A bit more info.
Because I got a spidey sense from my first two traverses with the 1sec, I reran both instruments over a traverse at the same time.
5 stations, each had a tripod and tribrach setup (so we never had to transfer tribrachs or legs).
Professional grade Leica circular prisms.
We'd use the 1 second instrument, then take it off and shoot the same set with the 3 second.
4 sets on each. left and right face.
3" instrument 1:178,000 or so.
1" instrument 1:78,000.
It is. Both instruments were calibrated in the last month. Both turn left and right face about as tight as possible. ATR results check with optical results. (just to ensure no ATR error)
As per my above post.
Because I got a spidey sense from my first two traverses with the 1sec, I reran both instruments over a traverse at the same time.
5 stations, each had a tripod and tribrach setup (so we never had to transfer tribrachs or legs).
Professional grade Leica circular prisms.
We'd use the 1 second instrument, then take it off and shoot the same set with the 3 second.
4 sets on each. left and right face.
3" instrument 1:178,000 or so.
1" instrument 1:78,000.
I like the parasol idea, but it's usually not suitable for parasols where I am. I try and cycle the robot from left face to right face regularly while I'm getting things setup to make sure it's evenly warmed, rather than cooked on one side.
If you know what you're looking for you can tell if it's cooked too.
We also won't run traverses that require high accuracy in the middle of a hot day either. We'll go out early so there's little direct heating.
4 sets by ATR, not by hand.
Look For A Circle Error
Set up a 3 point equilateral traverse. Set up 3 tripods, 2 with prisms and do all your work at the third. Using only 1 tribrach, set and level, then pencil mark the tribrach outline on the tripod head. Do a 4D 4R set, BSD is to left of instrument, angle right to FSD, flop scope, angle right, BSR, FSR, angle right, BSR, angle right, FSR, flop scope, angle right, BSD, angle right FSD, 2D2R complete, repeat.
Redo, 4D 4R but always turn angle left. What you are looking for is a skip in the circle count.
Loosen tribrach and turn on tripod head 60° insided pencil line, relevel, redo two sets, then repeat tribrach rotation. you may be able to isolate an area of circle error.
Just because these are modern instruments does not mean they are not subject to some old fashioned errors.
Paul in PA
Look For A Circle Error
Thanks Paul.
I was thinking of doing something very similar to that. I'll use your suggestion.
I should be getting Leica's documentation on their check methods in the next day or two as well.
Check The Verticle Axis
Darn few traverses are level, so check the vertical axis. Set a point far enough from a utility pole that you can see the top with 75° zenith angle. Using the good gun sight a bolt head near the top, set zero and bring the gun down to level, pay attention if and how much the horizontal angle is to get back to zero, set a nail, flop the scope sight the nail, rezero, raise the scope to the bolt and make notes. Repeat 3 times, then use the new gun.
Seldom is the vertical axis perfect, but there is internal software to compensate for it.
You should be able to determine the vertical axis error and if it is being properly compensated.
D&Rs should compensate those errors out but you never know until you check.
Paul in PA
We'd use the 1 second instrument, then take it off and shoot the same set with the 3 second.
4 sets on each. left and right face.
3" instrument 1:178,000 or so.
1" instrument 1:78,000.
What are the differences at each station?
I may have missed something but others pointed out relative differences can be misleading in terms of actual close ratios.
Just out of interest what happens if you keep repeating your observations and use different backsight settings?
Electronics are supposed to take care of ALL our problems but they still have to read the circles and I would expect they are just as prone to errors as in the old days of maunal reading circles.
> I like the parasol idea, but it's usually not suitable for parasols where I am. I try and cycle the robot from left face to right face regularly while I'm getting things setup to make sure it's evenly warmed, rather than cooked on one side.
>
> If you know what you're looking for you can tell if it's cooked too.
>
> We also won't run traverses that require high accuracy in the middle of a hot day either. We'll go out early so there's little direct heating.
Hi CSS,
My reply was directed at Eric's scenario as to why his situation could've occurred. I don't think you're getting the wild splits he was getting. BTW I really admire your work it's almost identical to my company model.
Cheers,
Ralph
Paul
You have described perfectly the check for the HORIZONTAL axis.
A few comments:
First, just because the instrument came back from the shop does not necessarily mean it is good. We have had experiences where in the process of checking and calibration instrument and prism constants have been messed with. You need to recheck those.
On a topcon for example, there is the famous prism offset, but there is also an instrument offset that needs to be validated.
Also in recently rereading my topcon manual I see that there is a procedure to check the vertical compensators. Something I had never thought of checking before.
I assume other instruments have similar issues.
So if you are setting up both instruments in the same tribrachs and getting different closures (swapping in and out) it has to be something in the instrument.
Besides the above I have to wonder if the shop actually had a good baseline to test on. Go visit a NGS baseline and follow the full procedure prescribed on as many of the marks as possible. Any descrepency in distances should show up and eliminate at least one aspect of doubt.
We never assume an instrument is correct after it comes back from the shop. They are always calibrated, by us, before use.
They also get run over the EDM baseline regularly, as well as checked against each other, against GPS, and against the network marks across our state.
Distance wise, these instruments are very good. I'm 99% certain the problem is to do with the horizontal circle. It's just a matter of identifying where on the circle the problem is, and demonstrating that in a repeatable fashion.
Once I do that, I'm getting the instrument replaced. I've had enough problems with this instrument to call it a lemon. 😀
We try Ralph. Still got a lot to learn. 😀
Right Dave
The horizontal circle rotates about the vertical axis.
The vertical circle rotates about the horizontal axis.
Any way I meant to describe a check that the horizontal axis of the scope was parallel to the horizontal circle which is normal to the vertical axis such that raising or lowering the scope causes no change in the horizontal circle reading while following a plumb line.
Paul in PA
A Baseline Only Allows Checks Of The EDM
For checking other functions of a total station you are on your own.
Paul in PA
Leica 1200 specific: check the EDM mode, is it "Standard" or "Fast" ?
"Fast" means it won't update the angle with ATR after the first shot ... so if that's your EDM setting, in your redundant shots you are basically getting a new EDM distance along with measuring how well the robot repeats aiming the H&V of the first shot, rather than how well the robot centers on the target. And that robot-repeated-aiming is darn close but not quite as good the instrument spec for actually shooting the target. The "Fast" mode also has a slightly looser spec for distance.
Worth checking, had me puzzled for a while once.
Are you shooting this with the traverse program, some kinda "sets of angles" program, or each shot by hand (with ATR) ?
Standard EDM. I only use fast for topo style shots.
I am using the Leica traverse software.
Compare one or both of your bad instruments to the instruments that consistently get good results at the same time with the same tribachs. If you already compared the two bad ones to each other....maybe only one of them to your good one.
By the way, running the raw math around from one point back to the same point can be deceiving. Run it from different starting points will get different results. So any one traverse could have a range of misclosures from 1:60,000 to 1:200,000.
If you adjust out the angular error, you will get a consistent misclosure from each setup. (In the case above it will be bettrt than 60,000 and less than 200,000).
