Conference unifix::sailing

>    Has anyone tried entering the waypoints given in the Chartkit into
>    their Loran?   As compared with entering waypoints on the spot, I find
>    that the Chartkit can be off by about 00 00.1 +/-.  When trying to use
>    these for navigation, in particular in combination with points that
>    have been entered as they are passed, the error in the heading can be
>    significant.
    
>    Can anyone explain why there is a difference?  Is it my Loran?
    

From what I have read (and experienced) a LORAN's "repeatability" is 
excellent, i.e. the LORAN can remember very accurately where a given
which it passed was when it was entered - it's internal calibration is very 
good. However, from LORAN to LORAN, the external calibration will vary more.
The readings on one LORAN will be differ by a greater degree with another,
than it will differ from itself when returning to a waypoint. As to why,
well I have this book...

    During last weekend I tried out the Chartkit way points. I found
    that you were correct on the accuracy, within .1 +/-. Since the
    position is rounded to the nearest .1, you can expect a +/- .05
    error. Conpound this with the normal 100 to 200 foot accuracy of
    the loran, the Chartkit numbers are pretty good. In heavy fog you
    should be in range of hearing the marker. I would not chart a course
    from waypoint to waypoint but from here to the next waypoint.

re .0:

There are many, many reasons why your loran does not agree with the 
Chartkit loran fixes. Some are:

First, the propagation speed of the loran signal is affected by 
environmental factors -- how wet the ground is, for example. Changes in 
propagation speed affect measured TDs.

Second, the charted position of buoys isn't necessarily the actual 
position. The Coast Guard gets them in almost the right place, but buoys 
in deep water have a long mooring chain and the buoy can easily shift 
position 0.1 mile or more. 

Third, the loran signal is buried in substantial background noise. This 
causes some uncertainty in determining the signal peak. Our loran 
displays TDs to the nearest 0.1 microsecond. Even with the boat 
motionless the displayed TDs erraticallly vary by 0.2 to 0.3 
microseconds. Admittedly our loran is an old design, and newer designs 
may be better. 

Fourth, you give the position difference in degrees and seconds. 
Different lorans use somewhat different algorithms for determining 
lat/long from TDs. 

Fifth, usually any correction to the TDs required to give the correct 
lat/long is the same over a moderately large distance -- say 50 to 100 
miles. I've seen, however, some significant local variations, 
significant enough to cause a disaster. For example, be careful around 
Winter Harbor, ME. Why, I haven't any idea. 

I've recorded the TDs of the whistle off Monhegan Island several times 
over the years. The numbers have varied by as much as a microsecond. I 
think that the short-term repeatability of loran is very good. The 
long-term repeatabilty is less good. 

The usual disclaimer is that loran is an aid to navigation. My loran 
navigates better than I do, but I always assume it might be wrong. 

    Many LORANs have correction factors built in based on your specific
    location.  I believe that this feature is called ASF.  I found a
    similar, what I considered to be excessive, error untile I found out
    how to enable ASF.  Once thas was done, things looked much more
    accurate.

    An additional, yet to be mentioned source of error, is the reference
    datum. That is TD lists and TD line on many charts are generated by
    calculation rather than in-the-field measurements. ASF correction
    provide some degree of improvement for local static effect, but as
    pointed out earlier, the variation over time (ground conditions)
    are harder to fix. 
    
    Most Loran chains have more than one "monitoring station" whose
    function is detect signal drift in the chain and thus allow the
    TD to be finely shifted. But, a chain covers are very large area
    so shifting TDs is not very good for effecting local corrections.
    
    Bottom line is that there are numerous variables which can errors
    and you must know the limits and behavior of your own Loran in the
    local area.
    
    Jerry
      

re .4 and .5:

If you know the exact location of the loran master and secondary 
stations, it is a simple matter of spherical geometry to determine your 
position from measured TDs, at least in theory. 

For all the reasons previously discussed, the position calculated from 
the measured TDs usually won't be your actual position. Some number of 
microseconds must be added or subtracted from the measured TDs to give a 
calculated position that agrees with your actual position. The number 
added or subtracted varies with position, and is, I think, called the 
ASF (additional secondary factor). Trouble is, the ASF model used may 
differ from loran to loran, and the model is most likely theoretical and 
fairly coarse (the corrections are approximately correct for a large 
area). The ASF model almost certainly will not know about local 
anomalies (such as Winter Harbor, ME). 

Measuring TDs is the most basic and fundamental thing a loran does. 
Everything else, especially ASF and lat/long conversions, adds 
uncertainty and error (while, admittedly, increasing ease of use). The 
more automatic a device is, the less you really know about how it 
functions and how (in)accurate it is. 

Because of all this, if you are in trouble and someone is looking for 
you, give him/her your measured TDs, not the computed lat/long. All 
lorans should give very nearly the same TDs at the same location 
regardless of what is later done with the TDs. The lat/long computed by 
different lorans from the same TDs may differ quite a bit.

I've compiled a list of TDs (measured with my loran) of buoys, etc. I do 
as much of my loran navigating as possible using measured TDs and use 
lat/long only when absolutely necessary. My loran happily accepts 
waypoints in TDs, so I don't have to worry about lat/long and ASF errors.

Note 518 was started as a place to list the measured TDs for various 
buoys, etc. Additional contributions would be helpful to everyone. 

Note 10 also discusses lorans. 

    I don't own a LORAN yet :^( but used to work on LORAN stations for
    our uncle :^) :^(.   I'm curious about some of the comments made
    here regarding accuracy.
    
    One note said that the claimed accuracy is +/- 100 feet.   My
    recollections from uncles training are that the chain accuracy is
    more like 200-300 yards at best and degrades from there in cases
    of bizarre geometry or extreme distance from transmitters.
    
    The +/- 100 foot number is suspiciously close to the distance that
    radio waves travel in .1 microseconds and I'm wondering if the various
    receiver manufacturers are misleading folks by stating the receiver
    RESOLUTION as chain ACCURACY.   If you have a receiver that resolves
    to +/- .1 microseconds this would be an easy misconception for both
    manufacturer (though I would be appalled) and customer.
    
    What are the various claims for accuracy that are being published
    by the manufacturers????   Do the claims match the theories proposed
    by the authors of the multitude of 'LORAN guides'??
    
    The way the system works is that the chain 'master' sends out a
    series of pulses which proceed to arrive at chain slaves after
    a VARIABLE (error contribution #1) amount of time.  The slaves then 
    receive these pulses in much the same way that you do, delay them by a 
    constant time (error contribution #2) and send out their own pulses.
    You are receiving a master signal after a variable amount of time
    (error contribution #3) and a slave signal (error contribution #4)
    in a set that measures the difference time of the arrivals
    (error contribution #5).   
    
    There are a lot of variables here and I just don't buy 100 feet.
    A good rule of thumb for any measurement is that the measuring stick
    should have graduations of 10 times more accuracy than the error
    you are trying to measure.  Thus to measure to the nearest mil you
    want a micrometer that can resolve to a tenth of a mil.   In the
    case of a LORAN that resolves to .1 microseconds the only thing
    I would stake my life on is the nearest 1 microsecond.   Indeed,
    this would work out to something like +/- 1000 feet and that is
    much closer to what I recall for LORAN chain accuracy.
    
    I'm really surprised to hear of people using this system to navigate
    up a channel or canal etc.   The system was not intended for this
    kind of use.   You might get away with it 99% of the time but oh
    that 1% !
    
P.S.:   I also recall that 1/2 hour before and after sunset our measured
    propagation time difference between master and slave went to hell
    quite predictably (it always went to hell).   It had something to
    do with solar flare and sunspot activity.   We used to send this
    daily variation to NOAA.   It had something to do with the ionosphere
    and the effect on our signals was to change the groundwave-skywave
    phase relationships slightly.    Anyway, your milage may vary with
    readings taken around these times (better to be taking pictures
    then anyway).

    The +/- .1 figure I was using was in LAT/LON. the 100 to 200 feet
    accuracy was my visual observation of ships location compared to
    marker when the loran said I was at waypoint. The Chartkit positions
    are all in LAT/LON, all in all, pretty good.

    Re .8
    
    That's a good summary of the interaction between chain master and its
    secondary, except that most (if not all?) Loran chain have been
    upgraded. Part of the upgrade involved eliminating the need for
    secondary to receive the master's signal before transmiting (after the
    proper delay) its own series of pulses.
    
    With the older method, the master was a single point of failure. If
    it went down the secondaries where down also, for lack of a timing
    reference. The newer method provides each station, master or secondary
    with its own highly accurate time standard. Thus if any one station
    is off the air the balance of the chain still functions. This is
    especially important where a single secondary operates as secondary
    or even the master of an adjacent chain (as into the N.E. chain).
    
    What this means regarding errors is that one source (interstation
    variation) has approached zero! Of course all other error sources
    are still evident.
    
    Jerry
     

re .whatever

Hmmm, +/- 0.1 minute of latitude is +/- 0.1 nm, about +/- 600 feet. 

Were I to use my loran to go some place I'd never been before, I'd
assume its accuracy to be no better than 0.25 to 0.5 nm because of the 
many error sources that have been mentioned. If I were returning to
location whose TDs I'd measured some time previously, I'd assume the
loran's error to be perhaps 0.1 to 0.2 nm. 

In very special cases, a loran is much more accurate. Over a short time
period (an hour or two or a bit more) the loran system and the
environment are probably stable enough to permit returning to a location
within 100 feet or. Beyond that time, I'd be dubious. An example: My
loran will give cross-track error to the nearest 0.01 nm (about 60
feet). One foggy morning we left Pulpit Harbor, ME, and entered the
center of the entrance channel (which is about 100 feet wide) as a
waypoint. As we feared, dense fog quickly closed in and we returned to
Pulpit Harbor. Using the loran's cross-track error to guide us, we hit
the center of the entrance channel on the first try. 

	The key issue here is that you are using LAT/LONG waypoints.  There are
two major sub-systems in a modern LORAN-C receiver...  The front end receiver
and the backend "nav features"...  The errors discussed in most of the previous 
notes were system errors that should be picked up the same pretty much by any 
receiver (there are receivers that receive marginal signals better, factor in
skywave stuff, etc.)  After all, what the front end does is receive the signal
and measure the Time Delay...  Not very complicated given today's 
microprocessors, and tough to screw up by a manufacturer cutting corners...

	The back end stuff is another story...  Every manufacturer uses different
programs for conversions from TD to LAT/LONG (and back)...  These programs use
the manufacturers version of ASF corrections in particular...  I have a SeaRanger
ASB 2000...  LAT/LONG waypoints input to it will vary as much as .5 nm from the
plotted TD's...  And having sailed with it for a couple years now, the plotted
stuff is usually pretty good... On the other hand, having spent some time at
'uncles school for boys' in New London, I'm perfectly happy plotting TD's 
on charts... (Each chart should inicate if the TD info is measured or
calculated)

	One thing to look for in newer LORANs is a statement indicating that
they are using DMA's new database, which is based on claulations, but with as many
actual measurements thrown in as they could....

	For a more detailed article on this, Pratical Sailor, about two years
ago, rated a bunch of LORANS in both front end and back end catagories...  It
was interesting where various manufactures put their money, front end
sensitivity vs sophisticated nav programming.

	The moral of the story is use LAT/LONG stuff only when you don't need
real accuracy, or if you have really checked out YOUR receiver..

For the benefit of those who don't subscribe to Practical Sailor (but
who should -- this is an unpaid endorsement as PS accepts no
advertising): 

The latest issue compares four bottom of the line loran receivers
(<$400). All four gave (to within 0.3 microsecond) the same TDs but the
displayed latitudes and longitudes varied considerably. Only the Apelco
has automatic ASF corrections, which result in its displayed
latitude/longitude being correct within maybe 600' (I read the article
quickly at breakfast). Without the automatic ASF correction the
displayed latitude/longitudes of all the units were off by as much as
1500'. Repeatable accuracy (not absolute) was very good -- a few boat
lengths (though I don't recall that the length of the boat was
specified). If you don't have a loran, the Apelco DXL6300 would be a
good investment. (I'm a little tempted to buy one just for its small
size and low power consumption.) 

    I use the small Apelco Loran and use Lat/Long readouts. (I think
    the model # is DKL 6300 or something like that, the really small
    unit).
    
    I never get more than 150 feet innaccuracy. I've done that in ASF
    all the way from Warren, Narragansett, to Providence, to Boston.
    
    
    Good enough for me!!!
    
    Regards,
    
    Ray

re .14:

Which shows two things:

1. The latitude/longitude accuracy of the ASF algorithm can vary with 
geographic location. I don't know how large an area each ASF data point 
covers, but I'll bet it misses some/many local anomolies.

2. You shouldn't trust that accuracy until you've confirmed it by actual
experience. Would you be willing to bet your life on that accuracy when
making a landfall on, say, Mt Desert Rock in dense, dense fog? I wouldn't. 

As every loran manual says (or should say): Loran is an aid to 
navigation.

signed,

cautious_sailor_who's_only_hit_one_rock_in_13_years

    
    I believe (and what I have been reading seems to agree) that LORAN
    offers information that is Repeatable.  However, the information
    is not necessarily Accurate.  It is good at getting you back to
    the same spot.  It just you can't be sure where that spot is.
    
    I have made it a practice to save where I am as a waypoint when I
    anchor at night and then compare that reading to one taken the next 
    morning.  If I have "moved" then I assume that any of the factors
    (propogation delay, sun spot activity, etc) are consistant over the
    distance I can sail in a short period of time and I add the range
    and bearing to readings I take over the next few hours.
    
    I had not considered the morning/evening deviation (when the sun is
    low) although I should have thought of it (I was trained by the same
    Uncle).
    
    Any comments on this practice?