Community Forums

Share:

VRS vs Base in site calibration  

Page 1 / 2

cgoodwin1212
Posts: 2
Member
(@cgoodwin1212)
FNG
Joined: 4 months ago

First off, I am new to this site and it has been very helpful so far in shedding light on many topics. However there is one thing I can't find.

 

I just started a new job in the private sector coming from our state's DOT and they do a site calibration on every job that they do. Coming from the DOT we used Access, the network with an RTK bridge, and typically adjusted to one CP central to the project and went from there for most projects. I loved those tools, they were easy to use, powerful, and fast. I became very familiar with them and did all of our jobs on that. However at the new job, they are running SCS900 and a base/rover set up for pretty much everything that they do surveying.  

 

I am trying to convince them to invest in TSC7's with access w/roads for our surveyors and to use the network more to expedite work flow but a couple of them are afraid of change. My biggest hurdle is convincing them about the network for site calibrations. I did a quick calibration with access and the network around our shop and showed them it can be done, and then exported a DC file to SCS900 set up a base/rover and showed them it would still work. However, if there was a calibration done with a base/rover, if I try to use the network I am off by a couple feet in either direction. I read, somewhere on here, that you can use a base rover with a network calibration, but not the other way around. Unfortunately they didn't go into detail so I am a bit confused as to why. If anyone can shed some light that would be awesome!

8 Replies
1 Reply
JaRo
 JaRo
Member
(@jaro)
Joined: 9 years ago

1,000+ posts
Posts: 1501
Posted by: cgoodwin1212

However, if there was a calibration done with a base/rover, if I try to use the network I am off by a couple feet in either direction. I read, somewhere on here, that you can use a base rover with a network calibration, but not the other way around. Unfortunately they didn't go into detail so I am a bit confused as to why. If anyone can shed some light that would be awesome!

That few feet difference is probably because someone did a "HERE" shot. They are taking an approximate LLH and pairing it with solid coordinates.

Reply
MightyMoe
Posts: 6218
Member
(@mightymoe)
5,000+ posts
Joined: 9 years ago

You can calibrate using the older versions of the software and a VRS.

Maybe calibration isn't what we are discussing cause you can use a base/rover with a network calibration, and you can use a network with a base/rover calibration. The user has to understand the data to make it all work. A calibration is just comparing the XYZ data to the LLH data and doing a calculation to merge them together. 

Back on my high horse I will say that calibrations (with Trimble anyway) do things to the underlying data I don't like, so I don't use them, network or base/rover. There are better ways to handle the data. 

Of course you can't use the VRS with a Base/rover calibration that is on a different system than the VRS without doing a VRS calibration to the XYZ numbers if that is what you attempted. You aren't on the same LLH's.

Reply
JaRo
Posts: 1501
 JaRo
Member
(@jaro)
1,000+ posts
Joined: 9 years ago

For the Texas Central and North Central zones, I have created template job files that I use for TxDot work. All I have to do is key in the TxDot Surface Adjustment Factor for the job or the County that the job is in. I can do this in Access or Survey Controller. It can probably be done for your state also.

If the guy building the model for the dozers needs a calibration file. I do a box calibration in the data collector without taking a single shot. This also tilts the plane to best fit the geoid in a smaller jobsite area (as opposed to the area of a state or zone). Taking the geoid out makes processing the data on the dozer much faster. It can also be exported in a .DC file for SCS900.

James

Reply
Brad Ott
Posts: 4592
Member
(@brad-ott)
2,500+ posts
Joined: 9 years ago

Maybe the few feet has to do with a US FEET setting in one controller & an INTERNATIONAL FEET setting in the other?

Reply
cgoodwin1212
Posts: 2
Member
(@cgoodwin1212)
FNG
Joined: 4 months ago

@mightymoe, what are you seeing it do to the underlying data that you don't like? From my understanding its just doing a least squares adjustment for its horizontal, and then creating a tilted plane based either off a geoid, or ellipsoid height?!

 

@brad-ott I thought the same thing, but it was the same job opened up on the same collector using the same rover. Only difference was I connected to the network after the base had calibrated the site. We're on the far east side of the state (5xxxxx, 27xxxxx), so my easting value would have been over twice the distance off, which it wasn't.

 

@jaro we are using the calibration for our contractors. For what these contractors are using scs900, and their understanding of the software, that I've noticed in my month here, they need the calibration file otherwise they're lost and confused. I like what you said about the 'here' function. We'll just stab in a piece of rebar anywhere and set up a base  and go around and start calibrating and then save the corridinates of the rebar. SCS, from what i've seen so far, doesn't have a place to put in my LLH for the state plane im working in like access does.

 

 

Reply
1 Reply
MightyMoe
Member
(@mightymoe)
Joined: 9 years ago

5,000+ posts
Posts: 6218

One problem with calibrated files are the elevations. There is no way to make them work well, only with a very small site will they work at all where I do most of my work. A city block would be OK I suppose. I found out that even the small local college was too big an area for a calibrated file. No matter how I would calibrated that area it would never work. As Geoid models became more accurate it was clear that they worked much better modelling the local undulations and that's what I use for elevations when using GPS-Geoid 12 with any new project.

Of course moving past any calibration can create really bad result. And it doesn't take long for awful things to happen. Machine control tends to need a calibration for some reason, a local mine tried to cut in an access road in a busy geoid area and found themselves almost 6' off as they moved away from the mine to the far extent of the new access road. They tried to push a calibration using dozers and scrapers based on control surrounding the mine way too far.

That is one component. The other is what it (at least Trimble) does to the horizontal data. The calibration changes the Local Lat, long, heights and creates a new set of data under Local Coordinate. The Global Coordinate should remain fixed, but for some reason Trimble decided to adjust the Local Coordinate and that is what it uses in it's calculations. In other words, the calibration changes the latitude, longitude values used to do calculations. They will be correct at the hinge point, then expand away from there, if you need a correct latitude, longitude value (I often do) you need to be very careful in a calibrated file not to grab the wrong one.

They were sold back in the 1990's as a way to kinda make elevations work sorta with GPS and to merge into conventional XY control quickly. But into today's world I don't understand why anyone uses them, except machine control guys who say they have to.

Much better to use a projected file, leave the Lat, Long data unmolested, have metadata that can be correctly expressed, data that can be recreated by another user and simplify your coordinate systems.

Reply
Page 1 / 2
Default
:)
:d
:wink:
:mrgreen:
:neutral:
:twisted:
:arrow:
:shock:
:???:
:cool:
:evil:
:oops:
:razz:
:roll:
:cry:
:eek:
:lol:
:mad:
:sad:
:!:
:?:
:idea:
:hmm:
:beg:
:whew:
:chuckle:
:silly:
:envy:
:shutmouth: