I had one of those once a career flukes this morning.
I set the instrument about twenty feet off the front of a house and nearly aligned with the left side wall and proceeded to resect from existing control in the street. I was able to see the first nail at low rod but on the second I had to raise the rod to 8' to clear the trucks in the yard. I was shooting forward and reverse and using a bi-pod so I was a bit surprised when the angles to the second point were out of tolerance. I had it re-shoot the second point, got a good set of angles, hit solve, and the DC responded "No Solution". At this point I just canceled the resection and started over, using the high-rod point as the first resection point this time. After shooting the low-rod point I pressed solve and got a closure of 5 with an angular rms of 3000.
I stepped out of the street to scratch my head and figure out what went wrong. The first clue was that the DC was showing an angle of roughly 180° between the two control points when I knew the real number should've been in the 90-100° range. The second clue was that when I had arrived at the low-rod point, the instrument was facing the house instead of the street. I had expected to lose lock when I walked between the points due to trees and bushes but thought it odd that it had turned toward the house while searching for me. I was pretty sure there was only one explanation for the odd results.
I had the trucks moved so that I had clear line of sight to the instrument from both control points and repeated the resection a third time getting good numbers this time. From the control point where I originally had to raise the rod I could see a reflection of the instrument in the left most window on the front of the house. To prove my theory, I used the joystick to rotate the instrument to face the window until I could see the red search light from the instrument reflected in the window then pressed search on the DC. Sure enough, the instrument locked onto the reflection of the prism in the window and I was able to move around and take shots as long as I could see the instrument reflected in the window.
If I hadn't seen it; I wouldn't have believed it. So here's a picture.
It's a bit tough to see, but you can tell that the instrument and prism aren't facing each other and you can see a faint reflection of the instrument in the left most window.
Reason #545 to never just assume that just because you pushed the correct buttons in the correct order that the answer must be right.
> If I hadn't seen it; I wouldn't have believed it. So here's a picture.
I have seen it. It used to happen a lot more in the early days of these auto-tracking things. That was the selling point of the Trimble 5603 (ie/ Geodimeter 600) with the diode tracking, you got this less often. But it was still possible.
> > If I hadn't seen it; I wouldn't have believed it. So here's a picture.
> I have seen it. It used to happen a lot more in the early days of these auto-tracking things. That was the selling point of the Trimble 5603 (ie/ Geodimeter 600) with the diode tracking, you got this less often. But it was still possible.
Yeah, mirrored windows will do that with the 600/5600s. It's about the only time you have a chance of getting bad shots.
I've shot my 5600 & my old 4400 (600's grand daddy from mid-90's) through windows and things looked ok. I proceeded, but always spent the time to check back from a not so ideal setup to satisfy me. Should do that anyway, but that's why we buy robots.
Stephan's issue is really weird, but at least now he knows what not to do next time.
I had my Trimble 5600 locked on me one time while setting a control point in robo mode. I moved ahead and BS where I came from. 20' long for distance??????? How can that be????? Reset up on original control point and turned to the control point (20'off pt) and looked through the eye piece on robo. The gun had locked on the diode but had shot a reflective traffic sign that was blocking the view of the prism portion of the foresight (.4+-above diode). God has a way of kicking us in the teeth when we don't do our checks. (There were times when I would have just left a nail for a BS and never checked back site distance) Live and learn! Jp
Cool. Next time try for an Alley Oop.
What surprised me most was that the instrument found the prism reflection in the window to begin with. I initiated the search for the prism with the RC-3 so the robot should've been scanning for the beam from the RC-3 rather than looking for reflective surfaces. The RC-3 beam is narrow and at that distance I usually have to have it pointed carefully at the instrument or it will turn circles endlessly until it times out. Given the distance between the instrument and the window it shouldn't have even been searching in that direction. Too bad the control point was in a road where I couldn't leave the rod unattended. I would have been interesting to leave the rod setup and go take a look from the instrument position to see what it was seeing that caused the craziness.
Possibly somebody walked by the target and blocked it at the critical instant. Or a bird flew by.
Likely the RC3 was bouncing off the window as well.
In the picture I've intentionally turned the prism and RC-3 toward the window. However, when the initial goof happened, I had the RC-3 pointed at the instrument. Maybe it reflected off something on one of the trucks to the window and then to the gun in a JFK magic bullet sort of scenario.
One other possibility is that the beam from the contractors rotating level that was setup just inside the garage door on the left side of the house caused the instrument to get confused.
One day last fall I kept getting a bad backsight delta. Guy had parked his lowrider between me and the gun, and his mirror polished front rim just happened to fall on line. Hit "lock" and the robot went down instead of up, picked up the rim, and gave a bad D about 5 times before I figured out the ruse.
> I had my Trimble 5600 locked on me one time while setting a control point in robo mode. I moved ahead and BS where I came from. 20' long for distance??????? How can that be????? Reset up on original control point and turned to the control point (20'off pt) and looked through the eye piece on robo. The gun had locked on the diode but had shot a reflective traffic sign that was blocking the view of the prism portion of the foresight (.4+-above diode). God has a way of kicking us in the teeth when we don't do our checks. (There were times when I would have just left a nail for a BS and never checked back site distance) Live and learn! Jp
Way before the advent of robotics, I had similar problem road design survey. The crew went out and collected the topographic data over a week period. When the I processed the data in the office we notices a gap in the collection. The crew swore they had got the area. After closer examination, the shots were all there but in a line. It turns out the instrument person was shooting through a barricade to the PC on the rod. The instrument had collected the correct angles, no surprise give since this was pre-tracking, but the EDM collected distances off the reflective tape on the barricade. Needless to say we had to come up with some new field procedures.
