Accuracy Issues

Sdrake14

Sdrake14

Active Member
Hi Jim,

I have noticed this phenomena as well. And I am curious. I can not say that I can relate it to an accuracy issue...
In fact I have (in the back of my head) actually intentionally oriented the unit to north when it seems to struggle with a fix, after seeing comments along this line. But I know nothing! (Schultz)
 
Nate The Surveyor

Nate The Surveyor

Well-Known Member
I was told long ago that there is a "ghost" that sticks up above any gps antenna. This "ghost" (electronic thing sticking up) varies between recievers. Therefore, they build an antenna, then model, and map its individual parameters. This is used in the equasion for all shots.
Now, as these units get more and more complex, maybe it's shaped like a large sweet potatoe. (angles out to the NE, or SE, as it goes up)
So, changing orientation, moves this ghost around, sometimes gaining more, or reducing the amount of data.
Maybe someone with more knowledge than I could confirm, or set aside this idea, as i cannot remember the source.
N
 
Jim Campi

Jim Campi

Active Member
Most of the work I do with the LS and T2 involves precise topo work. After taking 10 to 20 shots across a lot, I instinctively know the proper orientation of the LS to obtain the fastest fix. This is primarily relevant when there are overhead obstructions.

Orienting the LS on a per shot basis clearly increases productivity and I use this method on nearly every shot on parcels without a clear view of the horizon. If you rotate the LS with while reviewing a sky plot of the available sats/constellations, it's clear the signal to noise ratio changes on multiple sats. You will also see that that the number of available sats vs sats blocked by obstructions changes. This results in an optimal orientation per shot. In practice I find that of given parcel at a particular time and day have an optimal azimuth over a localized area. This translates to areas of a parcel that acquire a fix faster at a particular orientation.

If all you are doing is boundary work, then this will not be as useful.

To be clear, I notice this on virtually every use of the equipment.
 
Jim Campi

Jim Campi

Active Member
One more point, I'm slightly over 6 ft tall. My head seems to obstruct signals as well. When I become frustrated and the LS will not fix, bending over and tying my shoe can make a difference. No idea if this really impacts the time to fix or its a coincidence.
 
Jim Campi

Jim Campi

Active Member
Shawn Billings said:
Are both the base and the rover in open sky or under canopy?
Click to expand...
Hi Shawn, I always make a serious effort to set up the base with as much open sky as possible. I don't believe I have ever set it up under a canopy.

I just noticed that was not in response to me however; I will leave it for clarity's sake...
 
Sdrake14

Sdrake14

Active Member
LOL you are killing me Nate! I was already having images passing my head of (public relations) closing one eye on the horizon, mutter some mumbo jumbo, pitch dirt in the air, bow to the north, adjust the LS, and "bing" the shot....

That aside...hope this is true and the developer genius' could pipe in since the real time QC information we get with the Javad, and none other, gives it a clear advantage over units not providing this info at the shot level. It seems we are able to analyze this because of the volume of information we get during measuring...
 
Matt Johnson

Matt Johnson

Well-Known Member
5PLS
Jim Campi said:
Is it best practice to orient the LS with an Azimuth of 0 degrees (assuming the base is oriented in the same direction)?
Click to expand...

Both DPOS and OPUS are now processing using absolute phase center variations now so the receivers need to be aligned to north to get good vertical results (T-2 and T-1M record button and green triangles to the north, LS display to the north with the compass reading 180 degrees). Processing with phase center variations mostly improves the vertical results. A good paper about this topic is Influence of GPS antenna phase center variation on precise positioning.

I have not done any testing to see if fixing time decreases when both receivers are aligned to north. If both GNSS antennas have similar phase center variations then I expect this may help some.
 
Sdrake14

Sdrake14

Active Member
Good call on that article Matt....it looks like this may also explain poor results in elevations (partly) when using RTN. (different antennas and all)
 
Matt Johnson

Matt Johnson

Well-Known Member
5PLS
Jim Campi said:
One more point, I'm slightly over 6 ft tall. My head seems to obstruct signals as well. When I become frustrated and the LS will not fix, bending over and tying my shoe can make a difference. No idea if this really impacts the time to fix or its a coincidence.
Click to expand...

I would suggest you use a long pole tip so that the display is near eye level.

1524670942944.png


http://www.javad.com/dynamic/Shop/Product/6050
 
Jim Campi

Jim Campi

Active Member
Matt Johnson said:
Both DPOS and OPUS are now processing using absolute phase center variations now so the receivers need to be aligned to north to get good vertical results (T-2 and T-1M record button and green triangles to the north, LS display to the north with the compass reading 180 degrees). Processing with phase center variations mostly improves the vertical results. A good paper about this topic is Influence of GPS antenna phase center variation on precise positioning.

I have not done any testing to see if fixing time decreases when both receivers are aligned to north. If both GNSS antennas have similar phase center variations then I expect this may help some.
Click to expand...

Matt, great reply. For some reason I thought the LS should have compass reading of 0 not 180. All of my comments are related to time to fixed position/number of engines fixed not accuracy (although I assume it's indirectly related) I don't recall a compass heading of 180 being particularly helpful either. I will check this out in the next few days. Thanks!
 
Sean Joyce

Sean Joyce

Well-Known Member
Jim;
I orient the base to north, and try to orient the L.S. to south, which means the screen is facing north. Matt Johnson has a thread on this last year I believe.
This does have a positive influence on the fixes most of the time, but sometimes just rotating the L.S. randomly can help with the fixes in a real bad spot..
 
Darren Clemons

Darren Clemons

Well-Known Member
Sean Joyce said:
Jim;
I orient the base to north, and try to orient the L.S. to south, which means the screen is facing north. Matt Johnson has a thread on this last year I believe.
This does have a positive influence on the fixes most of the time, but sometimes just rotating the L.S. randomly can help with the fixes in a real bad spot..
Click to expand...
We do this quite often as well Sean. It "seems" to most always give a click or two very quickly when rotated - sometimes it's just a certain tree or branch that I think, by rotating, gives a slightly different angle for each signal to get through.
 
Tyler

Tyler

Member
I have a question regarding the accuracy of repetitive shots on the same point.

What is the best way to achieve an accurate location using the cluster average tool?

I would think taking a shot on a point and then rotating the rod 180 degrees would be the best way to find an accurate position once its been averaged. When I tried rotating 180 degrees there appeared to be about 0.05 feet of error when taking back to back shots.

However, I have read on this forum that orienting the rover to North will give a more accurate result which may be a bit beyond my current knowledge.

To achieve the most accurate position should I take redundant measurements with the rover all aligned to North or should I rotate the rod to compensate for any level errors?

What is the horizontal accuracy typically achieved with the Triumph LS and Triumph 1-M if I was to state it on a survey or to a client?

We have a new client who has asked us to provide him with a boundary survey that meets the horizontal requirements of a second order - class 1 survey which has a minimum distance accuracy of 1:50,000. I was wondering what we are able to achieve with the LS and if this is a possibility of achieving this kind of accuracy for our client.

-Tyler
 
Matthew D. Sibole

Matthew D. Sibole

Well-Known Member
5PLS
I suggest taking for 4 4 minute shots rotating the pole 90 degrees after each shot. Then do a cluster average of each corner. You can use the relative accuracy calculator to determine the accuracy of your averaged shots. If it does not meet your requirements go locate them again and delete the old average points and re-average.
 
Tyler

Tyler

Member
Matthew D. Sibole said:
I suggest taking for 4 4 minute shots rotating the pole 90 degrees after each shot. Then do a cluster average of each corner. You can use the relative accuracy calculator to determine the accuracy of your averaged shots. If it does not meet your requirements go locate them again and delete the old average points and re-average.
Click to expand...
Thank you for the response, I will look into the relative accuracy tool and see if this procedure will work for meeting the accuracy requirements. So it sounds like it is more accurate to rotate the rod and average the measurements rather than taking multiple measurement with the rover pointing to North?
 
You must log in or register to reply here.
Top