Following up on another thread wherein I was trying to wrap my head around the entire subject of "Adjustment of Compensation of Systematic Error of Instrument" on page 17-10 of my Topcon GTS 255 manual, I did a more precise observation of the horizontal angle changing as a result of vertical angle travel. Here's what I got; haven't plotted them yet because I really don't know what to make of the numbers.
Process: Leveled instrument reasonably well, using the vial only, on a solid stand. Both X/Y compensation was OFF (Although there appeared to be very little difference in these numbers with it on). All but the ones at V=0 and 10 were within a few seconds. V=90 is telescope horizontal.
[pre]
Vert Horiz
90 00 00 00
80 359 59 53
70 359 59 46
60 359 59 39
50 59 59 30
40 359 59 19
30 359 59 03
20 359 58 33
10 359 57 08
0 359 32 20
350 00 02 33
340 00 01 07
330 00 00 37
320 00 00 21
310 00 00 10
300 00 00 01
290 359 59 54
280 359 59 47
270 359 59 40
[/pre]
The "Constant List" in the instrument at the time I did this test is as follows: VCo = 0 23' 23"; HCo = 0 00' 09"; HAx = 0 00' 38".
I really wish someone could find the hand drawn sketch I posted ages ago that was a "stick figure" of a total station, identifying the axes where these angles are being measured, or post one themselves. That might lend a clue to what the numbers I've recorded mean.
Those constants seem grossly excessive.
The correction process is iterative.
You may have to do it several times (or more in this case), accepting the values of each iteration, until they reduce to just several seconds.
At least that's the way I do it in my Leicas.
After seeing those numbers, your scope is adrift from alignment by nearly 3min.
I would imagine that is beyond the range of the onboard digital correction.
That can be attributed to something as simple as the case being mounted around the scope improperly or a structural problem with the standards that the scope mounts.
My first conversation would be with a tech at the tuneup shop.
Make sure your level vial is calibrated before doing any other adjustments. You probably know the routine, but just in case ...
Level the instrument by the vial, turn 90 degrees, re-level. Then turn 180 degrees more and take out half the error with the tribrach screws and half with the adjustment to the vial itself. Repeat until the bubble is centered as you rotate the instrument.
Alternatively, use the tribrach screws until you can get the same (probably uncentered) bubble reading at any rotation of the instrument, and then adjust the vial to read level.
A Harris, post: 370006, member: 81 wrote: After seeing those numbers, your scope is adrift from alignment by nearly 3min.
I would imagine that is beyond the range of the onboard digital correction.
That can be attributed to something as simple as the case being mounted around the scope improperly or a structural problem with the standards that the scope mounts.
My first conversation would be with a tech at the tuneup shop.
If that's the case, why, in the very manual, would they show an example of results (constants) that are way greater than mine?
I don't think these have anything to do with the "tilt compensation" system. I think these have to do with the difference between the optical and mechanical axes of the system.
Bill93, post: 370007, member: 87 wrote: Make sure your level vial is calibrated before doing any other adjustments. You probably know the routine, but just in case ...
Level the instrument by the vial, turn 90 degrees, re-level. Then turn 180 degrees more and take out half the error with the tribrach screws and half with the adjustment to the vial itself. Repeat until the bubble is centered as you rotate the instrument.
Alternatively, use the tribrach screws until you can get the same (probably uncentered) bubble reading at any rotation of the instrument, and then adjust the vial to read level.
I'll do that. I've noticed that the bubble may be very slightly off (less than half a graduation...perhaps a third).
Keep in mind, I'm not doing this because I think there's something wrong with the instrument. My recent solars (Multiple D&R) prove that there's probably not. I'm doing it to find out what the internal compensation system is doing.
First and foremost: the vial. (Just like putting crosshair on a target. A little to one side is a miss.)
You get the bubble dead center, rotate 180, and if it's to one side, at all, move the bubble 1/2 way back. Repeat until the bubble stays in the same location, relative to a mark. And turn 90, repeat. Absolutely no perceptible movement. only then is the vertical axis vertical.
And you should be able to adjust the vial (adjusting the vial by its own adjusting screws, not with the tribrach screws) so that the bubble resides in the center, but that's optional.
If the bubble is a 1/4 mark towards the vertical circle side of the instrument, for instance, and you rotate 45, 135, it has to remain in the exact same position (relative to a mark, and the instrument chassis).
rfc, post: 369993, member: 8882 wrote: Following up on another thread wherein I was trying to wrap my head around the entire subject of "Adjustment of Compensation of Systematic Error of Instrument" on page 17-10 of my Topcon GTS 255 manual, I did a more precise observation of the horizontal angle changing as a result of vertical angle travel. Here's what I got; haven't plotted them yet because I really don't know what to make of the numbers.
Process: Leveled instrument reasonably well, using the vial only, on a solid stand. Both X/Y compensation was OFF (Although there appeared to be very little difference in these numbers with it on). All but the ones at V=0 and 10 were within a few seconds. V=90 is telescope horizontal.
[pre]
Vert Horiz
90 00 00 00
80 359 59 53
70 359 59 46
60 359 59 39
50 59 59 30
40 359 59 19
30 359 59 03
20 359 58 33
10 359 57 08
0 359 32 20
350 00 02 33
340 00 01 07
330 00 00 37
320 00 00 21
310 00 00 10
300 00 00 01
290 359 59 54
280 359 59 47
270 359 59 40
[/pre]The "Constant List" in the instrument at the time I did this test is as follows: VCo = 0 23' 23"; HCo = 0 00' 09"; HAx = 0 00' 38".
I really wish someone could find the hand drawn sketch I posted ages ago that was a "stick figure" of a total station, identifying the axes where these angles are being measured, or post one themselves. That might lend a clue to what the numbers I've recorded mean.
Hey rfc , assuming you levelled the instrument using the correct procedure as described by others, it looks like you've got a couple of things going on. First it looks like the trunion (telescope) axis is not perpendicular to the standing or vertical axis of your instrument. That's why you're reading about 3 minutes off at steep vertical angles. I don't know how you measured a direction to something at VA=0 tho. Secondly it looks like your crosshairs are out of alignment with the line of collimation of the telescope which is the difference in readings between FL and FR.
What was your test procedure? the job of the instrument calibrations constants is to prevent these kinds of differences when reading targets at different VA values. If there's not much difference then they ain't doing their jobs. Did u just turn the telescope gradually upwards without sighting something? That would make the results make more sense.
Conrad:
As for this test procedure, I sighted nothing. i leveled the scope, set the horizontal and vertical to zero (just by looking at the readout), and began to move the scope upward. I have recently (but not immediately prior to this informal test), set Vertical Datum=0, and done the "Adjustment of Compensation of Systematic Error" per the manual. I can print the whole page that describes that, but basically it has you sight something within 10 degrees of horizontal an equal number of times in each face, then something greater than 30 degrees in both faces; it logs the "Constants". I've done it multiple times and the numbers vary a bit, but (as someone suggested that this might be iterative), they've never appeared to go lower or approach zero.
I have never done the crosshair collimation adjustment but will at least check that next. I'm not sure anything can be determined about that from this test though, because I didn't use the crosshairs. Never looked through the scope. Still thinking about all this. Thanks for the input.
Well your procedure explains why you are getting those readings that go away from 0 as you wind the scope upwards or downwards. Modern total stations are able to utilise electronic compensation of angles rather than relying on the machining of perfectly square standing and telescope axes. Unless you happened to score a perfectly constructed total station you will need to keep turning the instrument to read 000 00 00 as you turn upwards or downwards if you want to trace a vertical line.
Because your telescope axis is not square to your standing axis by some 23'23" the instrument is correctly adjusting and displaying the new azimuth relative to the horizontal zero you set. This is normal behaviour. Adding or subtracting from this effect to a lesser degree is the misalignment of the crosshairs from a line perpendicular to the telescope axis. A smaller effect at 9" misalignment, but it affects the azimuth more as you tilt the telescope away from the horizontal.
The calibration values aren't supposed to be zero. They are meant to correctly reflect the current mechanical variation from perfectly square axes, and are determined by running through the procedure in the manual.
Maybe related..,
If the cal procedure is to repeatedly measure a hor angle with a BS that is level and a FS that is elevated, 15-25å¡, D/R, small angles like 2å¡ and near 90å¡, then both sights have be far enough away so that 1-2" is optically repeatable. If the sights are near, then 10" seconds of observational noise may return an artificially exaggerated compensation. i.e. it's averaging +/- 10", or more, of pointing error, not instrument error, and the arithmetic can't distinguish pointing err from instrument err. If the sights are far, or laboratory grade, then pointing errors should be smaller than the variation in the instrument. V and H. Calibration observations have to have lowest possible noise to reveal instrument induced err.
Faulty cal observations may return offsets that aren't realistic, and not repeatable.
IMO
Conrad, post: 370055, member: 6642 wrote:
Modern total stations are able to utilise electronic compensation of angles rather than relying on the machining of perfectly square standing and telescope axes.Because your telescope axis is not square to your standing axis by some 23'23" the instrument is correctly adjusting and displaying the new azimuth relative to the horizontal zero you set.
Adding or subtracting from this effect to a lesser degree is the misalignment of the crosshairs from a line perpendicular to the telescope axis. A smaller effect at 9" misalignment, but it affects the azimuth more as you tilt the telescope away from the horizontal.
What is a "standing axis"? The Trunion axis? And what's the "telescope axis"? These must be what VCo = 0 23' 23" is recording the difference between.
Isn't the telecope mounted on the trunion? How can it be out of whack?
rfc, post: 370117, member: 8882 wrote: What is a "standing axis"? The Trunion axis? And what's the "telescope axis"? These must be what VCo = 0 23' 23" is recording the difference between.
Isn't the telecope mounted on the trunion? How can it be out of whack?
The standing or vertical axis is the axis about which the body of the total station rotates around horizontally. The telescope, tilting or trunion axis is the axis about which the telescope spins. If these were perfectly square then no correction would be necessary. But as they usually aren't perfectly square then the total station must mathematically correct for the systematic error using the values derived during the calibration procedure. Now it's been a while since I used a topcon so I don't know which one represents the magnitude of the non-perpendicularity of the two axes but a quick calculation I did shows 23'23" would give corrections of around the magnitude that you are seeing. At VA of 20 40 60 and 80 degrees, I calculated corrections of 9", 21", 42" and 139" respectively.
As for how it can be out of whack, well, a good whack might do it?!? or it was manufactured to that tolerance, or it could have been poorly serviced if they have the facility to slide the telescope bearings up or down. If the bearings holding your total station are say 100mm apart then one being mounted 0.68mm higher would give you 23'23" of tilting axis misalignment. Though that kind of tolerance seems really high for any total station manufacturer.
Just to reiterate, the values are what they are, and if they were all zero then you either scored a perfectly manufactured and adjusted instrument, or more likely I'd suspect someone did a factory reset of the instrument and the values need to be determined for the first time since the reset. Doing the calibration several times doesn't get the values closer to zero, it will just give you several results around the same numbers and a sense of the precision of your results.
Conrad, post: 370133, member: 6642 wrote: Just to reiterate, the values are what they are, and if they were all zero then you either scored a perfectly manufactured and adjusted instrument, or more likely I'd suspect someone did a factory reset of the instrument and the values need to be determined for the first time since the reset. Doing the calibration several times doesn't get the values closer to zero, it will just give you several results around the same numbers and a sense of the precision of your results.
Thank you. That's a great help. If I forgot to turn ON compensation after this test and then went out and did two solars, back to back about an hour apart, could the 18" seconds I'm seeing between the first and second be explained by this? The sun was higher in the sky by 10 degrees or so (I'll check MICA).
rfc, post: 370142, member: 8882 wrote: Thank you. That's a great help. If I forgot to turn ON compensation after this test and then went out and did two solars, back to back about an hour apart, could the 18" seconds I'm seeing between the first and second be explained by this?
Yes it is possible. Even if you just turned off the level compensator and not the mechanical compensation, that may be enough to account for the discrepancy once added to a bit of timing and aiming error one has to deal with in solar observations.
Conrad, post: 370145, member: 6642 wrote: Yes it is possible. Even if you just turned off the level compensator and not the mechanical compensation, that may be enough to account for the discrepancy once added to a bit of timing and aiming error one has to deal with in solar observations.
I'm getting both the timing and aiming down great; just need to make sure the instrument is performing to it's published capacity. Thanks again.
Timing errors are so easily over these days, that there's no reason for timing errors beyond 0.05". And you can test hand-eye-coordination errors, and they're random errors that are pretty well reduced by averaging.
