times, plot them on a sheet of graph paper, discard the one ( or teo ) that are evidently 'well out', then pick one on the line.

B. Next, take the same group of sun shots, discard the same one or two as before, then sum and average the Hs's and times.

Plot the result of both A and B. Consider the difference in intercepts....

Discuss.

 

[davidej]

Couldn't you just take the time off the GPS?

If you had a working GPS I can't see why you would bother taking sights.

I thought the purpose of astro-navigation is to avoid reliance on GPS.

 

[prv]

If you had a working GPS I can't see why you would bother taking sights.

For practice?

I thought the purpose of astro-navigation is to avoid reliance on GPS.

Quite a lot of people I suspect do it mostly for fun, as an interesting challenge, or for the appeal of using an ancient technique not dependant on technology. It appeals to me on those terms, and I'll probably get round to learning properly at some point (I have a rough idea of the principles at the moment but probably couldn't work out a position unassisted). But I don't actually think it's 100% vital before setting out on an ocean passage.

Pete

 

[Kukri]

Don't, ever, polish the brass on a sextant.

It is not necessarily daft to take the time from the GPS as a means of rating the deck watch.

In the unlikely event that the GPS goes TU, you know the rate of the watch.

 

[Bosun Higgs]

establishing the rate, and verifying the consistency of the rate has always been part of the maths of nav. That work you referenced seems to have been carefully carried out, and is very encouraging. Cheap digital watches at £10 each, and a sextant at £200 - bullet (and solar flare) proof

But not cloudy sky proof, and still not very accurate. But I thought all you old fashioned navigators didnt like GPS because of the electronics / potential for failure / dud batteries etc. And whats in a digi watch but electronics and a battery.