Recently I have had the experience of also using a Topcon QS w/ FC 2500 and Topsurv SW and I started finding "wildshots". Not super common, but every once in a while when I'm processing the dtm, I'll find a shot which is not where it should be. Crazy thing is that these shots seem to be strictly a horizontal angle bust. For instance on one job, we shot a series of points along a ditch line located at the toe of a slope. During processing, one of the shots showed up being 20' into the cut bank, but the elevation was consistent with the ditch.
Another job, I shot a sign pole and when processed, the sign shot ended up being about 10' west of where it really should have been.
In both instances, the correct location of where we are sure the shot was taken can be obtained by adjustment of the horizontal angle of those particular measurements. Distance and elevation appear correct for these shots, only horizontal angle appears to be off.
Also, shots taken before and after the errant ones are good. I do frequent check shots to control with good results.
I've done 20 or so topos with this set up and everything usually comes out looking good. These wild shots are so far out that they are obvious. The nature of most of these topos is that relatively small errors would show up. Perhaps there are some small magnitude wild shots, but our frequent control check shots are consistantly tight.
Perplexing.
I found the same quirk using a Topcon 9003a robot with TopSurv on an FC-2500. I think it has to do with how TopSurv works. Like all robotic software, TopSurv is getting a constant stream of position info. When you press the button to take a shot in TopSurv it appears to store the last data set it received prior to the pressing of the button. Carlson SurvCE on the other hand shows you the constant position info but when you press the button it calls for a fresh data set and then stores the data.
TopSurv's method gives quicker shots but if the gun starts chasing cars, squirrels, or other shiny objects at nearly the same instant you press the button for a shot you'll get a good vertical angle, good slope distance, and a rogue horizontal angle. SurvCE's method is slower especially if you're in fine mode but I've never gotten a rogue HA since making the switch (all other equipment is the same).
For control points you can compensate by setting it to shoot face1 and face2 (not likely to get matching rogue HA's) but for topo shots you just have to be very aware of what the gun is doing. If I lost lock or caught the gun squirrel watching immediately after a shot, I staked the last shot or just simply re-shot it. There may be a setting to change the shot behavior in TopSurv but I moved on for other reasons before finding it.
That makes sense with what I've been seeing.
thanks.
I think he hit the nail on the head. With software/hardware combos trying to talk faster and faster in addition to the passive technology of most instruments out there we are seeing this more often, to just on Topcons only either. Most instruments, whether or not they lock on to something else still "look", although extremely quickly at other reflective objects in the direction of the prism. If by some rare occasion the point is stored while this is happening or a glitch occurs that slows the data down you may see some busted angles/distances. That being said, that particular brand is notorious for locking on reflective items other than the prism. I wouldn't be surprised it you were locked on something at or around the right elevation that wasn't the prism.
He's also right about Carlson. With Carlson you won't see the problem near as often, if you do it normally has to do with actually be locked onto something else otherthan the prism and not knowing. It's is why I say, leave the software to the software guys.
I agree , we had wild shots with Topsurv , but switch to Carlson one year ago . No issues.
I knew what this was just from seeing the subject line - I posted about this last year:
http://beerleg.com/index.php?mode=thread&id=173549
Stephen had the same answer back then, and I think he's right.
Although I've managed to avoid this happening since I posted, just last week while taking topo I heard the "lost lock" sound just after I stored a shot. I took another shot without moving the pole, and it turned out to be about 30 feet different.
You can avoid it by shooting from both faces, but that's quite a bit slower. I was thinking that maybe setting the instrument to measure more than once & average the measurements might prevent it, but I'm not sure if it records multiple angles as well as distances when it does that, so that might not help.
Just look at it as another one of those Topsurv "features"!:-)
Jeff
