Just to get going here, I??ll be the first to admit I??m a dummy (and early in my surveying career), but I??m interested in finding out where I went wrong, so I??m all ears.
I thought it would be easy to tighten up my wall target elevations by sighting them with a Trimble DiNi level, but I??m seeing 0.04?? to 0.055?? [vertical] difference between my DiNi ??measurements? and my robotic data, shot with a Trimble S5 and a mini prism. I would be disappointed with 0.02?? of disagreement between the S5 and the DiNi, but a consistent ~0.05?? means something I??m doing is *&@#$ed.
Here??s the robotic situation:
To shoot the wall targets, I used a Seco mini prism taken out of it??s harness thing. This mini prism has a little nipple attached to the back of the prism housing, and I stuck that nipple on the gnats ass of my wall targets which were stuck to huge concrete columns at around 5.20' from the floor. Having seen weirdnesses on other jobs, I found an EngineerSupply.com video about ??nodal points? and implemented what they discussed in that video by directing the face of the mini prism at the robot (my coworker sighted through the S5??s eyepiece to let me know if I was mis-oriented). I used a 0mm prism offset setting for this configuration.
All robotic shots, including sideshots, were ??repetition shots? (in SurveyPro parlance), i.e. Backsight-Face1, Foresight-Face1, Foresight-Face2, Backsight-Face2.
Backsight/foresight??s were to -30mm traverse prisms, and each leg of the traverse was roughly equal (with a nice long backsight of ~300?? to initiate the traverse).
All sideshots to wall targets were shorter or equal distances than my backsight shots. After station setups, my readouts would report my backsight delta??s, all were less than 0.02?? vertical from the previous setup??s foresight, often they were under 0.01?? from the previous setup??s foresight.
Also, I recently did all three calibration routines recommended for the S5, did them in similar climate conditions?? perhaps one note is that I performed the ??tracker collimation? on a -30mm traverse prism which is much larger than the mini prism I used for my robotic shots on the wall targets (not sure what the significance is there). I do not know the precisions of the traverse/mini prism, but we get them from a reputable dealer in our area (I know the mini prism is a Seco)..
Also, I dutifully monitored the S5??s electronic level bubble throughout each leg of the traverse, so I can verify that the level was less than 10? out of plumb in any direction for any given shot (the level compensator was on the entire time).
Here??s the DiNi Level situation:
As far as I??m concerned this thing lies to me all the time. From its readouts, I can level through an entire site and check in to my original benchmark within 0.002??. Professionally I think that??s great, personally I think that??s bogus, but whatever.
The level has crosshairs visible when looking through the eyepiece. For backsight/foresight shots on my traverse control, I basically point those crosshairs at the barcode of the level-rod, get the barcode into focus, then shoot. After a backsight on a known or computed benchmark elevation, a screen readout reports a Z which is approximately 5.0?? above that known/computed benchmark elevation?? I interpreted this as the HI or height of instrument. I also assumed that if I look through the eyepiece, those crosshairs show me where that Z is, i.e. I can spin that DiNi around and my crosshairs will locate my HI. Have a sneaking feeling that this is where I went wrong.
Anyway, turning a backsighted DiNi to a wall target, I see those crosshairs overlaid on my coworkers tape measure, which has 1.00?? pinned at the gnats ass of my wall target, and is dangling vertically. I note the Z of my backsighted DiNi readout, and then I count how many hundredths, plus or minus, my wall target is from the crosshairs with that Z.
Back at the Office:
Back at the office, I start adjusting my robotic data with my level data, none of my traverse elevations change by more than 0.015??, and my vertical misclosure for my entire traverse is around 0.015??.
The important part is that I improvised because I couldn??t use the level-rod to shoot the wall targets - I just followed the procedure above and wrote all the HI+/-tapemeasure elevations in my notebook.
I would love to hear any feedback, thanks.
It sounds like you are mixing and matching bar code and optical readings when using the DiNi.?ÿ This is a no-no unless you have carefully calibrated the bar code and optical readings to be in sync.?ÿ The bar code and optical crosshair systems in a digital level are not necessarily coincident.?ÿ You could go out in the survey yard and check this without returning to the job site.?ÿ In fact, you could determine how different those observations are at certain distances and perhaps even save yourself a return trip to the job site (although a return trip might be best).?ÿ I happen to know a young surveyor who did just that about 15 years ago after learning this lesson the hard way.
You should be able to use your DINI manually and get better results vertically than the trig from your robot. The DINI level should be "pegged". If it is pegged correctly it will give you the most reliable vertical differences. My guess is your level is out of adjustment and you need to dial it in.
1st question: where are the targets?
2nd question: can you "buck in" on a target and turn to another to see if they agree?
General observation: Your robotic data and level data should agree by at most 0.02 so you are correct to investigate.
I don't know much about the DiNi so this question might be relevant: does the level need to be "pegged?"
It sounds like you are mixing and matching bar code and optical readings when using the DiNi.?ÿ This is a no-no.
Yup. Agree - I missed that in the original post.
A few things immediately jump out at me:
1. Did you peg & adjust the DiNi before beginning work? Calibrate the bar code to manual staff?
2. You mentioned traverse legs of equal length. Did you balance backsight/foresight observations with the level as well?
3. Did you level through your traverse points? That would be my first move before starting to observe those targets, either with the level or the total station.
4. What's the reference mark for these elevations? I might have missed this, but I didn't catch what you were holding for the traverse vertical versus the level vertical. A single starting observation may be the culprit here.
?ÿ
Generally speaking, if I can't physically set a level rod on a mark, I'll level through my traverse points, adjust the traverse and level data together with appropriate weighting, and then use the total station to observe the targets.
A well-adjusted DiNi will almost always pummel a total station when it comes to transferring elevations, barring extreme topography.
I am heading out to the street now to do this - guessing it's the two-peg adjustment you're talking about. Had no idea there could be such a mismatch.. still a little unsure I understand but will look into it.
Someone could correct me on this, since it's been a while, but I think you're looking at three separate adjustments.
1.?ÿ Peg and adjust the level in bar-code mode
2.?ÿ Peg and adjust the level in optical mode
3.?ÿ Collimate the optical readings with the bar code readings.
If you did just Item 1, you could run beautiful levels all the time?ÿas long as you stayed in bar-code only mode.?ÿ?ÿOtherwise, if you want to mix and match, you need the other two adjustments.
Someone could correct me on this, since it's been a while, but I think you're looking at three separate adjustments.
1.?ÿ Peg and adjust the level in bar-code mode
2.?ÿ Peg and adjust the level in optical mode
3.?ÿ Collimate the optical readings with the bar code readings.
If you did just Item 1, you could run beautiful levels all the time?ÿas long as you stayed in bar-code only mode.?ÿ?ÿOtherwise, if you want to mix and match, you need the other two adjustments.
It's also been a while for me, but I am 90% sure that when the DiNi is pegged, it is just a single procedure and always done in the bar code mode. Corrections are computed, and then at the very end of the procedure it will prompt you to flip the rod around and check the optical reading against a computed value that it displays on screen.
If those values are different by more than 1-2mm, the reticule needs to be adjusted with the allen key in the case. Once that is done, everything will match up between the bar code and optical readings.
As you mentioned, where folks get into trouble is by skipping that last step and then trying to mix and match.
Assuming the targets are flush with the wall, shooting them direct reflect makes more sense than fiddling around with a disassembled prism.
You guys rock, I've been hunting around for a manual - all I have is the Quick Reference Guide (which does briefly explain the procedure on page 43). Still looking for that manual but I'll stop when I find the main adjustment procedure (I was hunting around for mentions of bar-code/optical modes and collimations).
Thanks everyone!
The manual that I have is attached. Those things don't get updated very often so most of it should be relevant.
I liked the Forstner method for adjustment....can't remember why, think I found it faster/easier.
Posts like this are what give me a glimmer of hope for the profession.
There's not always one perfect way to do things, and if something goes sideways it is critical to figure out what went wrong. Not point the finger or say "it's not my problem". There's a time and place for that, but only after all other options have been explored.
I for one commend you for immediately looking for a solution rather than just complaining and/or giving up. The problem-solver mindset is a rare thing (even among surveyors).
Is your level loop .02 feet or .02 cm either way if the didni is pegged and adjusted you should be able to run a very far distance level runs using appropriate procedures and get much better than that. My guess is level is out of adjustment. ?ÿAlso is this an invar 1 piece level rod. And once you have that dini in adjustment use your tape and ck it and any variation of it against the dini at a few spots. This i say only because you are mixing them. Now if you just read the tape on the targets only and not mix the invar rod run levels through targets themselves turning through them. This will give you the height differences between targets. ?ÿIf your level bs and fs are spaced correctly even distance and a rough 45 degree angle. ?ÿIt will balance the error in the level itself. ?ÿIf its out of adjustment any. Just read above and below so you change your height some. Keep up with sign negative and positive readings . You are getting some great advice here.?ÿ
?ÿNext time you run into something like this. Remember you can just do height difference between targets as well. ?ÿNo elevation needed to see how the fit with each other compared to the trig levels from a total station. ?ÿ
Ok, I'd love to say third time is a charm, but I'm still feeling uneasy about this.
I've run the "Forstner" peg adjustment a few times now, this last time I made quadruple sure the rod was nice and level during the readings (although I think the DiNi was still warming up in the sun cause for the first time ever I couldn't level it in multiple directions - I'm starting to suspect gremlins). The first part of the adjustment uses the Barcode side of the rod, and resulted in "line of sight corrections" which came down to ~16" (from ~22").. I accepted the new corrections.
Having had the 0.05' of error I spoke of in the original post, I was expecting the second part of the adjustment (which involves the reticule or crosshairs) to need a fair amount of adjusting. For the second part, the rod stays at the last shot made during the Forstner routine, but gets rotated 180?ø to turn the metric graduations toward the DiNi. The DiNi readout tells you what height you should see when you look at the metric gradations through the eyepiece. Mine said 1.512meters - I had my buddy look through the eyepiece to make sure I wasn't missing something - it was as close to 1.512meters as is possible.
So I guess my question now is: does the DiNi adjustment routine basically re-calibrate the barcode & reticule together? (in which case the last page of the calibration routine, where you turn the rod 180?ø and check the metric graduations, is that just for triple checking? Why didn't I have to adjust the reticule if I was previously finding errors of ~0.05' at no more than 60' away?)
First lets go in order. If you can??t get the dini to level then that needs adjustment first. It has been many years but I would set dini up and use two foot screws set the dini parallel to these two foot screws sometimes called leveling screws. Use these two and adjust bubble and the third. Then rotate 90 use only one. Go 180 take half the error. Go back repeat. If still will not stay in adjustment take half error out with adjustment screws for bubble. Repeat all steps slowly getting it in level. Then peg check or c check everything again. Hope this helps. You have error in readings you may have to adjust cross hairs manually per manual if I remember correctly?ÿ