Just getting into the GPS scene and was wondering what the protcol is for using two bases and one rover. How do you initialize the rover if using nongrid gontrol points. Thanks

Thanks for all the help....it was greatly appreciated

Ben,
I don't know if you'll be checking back since it seems no one is responding.

In regards to your question, I believe that you will build a stronger network quicker if you rove two units rather than one. That allows you to process vectors between your GPS points. Otherwise, your GPS points, though probably good since you have data between both base control points and each point, are not interconnected with GPS vector data.

If you occupy one point as principal control for the project and "leap frog" the other two as a general strategy, this allows you to compute a loop closure through all your control.

3 receivers will double your productivity over two receivers, BUT I am not a fan of two bases and 1 rover. You end up with a bunch of independent triangles. Besides that, you really have no redundancy at the rover end. Yea, sure you will have a closed triangle that may give you warm fuzzies, but you don't have an independent check on closures. Any errors at the rover end will be transparent in the so called closure.

3 receivers work good, just not in the 2 base, 1 rover mode.

Shelby

with two bases if one vector to your rover fails because of an obstruction, the other will likely pass.

No offense Shelby, jmho.

Lew,

So how do you know that the one that "passes" is any good. Just because the software resolves the ambiguity does not make it correct without some independent verification.

I am saying that if you have an obstruction at your rover, or more likely multipath, it is going to affect both vectors. I inherited a project in 1990 when I went to work for a place that had done a project like this, each triangle closed BUT they where shifted in space from where they should have been. If you are not doing independent observations, i.e. a new occupation with different satellites, you are mostly kidding yourself. Sure 99.9% of the time it will probably work, but not having a truly independent occupation will bite you eventually. I used to think doing an antenna dump would ALWAYS work at rooting out a problem too. I have had antenna dumps reinitialize and provide two bogus answers that were the same, there simply is no substitute for an independent observation.

No disrespect meant.

Shelby

Shelby,

First I must say I have not used the Promarks.

I agree with Shelby, if the bases are in an open area with no obstructions and you get a bad vector. It would likley be from an obstruction in the rover, hence both vectors bad.

Seems it would make sense for a solo operator to run a base and two rovers. One rover always in a "stand-by" positon. When you loose lock, go to the "stand by" and swap rovers. Let the loss lock rover re-initialize on the known point, while running the other "stand-by" with lock.

Does this make sense?


Maybe not, since if you have the known point close by you could just re-initialize the single rover.
Lee Green
Modified By Lee Green on 3/2/2003 at 5:39 PM