Hello,
I'm looking at an alternative way to assess the angular accuracy/precision of my total station. I say alternative because I can't be bothered taking a bunch of gear to an open space on the weekend. I was willing, however, to take an instrument home and sight some targets stuck to the wall. I wanted to assess the standard deviation of the observed angles between my tightly spaced targets over different parts of the circle. I managed to sample a cumulative 76 degrees of the circle, ~6 degrees at a time, in 14 minute increments from 3 areas corresponding to the positions of the instrument lugs.
Now can anybody think of a reason (apart from circle eccentricity which is allegedly taken care of by double circle sampling in this instrument) why analysis of bigger angles between targets spaced around the horizon would give a 'better' assessment of the total station accuracy than my heavy sampling of the circle in ~6 degree increments at a time?
If it's deemed to be an ok method then i'll post the rounds and I'll invite you to run the angles through excel (or whatever) to confirm my analysis and guess the instrument specs. Or if my method sucks then I'll forget I ever went through the exercise!
Cheers.
>I wanted to assess the standard deviation of the observed angles between my tightly spaced targets over different parts of the circle. I managed to sample a cumulative 76 degrees of the circle, ~6 degrees at a time, in 14 minute increments from 3 areas corresponding to the positions of the instrument lugs.
I'm not quite understanding the procedure you've described. Could you elaborate a bit, particularly as to how circle was rotated without altering the instrument centering?
At first impression, I'd think that measuring the same small angle multiple times wouldn't really capture the effect of the quality of the machining of the instrument center. At least, I'd guess that the centers are more uniform over small arcs than they are over the entire 360 degrees that the instrument turns.
You need to look at the positional tolerances needed for your test. What is the expected pointing accuracy of the instrument? What working distance can you get?
For instance, a 5 second change in angle at 10 feet is 0.003 inch and I don't think you can rely on that kind of stability even with the best tripod.
When Shooting At Close Tagets +/-5" Is Optimisic.
Distant targets are used to discern some differences in angles.
Paul in PA
I disagree
With my T3's or T3000's I can turn sub-second angles time after time in my basement at targets about 20' away. Now I will grant you that those are ideal conditions and my eye gets attuned to the gun & targets after a while, but it can be done. In fact most of the time my angles will come up sub-half-second.
Expain Your Targets To Conrad
I doubt he is using what you use.
Paul in PA
>
> I'm not quite understanding the procedure you've described. Could you elaborate a bit, particularly as to how circle was rotated without altering the instrument centering?
>
> At first impression, I'd think that measuring the same small angle multiple times wouldn't really capture the effect of the quality of the machining of the instrument center. At least, I'd guess that the centers are more uniform over small arcs than they are over the entire 360 degrees that the instrument turns.
Hello Kent,
After the first two sets I'd release the instrument from the tribrach, rotate 120 deg and drop it into the next lug position. After the third I'd rotate the instrument on the plate so that it was now pointing back at the first target, I'd re-centre as best I could and go again in each of the three lug positions. This way I never used the same part of the circle.
So your concern is that I wouldn't get to see the full effect of things like short and long periodic circle gradation errors, and even though I was only 2.4 metres from my target you'd expect to see an unrealistically good estimate of the accuracy of the instrument?
Cheers.
> You need to look at the positional tolerances needed for your test. What is the expected pointing accuracy of the instrument? What working distance can you get?
>
> For instance, a 5 second change in angle at 10 feet is 0.003 inch and I don't think you can rely on that kind of stability even with the best tripod.
Hello Bill,
I got 2.4 metres distant from my target. The pointing accuracy Is one of the things I was trying to determine. If nothing else this should be the only sure thing that should come of this exercise.
So with any tripod you'd expect 'poor' results from my test?
Cheers.
Expain Your Targets To Conrad
> I doubt he is using what you use.
>
> Paul in PA
Hello Paul,
And this to me was half the fun of the exercise. I don't know If anybody wanted to play ball but I was hoping to provide a starnet input file or similar and people could use it to guess the quality of the instrument I was using, without the prejudice of knowing what I was using.
Expain Your Targets To Conrad
> I was hoping to provide a starnet input file or similar and people could use it to guess the quality of the instrument I was using, without the prejudice of knowing what I was using.
I'm game. Sounds like a good Starnet exercise. And since you said "guess", I wouldn't even need to be an expert!
What Do You Use For Precise Targets
Possibly more important than the instrument at short range.
Paul in PA
> After the first two sets I'd release the instrument from the tribrach, rotate 120 deg and drop it into the next lug position. After the third I'd rotate the instrument on the plate so that it was now pointing back at the first target, I'd re-centre as best I could and go again in each of the three lug positions. This way I never used the same part of the circle.
>
> So your concern is that I wouldn't get to see the full effect of things like short and long periodic circle gradation errors, and even though I was only 2.4 metres from my target you'd expect to see an unrealistically good estimate of the accuracy of the instrument?
I've never tested the centering errors associated with releasing an instrument and replacing it in the tribrach with a different set of lugs engaged. The literature I have (M.A.R. Cooper's "Modern Theodolites and Levels" 2nd Ed Pg. 158), suggests a centering error on the order of 0.006mm might be realistic for well-made equipment. At a range of 2.5m, that would result in an error at most of 00'00.5", so the method you describe may not be completely unfeasible due to centering errors.
What is the type of target that you're using at 2.5m?
Expain Your Targets To Conrad
>I don't know If anybody wanted to play ball but I was hoping to provide a starnet input file or similar and people could use it to guess the quality of the instrument I was using, without the prejudice of knowing what I was using.
That exercise would be straight forward enough. There would have to be some assumptions made about the centering error of replacement in the tribrach, of course. If there is any guess work involved, my guess right now is that the results will tend to indicate a smaller standard error of directions than probably is realistic, not the other way around.
>
> I've never tested the centering errors associated with releasing an instrument and replacing it in the tribrach with a different set of lugs engaged. The literature I have (M.A.R. Cooper's "Modern Theodolites and Levels" 2nd Ed Pg. 158), suggests a centering error on the order of 0.006mm might be realistic for well-made equipment. At a range of 2.5m, that would result in an error at most of 00'00.5", so the method you describe may not be completely unfeasible due to centering errors.
>
>
> What is the type of target that you're using at 2.5m?
Kent,
I'm using an adhesive backed A4 label with target stripes laser printed vertically at the right width to be able to 'pinch' the target with the double vertical hairs. It's on a particle board wall which should have a low expansion due to thermal effects.
What Do You Use For Precise Targets
Paul,
I'm using laser printed targets on an adhesive A4 just wide enough to 'pinch' with the double hairs.
Expain Your Targets To Conrad
>
> If there is any guess work involved, my guess right now is that the results will tend to indicate a smaller standard error of directions than probably is realistic, not the other way around.
Kent,
are you guessing orders of magnitude like a 10" to 1" change or more like 3" to 1". Is this just gut feel?
Would this thread be messed up too much If I just dumped the whole 13 rounds to 26 targets right here in starnet format? Is there a better way?
> I'm using an adhesive backed A4 label with target stripes laser printed vertically at the right width to be able to 'pinch' the target with the double vertical hairs. It's on a particle board wall which should have a low expansion due to thermal effects.
The label is on some non-hygroscopic material? I'm game to run it all through Star*Net to see what it looks like.
Expain Your Targets To Conrad
> are you guessing orders of magnitude like a 10" to 1" change or more like 3" to 1". Is this just gut feel?
It's just an intuition. My expectation is that the results would optimistic by less than a factor of 2, i.e. apparent s.e. not less that 0.50 x actual s.e. per DIN test procedure.
> Would this thread be messed up too much If I just dumped the whole 13 rounds to 26 targets right here in starnet format? Is there a better way?
I don't see any reason not to try to post it here so that anyone interested can just cut and paste the input file. I wouldn't use the "pre" tag because I think it makes cut and paste more difficult as carriage returns need to be inserted. May be wrong on that point.
> I've never tested the centering errors associated with releasing an instrument and replacing it in the tribrach with a different set of lugs engaged. The literature I have (M.A.R. Cooper's "Modern Theodolites and Levels" 2nd Ed Pg. 158), suggests a centering error on the order of 0.006mm might be realistic for well-made equipment.
John Greenwood at Fermilab published a paper in 1995 or so on high-precision measurement in the accelerator tunnel using a Geodimeter 640 and associated equipment. His remarks about tribrachs:
> Our finding concerning the Wild tribrach is that the centering from one insertion of a device to the next insertion of that same device in the same tribrach, is not likely to be better than 0.15mm. This means that even though we are using the Wild NL to center over the point, when the NL is removed from that tribrach and any other tribrach-adapted device is mounted, it will not repeat mechanically better than this figure.
I hope I got this right - File part 1
# Input Field File Survey
.Units METERS
.Units DMS
.Order AtFromTo
.Separator
.Delta Off
.2D
.SCALE 1.000000000000
# Job : TEST
# Date : 11/16/2014
# Time : 13:29:19
C S1 0.000 0.000 0.000 !!!
B S1-1 0-00-00 !
DB 0
DM S1 1 18-15-25.35 2.400
DM S1 2 18-29-35.04 2.400
DM S1 3 18-43-48.90 2.400
DM S1 4 18-58-10.65 2.400
DM S1 5 19-12-22.62 2.400
DM S1 6 19-26-29.46 2.400
DM S1 7 19-40-33.08 2.400
DM S1 8 19-54-40.96 2.400
DM S1 9 20-08-52.45 2.400
DM S1 10 20-23-03.72 2.400
DM S1 11 20-37-14.62 2.400
DM S1 12 20-51-29.20 2.400
DM S1 13 21-05-38.06 2.400
DM S1 14 21-19-42.55 2.400
DM S1 15 21-33-48.45 2.400
DM S1 16 21-47-53.40 2.400
DM S1 17 22-02-01.14 2.400
DM S1 18 22-16-11.54 2.400
DM S1 19 22-30-29.86 2.400
DM S1 20 22-44-39.16 2.400
DM S1 21 22-58-45.69 2.400
DM S1 22 23-12-51.60 2.400
DM S1 23 23-26-52.10 2.400
DM S1 24 23-40-52.13 2.400
DM S1 25 23-54-57.24 2.400
DM S1 26 24-09-05.85 2.400
DE
DB
DN S1 1 258-15-04.91
DN S1 2 258-29-13.74
DN S1 3 258-43-29.68
DN S1 4 258-57-47.92
DN S1 5 259-12-02.57
DN S1 6 259-26-09.31
DN S1 7 259-40-13.01
DN S1 8 259-54-21.28
DN S1 9 260-08-32.45
DN S1 10 260-22-43.40
DN S1 11 260-36-53.96
DN S1 12 260-51-08.37
DN S1 13 261-05-16.70
DN S1 14 261-19-21.34
DN S1 15 261-33-28.49
DN S1 16 261-47-34.17
DN S1 17 262-01-42.39
DN S1 18 262-15-54.05
DN S1 19 262-30-10.47
DN S1 20 262-44-19.61
DN S1 21 262-58-26.72
DN S1 22 263-12-32.09
DN S1 23 263-26-34.16
DN S1 24 263-40-34.46
DN S1 25 263-54-39.72
DN S1 26 264-08-46.02
DE
DB
DN S1 1 138-16-15.07
DN S1 2 138-30-24.74
DN S1 3 138-44-39.78
DN S1 4 138-58-58.91
DN S1 5 139-13-12.89
DN S1 6 139-27-19.03
DN S1 7 139-41-22.54
DN S1 8 139-55-30.41
DN S1 9 140-09-40.98
DN S1 10 140-23-52.07
DN S1 11 140-38-03.34
DN S1 12 140-52-18.09
DN S1 13 141-06-27.68
DN S1 14 141-20-31.98
DN S1 15 141-34-37.94
DN S1 16 141-48-42.99
DN S1 17 142-02-50.06
DN S1 18 142-17-02.30
DN S1 19 142-31-17.93
DN S1 20 142-45-28.09
DN S1 21 142-59-34.13
DN S1 22 143-13-38.67
DN S1 23 143-27-41.20
DN S1 24 143-41-41.38
DN S1 25 143-55-45.85
DN S1 26 144-09-52.68
DE
DB
DN S1 1 144-14-35.99
DN S1 2 144-28-46.97
DN S1 3 144-43-00.55
DN S1 4 144-57-20.61
DN S1 5 145-11-34.55
DN S1 6 145-25-39.70
DN S1 7 145-39-43.01
DN S1 8 145-53-50.95
DN S1 9 146-08-01.90
DN S1 10 146-22-13.39
DN S1 11 146-36-24.78
DN S1 12 146-50-39.35
DN S1 13 147-04-48.37
DN S1 14 147-18-51.33
DN S1 15 147-32-57.53
DN S1 16 147-47-02.65
DN S1 17 148-01-10.48
DN S1 18 148-15-21.40
DN S1 19 148-29-37.23
DN S1 20 148-43-47.34
DN S1 21 148-57-54.03
DN S1 22 149-11-59.14
DN S1 23 149-26-01.64
DN S1 24 149-40-01.30
DN S1 25 149-54-05.82
DN S1 26 150-08-12.07
DE
DB
DN S1 1 150-11-34.67
DN S1 2 150-25-45.63
DN S1 3 150-40-01.36
DN S1 4 150-54-18.58
DN S1 5 151-08-33.02
DN S1 6 151-22-39.13
DN S1 7 151-36-43.08
DN S1 8 151-50-50.94
DN S1 9 152-05-02.24
DN S1 10 152-19-12.89
DN S1 11 152-33-24.39
DN S1 12 152-47-38.49
DN S1 13 153-01-47.29
DN S1 14 153-15-51.45
DN S1 15 153-29-59.32
DN S1 16 153-44-04.25
DN S1 17 153-58-09.84
DN S1 18 154-12-22.55
DN S1 19 154-26-37.68
DN S1 20 154-40-47.85
DN S1 21 154-54-54.37
DN S1 22 155-08-59.25
DN S1 23 155-23-01.50
DN S1 24 155-37-01.44
DN S1 25 155-51-05.01
DN S1 26 156-05-12.06
DE
DB
DN S1 1 270-10-27.66
DN S1 2 270-24-38.26
DN S1 3 270-38-52.67
DN S1 4 270-53-10.66
DN S1 5 271-07-26.06
DN S1 6 271-21-30.86
DN S1 7 271-35-34.83
DN S1 8 271-49-43.67
DN S1 9 272-03-54.33
DN S1 10 272-18-05.64
DN S1 11 272-32-15.59
DN S1 12 272-46-29.75
DN S1 13 273-00-39.15
DN S1 14 273-14-43.55
DN S1 15 273-28-51.26
DN S1 16 273-42-55.12
DN S1 17 273-57-01.86
DN S1 18 274-11-12.26
DN S1 19 274-25-29.90
DN S1 20 274-39-39.84
DN S1 21 274-53-45.80
DN S1 22 275-07-50.94
DN S1 23 275-21-52.25
DN S1 24 275-35-52.79
DN S1 25 275-49-57.53
DN S1 26 276-04-04.05
DE
DB
DN S1 1 30-11-04.97
DN S1 2 30-25-13.99
DN S1 3 30-39-30.35
DN S1 4 30-53-48.46
DN S1 5 31-08-02.55
DN S1 6 31-22-08.65
DN S1 7 31-36-12.31
DN S1 8 31-50-20.57
DN S1 9 32-04-31.35
DN S1 10 32-18-42.49
DN S1 11 32-32-53.79
DN S1 12 32-47-08.37
DN S1 13 33-01-17.79
DN S1 14 33-15-21.92
DN S1 15 33-29-28.13
DN S1 16 33-43-33.75
DN S1 17 33-57-41.53
DN S1 18 34-11-52.40
DN S1 19 34-26-08.87
DN S1 20 34-40-19.38
DN S1 21 34-54-25.54
DN S1 22 35-08-31.42
DN S1 23 35-22-31.94
DN S1 24 35-36-32.12
DN S1 25 35-50-37.40
DN S1 26 36-04-44.57
DE