Did a small road topo today with our Focus30, robotic. When I brought in the points and ran field to finish it looked horrific. As I looked at the shots I realized that about a third fell out of place by as much as 30'. These were striping, b curb, t curb, walk shots, one set would be fine, on the next set the walk would be 10' too far from gun and 20' into the road. Next sequence fine then a stripe shot would be behind the walk, 15' too close to gun. Double yellow line zigged all over, stop bar was diagonal across the road.
At 1st I thought it may have caught car reflectors but its on both sides of the road with no pattern. Data is in the right place globally, just random shots are out of place.
Looks the same from a .csv directly exported and processed raw.
I've never seen this before, any insight appreciated.
Are you perhaps piking up the refection of the prism in windows?
Mark Mayer, post: 420170, member: 424 wrote: Are you perhaps piking up the refection of the prism in windows?
No buildings near my work and it happened looking straight up a CSW with no obstructions at all, as well as looking through traffic. Baffling, will teach me to not take the Sokkia...
Mark Mayer, post: 420170, member: 424 wrote: Are you perhaps piking up the refection of the prism in windows?
Wish you the best of luck on figuring this out. Multiple errors like you describe would make my neanderthal brain want to pay the techs to check out the gun, I'd surely do a test run in a known and controlled environment to check things out first.
How old is the robot? Have it been checked? Sounds like the board may be failing. Had an old trimble, some shots the distance will literally be out of this world. Then sent it in and they said the edm board was on its death bed and had to be replaced.
I am watching this thread with interest, since my intended next robot might have been a Focus.
Like Sireath above, I had a Topcon 303 that started giving funny answers. I first noticed it when nailing house footings. The instrument man would tell me to come toward the gun .02, then on the very next shot it would be go away .02 I'd have thought it was me, but I was measuring the distances and dead steady on the rod and it started happening more and more often. Then one day it started busting backsites on a boundary by hundreds of feet and we figured out that it had a problem. I think they called it the "chopper motor" but I remember that it had to do with the timing of the edm signal.
The back site busts sound like and issue, hope you fired the i-main that thought 0.02x0.02 wasn't good enough. Hopefully those busts weren't in 0.02' range, might put you out of business, accuracy you know.
BK9196, post: 420185, member: 12217 wrote: The back site busts sound like and issue, hope you fired the i-main that thought 0.02x0.02 wasn't good enough. Hopefully those busts weren't in 0.02' range, might put you out of business, accuracy you know.
The point of telling that part of the story was to show that the errors started out small and infrequent before it completely broke.
For what its worth, when nailing footings for block work, I always try to keep the tolerance below a hundredth. That way I leave more of the error budget for the masons.
BK9196, post: 420177, member: 12217 wrote: Wish you the best of luck on figuring this out. Multiple errors like you describe would make my neanderthal brain want to pay the techs to check out the gun, I'd surely do a test run in a known and controlled environment to check things out first.
I did a few checks on known points throughout the day, they were fine. There seems no rhyme or reason to errors. Gun was out 2 days last week with another crew, no issues.
sireath, post: 420178, member: 9370 wrote: How old is the robot? Have it been checked? Sounds like the board may be failing. Had an old trimble, some shots the distance will literally be out of this world. Then sent it in and they said the edm board was on its death bed and had to be replaced.
It's a little over 3 years. Was calibrated last summer.
Brad Ott, post: 420180, member: 197 wrote: I am watching this thread with interest, since my intended next robot might have been a Focus.
For the $ I'd say go Sokkia. Works with any software, search/lock with PR remote is way beyond the Focus, holds lock better. The Ranger3 has had the radio die once and the charging port literally fell out. Both repairs were over $350.
Stephen Ward, post: 420193, member: 1206 wrote: The point of telling that part of the story was to show that the errors started out small and infrequent before it completely broke.
For what its worth, when nailing footings for block work, I always try to keep the tolerance below a hundredth. That way I leave more of the error budget for the masons.
I rarely use this instrument. Had to get the job done and it was the only one left. Now I have to do it all again, 1.5hr drive each way to top it.
Luke J. Crawford, post: 420201, member: 11382 wrote: I rarely use this instrument. Had to get the job done and it was the only one left. Now I have to do it all again, 1.5hr drive each way to top it.
I traded in a Topcon GPT-3005w because it would give a bad shot about every 1000 shots or so. We had the EDM board replaced once and a few months later it started doing it again so we traded it because we couldn't trust it.
BK9196, post: 420185, member: 12217 wrote: The back site busts sound like and issue, hope you fired the i-main that thought 0.02x0.02 wasn't good enough. Hopefully those busts weren't in 0.02' range, might put you out of business, accuracy you know.
.02 is too much in a lot of situations.
Sent from my Pixel XL using Tapatalk
Stephen Ward, post: 420210, member: 1206 wrote: I traded in a Topcon GPT-3005w because it would give a bad shot about every 1000 shots or so. We had the EDM board replaced once and a few months later it started doing it again so we traded it because we couldn't trust it.
If it was 1-1000 I'd say ok for this job, > 300 points of which 75å± are obviously bad and I'm not inclined to trust the rest now either.
Luke J. Crawford, post: 420200, member: 11382 wrote: For the $ I'd say go Sokkia. Works with any software, search/lock with PR remote is way beyond the Focus, holds lock better. The Ranger3 has had the radio die once and the charging port literally fell out. Both repairs were over $350.
Actually, I'd like to retract that somewhat. If most of your work is topo and boundary, the Spectra Focus is a great gun. It's fast, not too power hungry and lighter than most other robots. If you do primarily construction layout, the Sokkia is nice as it allows full function of the gun as a manual would even when connected to a collector whereas the Spectra only allows for interface through the collector.
The GeoLock of the Spectra/Ranger combo works very well in areas with sky view and there is a little distance (å± 50') between the rod & instrument. In dense woods or in tight construction sites the Sokkia/PR has an advantage.
ease, post: 420212, member: 1584 wrote: .02 is too much in a lot of situations.
Sent from my Pixel XL using Tapatalk
simple fix for that .02. swap the ' for mm...
Think the issue Spectra dealer said they've seen issues such as this when the antenna on the Ranger3 gets older and flexed a lot. I threw on a an extra one we had (used) and re-did the same survey. Twice I got a bad distance on my BS checks, an even 210' to long, but that is a far cry better than every third shot. It also connected faster and didn't lose connection randomly. On the way back to the office I stopped and grabbed a new antenna, the design of which they've changed, likely due to this issue. Assuming, hoping the issue is fixed, if so this will end the thread.
