'Double-ing(?) angles with a total station
Paul
In article <01bbd10a$d11f7b00$7046...@neilkg.netshop.net>,
"Neil Giesbrecht" <Nei...@netshop.net> wrote:
>With modern instruments....
[...]
Does the circuitry of these
>new guns make such a routine[doubling] unnecessary?
What "circuitry" does not eliminate are errors caused by mechanical things,
such as eccentricity of the spindle bearing thingie and collimation error,
etc. I dont think that tolerances in instruments has really improved much in
hte last 50 years, and electronic readouts give a false sense of precision. I
have a mechanical clock that keeps better time than my digital clock radio.
Neil Giesbrecht
With modern instruments is turning the angles twice on opposite sides of
the 'plate' necessary? In my case I own a Leica TC600 and turning angles
(actually bearings) twice on opposite faces is basically a pain in the ass
since the display is only on one side of the gun and it doesn't 'hold' the
foresight to use as the subsequent backsight. Does the circuitry of these
new guns make such a routine unnecessary?
It has me wondering
Jim Frame
In article <01bbd10a$d11f7b00$7046...@neilkg.netshop.net>,
"Neil Giesbrecht" <Nei...@netshop.net> wrote:
>With modern instruments is turning the angles twice on opposite sides
>of the 'plate' necessary?
I asked this same question some years back, and was told that modern
total stations have at least two sensors spread about the angle plate,
and that readings are averaged from all sensors. If this is true (I
have no independent verification) you're not likely to eliminate much
error by using different starting points on the plate.
Note that you still need to turn direct and inverted in order to remove
axis error.
---------------------------------------------------------------------
Jim Frame jhf...@dcn.davis.ca.us (916) 756-8584 756-8201 (FAX)
Frame Surveying & Mapping 609 A Street Davis, CA 95616
-----------------------< Davis Community Network >-------------------
Paul Kunkel
"Neil Giesbrecht" <Nei...@netshop.net> writes:
>With modern instruments is turning the angles twice on opposite sides of
>the 'plate' necessary? In my case I own a Leica TC600 and turning angles
>(actually bearings) twice on opposite faces is basically a pain in the ass
>since the display is only on one side of the gun and it doesn't 'hold' the
>foresight to use as the subsequent backsight. Does the circuitry of these
>new guns make such a routine unnecessary?
If every moving part on an instrument were adjusted or compensated to
pinpoint precision, then it would make no difference. But if the
instrument could not go out of adjustment, then there would be no need for
a calibration procedure. It is not a perfect world. By using both faces,
we make that a moot point. The error, big or small, is turned against
itself. Without this procedure, we are taking a lot for granted.
I once test-ran a very inexpensive Leica total station at a
convention. That may have been the TC600. It appeared to be built for
staking and topo, the kind of work that requires little high-order
traversing. One-sided angles may be safer now that such things as
dual-axis compensation and stored error constants are available, but does
that instrument even have those features?
Kunkel
car...@aol.com
Turning direct and reverse angles is still necessary if your intention is
to get the best from your instrument as well as your instrument operator.
The best made instrument will often turn our bad results if the right
proceedures are bad.
Neil Giesbrecht
Thanks for the responses guys!
This instrument (Leica TC600) does have dual axis compensators and I do
keep the two collimation errors adjusted. Also it has a SD of 5". It is
best suited for topographic pickup which is what I do mostly. I will
continue with the standard procedure of turning the angles on the 2 faces
on traverse as it is also a good check on my booking. Until I rely entirely
on electronic data collection I will play it safe (unless I want to risk
having to return to occupy that point).
(I used the word 'circuitry' with hesitation for lack of a better word at
the time.)
Peter Ehlert
Neil Giesbrecht wrote:
>
> With modern instruments is turning the angles twice on opposite sides of
> the 'plate' necessary? In my case I own a Leica TC600 and turning angles
> (actually bearings) twice on opposite faces is basically a pain in the ass
> since the display is only on one side of the gun and it doesn't 'hold' the
> foresight to use as the subsequent backsight. Does the circuitry of these
> new guns make such a routine unnecessary?
>
"Modern instruments make VERY accurate measurements.
We no longer have a need for doubling."
--> I made that statement when I got my first Wild T1 in about 1973.
I was right then and am still right.
Doubling does NOT increase the _accuracy_ enough to warrant the effort.
The REAL reason to double is (and always has been) to eliminate People
Errors.
1. We ALL make transcription errors. It is Very common to swap numbers,
write unclearly, record the vertical instead of the horizontal, . . . .
The list is endless. "I am very careful and don't do that": BS, we are
human. "I use a data collector": not out of the woods yet!
2. We ALL make pointing errors. In the best of conditions we get
complacent and make silly, sloppy errors; in tough conditions we can be
mistaken. Pointing to the wrong sight can and does happen, pointing a
little off the target happens too.
A bad double tells you that you have an error someplace. This is
probably the only way the error can be found. Balancing a traverse does
no more than spread the error thru your Good work.
**>>> Don't Skip Doubling, you will screw up, guaranteed. <<<**
We are getting lost in technology and forgetting real life conditions.
----- (you may replace "doubling" for "redundant independent
measurement" at any point in this discussion) -----
--
Peter Ehlert mailto:pet...@sonic.net
Oooo.
.oooO ( )
( ) ) /
\ ( (_/
\_)
Sea Dance Internacional, SRLdeCV
Chris Hall
Neil Giesbrecht wrote:
>
> With modern instruments is turning the angles twice on opposite sides of
> the 'plate' necessary? In my case I own a Leica TC600 and turning angles
> (actually bearings) twice on opposite faces is basically a pain in the ass
> since the display is only on one side of the gun and it doesn't 'hold' the
> foresight to use as the subsequent backsight. Does the circuitry of these
> new guns make such a routine unnecessary?
We use a TC-600 at Algonquin College in Ottawa. We take angles on both
sides of the plate, and yes, it is indeed a pain in the ass. The sales
rep from Leica told us that a second keyboard and display could be
attached
to the other side of the unit. I'm already very impressed with the
TC600,
and a second keyboard would make it even better.
--
Humans are merely a life support system for their cats.
=======================================================
Chris J. Hall j...@synapse.net Ottawa ON
Canada
=======================================================
Chris Hall
Paul Kunkel wrote:
> I once test-ran a very inexpensive Leica total station at a
> convention. That may have been the TC600. It appeared to be built for
> staking and topo, the kind of work that requires little high-order
> traversing. One-sided angles may be safer now that such things as
> dual-axis compensation and stored error constants are available, but does
> that instrument even have those features?
As a side note, we were only able to accomplish a 1:27000 accuracy on a
13 angle, 2100m traverse. Then again, we are only students, however we
took 4 or 5 sets at each station and were out by 24" on the whole
traverse.
I don't know if that's good or bad for a total station that really isn't
made for traversing.
Peter & Carmel Spierings
In article <01bbd10a$d11f7b00$7046...@neilkg.netshop.net>, Nei...@netshop.net says...
>
>With modern instruments is turning the angles twice on opposite sides of
>the 'plate' necessary? In my case I own a Leica TC600 and turning angles
>(actually bearings) twice on opposite faces is basically a pain in the ass
>since the display is only on one side of the gun and it doesn't 'hold' the
>foresight to use as the subsequent backsight. Does the circuitry of these
>new guns make such a routine unnecessary?
>
From my experience in deformation and general control surveys, the use of
a Total Station does not relieve the Surveyor from the responsibility of
applying basic principles and precautions in carrying out any type of
survey. Many surveyors regard the Total Station as a magic wand and they
conveniently overlook the fact that at each set-up the theodolite has to
be fully calibrated anew. Only when this is done can we speak of
advantages of the Total Station.
The use of Least Squares in adjusting traverses is quite meaningless as
the redundancies in a traverse are very small in number. The use of
Least Squares assumes that, apart from small random errors in the
measurements, the number of redundancies is significantly large.
Nonetheless, many surveyors keep on using least squares to adjust
traverses. Bowditch would be much more preferable.
Paul Kunkel
Chris Hall <j...@synapse.net> writes:
>As a side note, we were only able to accomplish a 1:27000 accuracy on a
>13 angle, 2100m traverse. Then again, we are only students, however we
>took 4 or 5 sets at each station and were out by 24" on the whole
>traverse.
>I don't know if that's good or bad for a total station that really isn't
>made for traversing.
Let me suggest that it may be misleading. One major source of
theodolite error is eccentricity of the trunnion axis, the axis about
which the scope plunges. If the trunnion axis is not horizontal, then as
the scope is tilted, the line of sight will cut a plane which is not
vertical. This will lead to angular error, but it is quite possible to
get a deceitfully good closure in spite of it.
Suppose that the trunnion axis is skewed clockwise from the gunner's
perspective. Now suppose that a traverse is being run down a steep hill,
and that the instrument is not being inverted. All of the angles will
appear larger than they actually are, say 15" larger. If it takes six
stations to get to the bottom, that comes to 1'30". After traversing
through a few more stations at the bottom, the party heads back up the
hill. This time the observed angles are 15" too small, because the
conditions have been reversed. If they use six stations going up the
hill, then the angular error will have been completely canceled. That
is not a good traverse though.
Those who read from both sides and use the same instrument every day
may notice this effect when they reduce notes. The error will tend to be
greater on the steep shots, and its sign will depend on whether the
traverse is running uphill or downhill.
Kunkel
Kent McMillan
In article <56l715$4...@grissom.powerup.com.au>, di...@powerup.com.au (Peter & Carmel Spierings) says:
>
>The use of Least Squares in adjusting traverses is quite meaningless as
>the redundancies in a traverse are very small in number. The use of
>Least Squares assumes that, apart from small random errors in the
>measurements, the number of redundancies is significantly large.
>Nonetheless, many surveyors keep on using least squares to adjust
>traverses. Bowditch would be much more preferable.
>
In a typical closed traverse there are at least three additional measurements
(two angles and one distance) beyond what is minimally necessary to calculate
the co-ordinates of points positioned by the traverse. Plainly the fact that
there is a need to adjust the measurements at all means that some redundancy
exists in the system of survey measurements.
The use of the method of Least Squares simply requires that there be something
to adjust, one more measurement than is needed to determine the figure of the
survey, and some realistic estimates of the standard errors of the survey
measurements. I assume that there is no problem estimating the standard errors
of the measurements made in the course of a project, even if the manufacturer's
specifications are in doubt, since widely accepted procedures such as those of
DIN 18723 are available for that purpose.
In the back numbers of the British publication, "Survey Review", there is an
article upon the relative inadequacy of Bowditch's Rule (oddly known as the
Compass Rule in the U.S. although Nathaniel Bowditch was American), P. Berton
Jones' "A comparison of the precision of traverses adjusted by Bowditch Rule
and by Least Squares" (No.164, April, 1972).
Am I right in thinking that you used Least Squares in adjusting the deformation
and control surveys you mention? Is your aversion to Least Squares based in
part upon having used very awkward software? If so, I can certainly sympathize.
Really though, not having a decent computer and software is the only reason I can
think of for not using Least Squares adjustments on every project. Good programs
like Star*Net and Star*Lev are very easy to use and have very short run times,
unlike some old mainframe programs that were not and did not.
Kent McMillan, RPLS
Austin TX
Bryan Bunch
"Neil Giesbrecht" <Nei...@netshop.net> wrote:
>With modern instruments....
>[...]
> Does the circuitry of these new guns make such a routine[doubling]
> unnecessary?
I have the double angle discussion with a fellow land
surveyor at least once in any given year. My philosophy on
doubling has nothing to do with accuracy, as a well adjusted
and operated instrument will provide adequate precision and
accuracy for the majority of general surveys performed.
The effects of multiple angles on accuracy becomes academic
for most applications as error ratios of 50000 or greater
are routinely achieved on traverses with many legs and
perimeters less that 10000 feet.
The real reason to double angles is to provide an on the
spot check for measurement or recording errors. Instrument
precision will never eliminate the potential for human
errors and blunders. Taking multiple observations is the
only way to identify and minimize human error, as well as
identify an instrument that has gone out of adjustment (gee,
I had that total station adjusted last year).
There is not one good reason for not turning multiple angles
on traverse legs and other significant observations. The
"time is money" argument never washes with me. One busted
angle not identified on the spot in the field will destroy
the economic viability of any survey, and can ruin the
ability to meet tight deadlines, when time really is money.
The time savings is never worth the risk.
--
Bryan Bunch
bwb...@skn.net
--
Melton Wiggins
I agree absolutely. I want a full repetition of each angle and
distance. Usually I am recording data manually in a field book, and I
like to double check what I recorded the first time. I have never had to
rerun a traverse due to instrument error, but I have due to human error.
Melton Wiggins
artwo...@aol.com
>>With modern instruments is turning the angles twice on opposite sides of
>>the 'plate' necessary? In my case I own a Leica TC600 and turning angles
>>(actually bearings) twice on opposite faces is basically a pain in the
ass
>>since the display is only on one side of the gun and it doesn't 'hold'
the
>>foresight to use as the subsequent backsight. Does the circuitry of
these
>>new guns make such a routine unnecessary?
I believe for some surveys In my state (AR) I must double angles to keep
within minimum standards for boundary surveys. I never double angles
for topo shots, but always double for traverse legs. Getting right on tack
causes more error thad not doubling angles!
ART
DCI Little Rock
10809 Executive Center Dr. Sw. #210
Little Rock AR 72211-6021
Tony Sprent
Paul Kunkel wrote:
>
>
>
One major source of
> theodolite error is eccentricity of the trunnion axis, the axis about
> which the scope plunges. If the trunnion axis is not horizontal, then as
> the scope is tilted, the line of sight will cut a plane which is not
> vertical. This will lead to angular error, but it is quite possible to
> get a deceitfully good closure in spite of it.
> (snip)
> Those who read from both sides and use the same instrument every day
> may notice this effect when they reduce notes. The error will tend to be
> greater on the steep shots, and its sign will depend on whether the
> traverse is running uphill or downhill.
>
> Kunkel
In fact taking horizontal circle readings on both faces (transiting)
will not elliminate the effects of a tilted trunnion axis due to plate
bubble missalignment. It is for this reason that for steeply inclined
sights, such as for astronomical observations it is essential to take
plate bubble readings on both faces in order to adjust the observed
horizontal angle between the reference and the sun or star.
Never-the-less where horizontal and vertical angles are required which
are free of the other systematic errors of horizontal and vertical
collimation, circle eccentricity etc then it is essential to take face
left, face right obs unless the instrument has suitable software,
sensors and calibration techniques to compensate for them automatically.
I would also disagree with the respondent who states that a Bowdich
adjustment is better than a least squares adjustment. while it is true
that the redundency is small, a ls solution will always be better
statistically than any approximate adjustment. Mind you the difference
between the two solutions may well be so small that it does not matter.
In the past Bowdich was used because of the lack of suitable simple ls
adjustment programs but these days there are any number available which
are capable of solving virtually any combination of angular and distance
measurements.
__________________________________________________________
| Dr Tony Sprent |
| Department of Surveying & Spatial Information Science |
| University of Tasmania |
| GPO Box 252C Hobart _--_|\ |
| Tasmania* / \ |
| Australia 7001 \_.--._/ ) |
| * / |
| Phone 61-3-62262106 Fax 61-3-62240282 International |
| 03-62262106 03-62240282 Australia |
| email Tony....@surv.utas.edu.au |
| http://info.utas.edu.au/docs/geomatics/ |
|_________________________________________________________|
Sam King
I have always believed that angles should be "wrapped", as a check in the field, against errors, as well
as, obtaining a higher degree of accuracy. I also shoot all of my first shots, with the data collector,
but take hard notes, as well. Then, I "plunge" the scope, and measure another set of angles. I find, that
many times, the rodman will shift, or not quite be plumb yet, at the time the shot is taken. I prefer to
let the EDM read continuous, observing several distances, as a check, before I write the distance down.
If I solely used the data collector, which takes only one reading per shot, then I would introduce
distance errors, into my field data. The data collector file, then comes in handy, when a number was
written down wrong, in the hard notes. Yes, I am aware that the data collector can record a variety of
angle turning methods, and take multiple distance shots, if you re-run an entire cycle, but I'm in the
business to make a living, and I don't have time to spend two or three times as long, waiting for all of
those shots to be taken. Does anyone else traverse this way? Does anyone else have any ideas for fast
and efficient traversing? Thank you for your input.
--
Sam King
rls...@iland.net
Jim Frame
In article <N.112296....@wbrg119.internetland.net>,
rls...@internetland.net (Sam King) wrote:
>I have always believed that angles should be "wrapped", as a check in
the field, against errors, as well
>as, obtaining a higher degree of accuracy. I also shoot all of my first
>shots, with the data collector, but take hard notes, as well.
[snip]
>those shots to be taken. Does anyone else traverse this way? Does
>anyone else have any ideas for fast and efficient traversing?
When I first began using a data collector I found it too cumbersome for
control traversing, so used a field book instead and relegated the data
collector to topo shots. As my level of comfort increased, I abandoned
that practice, and now record all traverse data with the data collector.
(This is in spite of the fact that the TDS routine for multiple angles
requires BS-FS-BS-FS rather than BS-FS-FS-BS, a particularly annoying
"feature.")
The value of having all the data in a single digital file outweighs the
disadvantage of being able to quickly view the individual measurements,
and I have come to trust the data collector to warn me about
unacceptable measurements.
Bryan Bunch
rls...@internetland.net (Sam King) wrote:
> I have always believed that angles should be "wrapped",
> as a check in the field, against errors, as well as,
> obtaining a higher degree of accuracy. I also shoot all
> of my first shots, with the data collector, but take hard
> notes, as well. Then, I "plunge" the scope, and measure
> another set of angles. I find, that many times, the
> rodman will shift, or not quite be plumb yet, at the time
> the shot is taken. I prefer to let the EDM read continuous,
> observing several distances, as a check, before I write the
> distance down. If I solely used the data collector, which
> takes only one reading per shot, then I would introduce
> distance errors, into my field data. The data collector file,
> then comes in handy, when a number was written down wrong,
> in the hard notes. Yes, I am aware that the data collector
> can record a variety of angle turning methods, and take
> multiple distance shots, if you re-run an entire cycle, but
> I'm in the business to make a living, and I don't have time
> to spend two or three times as long, waiting for all of
> those shots to be taken. Does anyone else traverse this way?
> Does anyone else have any ideas for fast and efficient
> traversing? Thank you for your input.
Please keep in mind that I am writing based on my own
limited experiences.
I believe that most instruments have settings that allow you
to specify the number of distance readings taken for each
shot before the average distance is displayed or
transmitted. This setting may be in the form of a simple
"coarse" or "fine" switch. I generally use coarse mode for
non-control work and fine mode for critical measurements.
The coarse mode usually completes the shots in half the time
for fine mode shots and is great for general topo and
planimetric shots. I am quite satisfied with either setting
as far as the desired accuracy of my measurements.
Since I always double for traverse legs and other important
observations, I measure the distance both direct and reverse
and let the data collector inform me of the difference
between the two (or more) shots. The amount of time
required to take the reversed distance measurement is
minimal (a few seconds). I even have the option of deleting
the offensive observation from the averaged measurements if
a blunder has been discovered.
A great majority of the time, the difference in distance
measurements are not enough to warrant further shots.
Although there are rare occations when the distance error
warrants at least one additional measurement, I find that I
have to take additional observations to tighten angular
accuracy more often than to correct distance for a jittery
rodman.
I also used both methods of recording observations
(hand-written and data collector) when I first acquired a
data collector. Now that I have modified my surveying
methods to conform with the quirks of using a data
collector, I only use hand written notes for sketches and
other important notes not readily entered into the data
collector. I am quite pleased with the traversing
efficiency my data collector has allowed me to achieve. The
note keeper (usually me) is no longer the time efficency
bottleneck in my field surveys.
Any proven practical method of traversing is acceptable
regardless of the technology used. The most important
aspect is that the surveyor in charge understand the method
used, and is satisfied it provides the accuracies required
for the project.