I can find plenty of testing/comment about how well these receivers with the Phantom 40 board do in tree cover/challenging conditions but little about the repeatability in ideal GNSS conditions.
For urban boundary work here I need <10mm in horizontal for control so not interested in the vagaries of NRTK, only using base/rover with <2km baselines. From what I see in online user comments for R12/GS18/BRX7 is that good repeatability (checking against total station work) for latest generation GNSS in canopy is still more like ~20mm at 95% confidence vs the 5mm of total station without too much effort. I've only used the GS18's myself and the experience lines up.
For those using the BRX7/S631, assuming clear sky (nothing within a 15° mask) and a reasonably thorough method like 20-30 epochs on tripod (or left/right pair to average out any pole runout), wait 20 mins, then shoot the point again what sort of repeatability are you seeing in HZ & V at 95% (not just the RMS screen readouts)? It's the precision of the observations I'm interested in not so much the absolute accuracy to datum.
Hi Luke, full disclosure, I do work at a shop as a support technician that sells Hemisphere S631 units. However, with the Hemisphere boards (the Carlson is simply a blue one) you have access to the SureFix option. SureFix runs two independent and unique RTK engines in parallel that will not provide a fix until in agreement. I personally have had great luck under canopy, near buildings etc with repeatability.
In theory, this provides something like a 99.999... % confidence level in the shot (I will admit this math is well above my pay grade). However, I would suggest trying a unit out if you can. I have tested Trimble, Topcon, Leica etc and I personally think the Hemisphere comes out on top, but I have worked with Hemisphere gear by the longest. In all honesty, in open sky conditions, any of the new receivers from any of the big manufacturers are going to be able achieve accuracy within 8 mm in your described conditions. What makes the difference now between receivers is how well they perform in difficult multi-path environments. Feel free to reach out to me if you have any questions, or are looking to talk to someone about this gear.
Nolan
I would assume that the BRX7 can get you what you are seeking repeatability wise with correct field procedures and conditions. It specs like Trimble and other leading brands. Now I will say canopy is different but if I am reading what you are wanting to do with being in the open then in that scenario I doubt it will not achieve what you are looking for but I doubt that with one shot. You have 8mm horizontal x 1 ppm. But with a couple observations given a gap in time of say 3 to 4 hrs pretty easily done. But don’t sweat the difference in those two observations at that gap being outside that range a bit even the simple mean of them gets you under that at times. Carlson has least squares and if I were running one today I would follow what I am doing with the Trimble. I set a base and run through my control wherever I need to set it. Move the base or a different day different time and run through it again. Any optical work total station or robot. Comes off of or resections off the gps control for tough areas and ties into other gps point or points. I build a mini rtk and robot network. Then perform least squares. We did one the past few weeks a 40 acre parcel Topo boundary. My uncertainty is in a couple hundredths range horizontally and max vertical is .04. That’s using the best of both worlds rtk base rover and robot. Lots of Topo in trees so short mini traverses all through the woods. But I use all the data and get rid of blunders.
Yes I found a number of your videos online when I was looking which were interesting. Some good work right there.
What I couldn't find much about when I was looking was repeatability of observations, not just dump and fix while standing at a point, but coming back 20 mins+ later and comparing with first observation. I find the 8mm+1ppm RMS on the spec sheet not any practical use, it's the real world 95-99% confidence interval stuff was I trying to find.
I've got Geomax gear (I'm down in NZ on the other side of world) and the specs of the Zenith60 Pro they've just released look suspiciously similar to the hemisphere board. Given Carlson are also in the Hexagon tent figured it wasn't too much of a stretch to assume they'd used it. The Zenith60 Pro has the excellent 1w Satel radio too, can't work out what the BRX7/S631 use though?
I'm never taking 1 epoch shots with GNSS, at least 3-5 for topo. For control/boundary point is 20+ epochs, dump, rotate pole and another 20+, wait 20mins and repeat.
Interesting you are still doing the topo under canopy with robot, assume that's because too slow with GNSS under canopy waiting for quality shots and then verifying them? However, most of the 'canopy' videos I found from North America don't look like the pine forests we get here with RTK radio eating needles.
So sorry. I did not mean to elude 1 shot as 1 epoch. I do 180 epochs for control. That’s one observation. Then 3 to 4 hrs later re observe that’s the 2nd shot.
As far as Topo in canopy. It is a little of both. One it can be slow and the other. Is loosing and gaining a resolved solution is iffy at times. Plus in a Topo shot you really only have a few epochs of data and so to me I have personally done so much testing in canopy that when you get a fix then record the Topo shot and all of a sudden it looses that fix. Then gets it again just doesn’t make for trusting the data.
I will say that I have run head to head on a site nasty canopy with a BRX7. What I noticed is if it gets fixed it holds. Now when I am running right beside it and see it going in and out with all the testing I have done I didn’t trust it and it showed in the results when we came back for 2nd observations. And to check the end results from independent check. So it really depends on the canopy and project requirements before I Topo that way with gps. I prefer to take a gamble and set control throughout the canopy with gps and then just come back and Topo from that with robot or whatever.
Time wise if you set up a robot and can get most of the data from one to two setups the robot wins. Vs gps rtk. Even in open. It’s really about that line of site and 2.5 to 3rd set up with a robot that rtk starts to win as only one set up for Topo line of site to base is irrelevant. That all depends on how quickly a crew can set up over a point and start Topo shots with instrument. Some I see takes 15 plus minutes to even get over a point and level so by the time they have back site and all its 30 minutes. Some can do it in a few minutes. Of course also all this depends on obstacles being in and around the area as well.
Hi Luke Carlson and Hemisphere are not under the Hexagon tent. Carlson is actually a stand alone private company who buy receivers from Hemisphere, while Hemisphere is under the Case New Holland umbrella now. GeoMax is a Hexagon brand, technically under Leica as a discount brand if my understanding is correct.
As for the Zenith 60, I know the base model uses a NovaTel board. The Zenith60 appears to use a Hemisphere board (based on the SureFix claim under RTK Mode), however I am not sure if it is the Phantom 40 (the one in the Hemisphere S631) or the Phantom 20 or Phantom 34 (which are a little bit slower but cheaper).
As for the repeatability videos, I am starting to lay the ground work for a more detailed over time repeatability test for people. I get a lot of requests for a more detailed video, and am starting to think about how to best go about it that is actually entertaining but has the information people need. It is logistically a bit of a difficult one to film without losing several days to just running tests and filming. Unfortunately, time wise for me this is best done in the winter, where in Canada (where I am based out of) I have no leaves on trees which is what most people want to see.
Yes - I like your videos where you compare time to fix of different antennas. I wish you would then compare GPS NEZ's with the location verified by a gun.
Getting a fix is great, but is the solution acceptable???
Hi Hoosier, I agree with you, I have some plans to do some NEZ tests in the future (and even some total station vs gnss comparison stuff). I am working with a new video guy he is helping us get more stuff out, so hopefully I can do something over the winter that fits into this, as it is definitely the most requested video.
I stand corrected about Carlson, not sure where I picked up that erroneous idea.
Yes the Zenith60 uses the OEM719 board which is what I guess the GS18/GS16 use although I'm sure you'll never get a straight answer out of Leica whether they have their own firmware or just repackage the Novatel firmware. I've currently got Zenith40's which also use the OEM719 board and seem as good as the GS18's I also use in some contract work I do, but no tilt.
As you noted it is the surefix, channel numbers etc. on spec sheet that made me suspicious the Zenith60 Pro has switched to Hemisphere board. Been very familiar with the Novatel boards so just trying to learn more about Hemisphere board.
Thanks for all your useful comments!
Heard back from Geomax, the Zenith60 Pro is using the Phantom 20 board. Looking at data sheets for the boards all I can see that is different is the size and ports but sounds like you have access to more info on the backend.
Hi Luke, interesting, good to know. I have not done a ton of testing on the P20 specifically, but my understanding is the P20 v P40 board are an indication of the number of pins on the board. This translates to how much data can be moved around. In theory this means may not be as good as the P40 in the really tough environments, but I can't confirm that. They are still very good boards, with the same RTK engine as the P40. I will try to dig around and see if I can find anything more specific.
@bm-nolan Hi Nolan Nice to see you here. I am also a loyal fan of hemisphere gnss. Recently, I want to know their core technological advantages in drones. I am studying aerospace-related knowledge recently and I feel that this company's technology is amazing. Can you help me?
@elieen-koyeela Feel free to shoot me an email at [email protected]! I'll see if I can scrounge some stuff up for you.
Hi Everybody,
I realize I am very late to the party, however, perhaps I can lend my personal experience with the Hemisphere S631.
A project requiring the setting of 256 lot corners. Houses were designed to "just fit" on each lot. Previous control was set by another firm using conventional equipment . (cut to the chase here) localization great. Base set on the safest point. unit orientation of base unit observed. Project required several days.
Each successive day, base set and oriented and control checked with Rover oriented the same. In ALL cases, the residual was less than 2/3 centimeter. Due to rains, part of the project was completed two weeks later. Check in residual was the same. Sure fix and auto tilt was employed. Rover and base units were oriented in the same direction for ALL days.
Short story long, the S631 units performed with excellence. The site; wide open sky. Base station was set on a tripod each different day. A different "HI" was observed,, and yet the S631 kept up it's diligence.
This is my experience with the S631 units requiring very tight repeatability.
Thank you very much
Thanks for the comments. I think most GNSS (even the old GPS only units) can usually check in around 5mm hz and 10-15mm v for repeatability (assuming poles/tribrachs are in excellent adjustment) in the wide open its around vegetation/canopy/buildings where the newer GNSS is being tested to see how well the repeatability holds, i.e., checking in on the same points several times using same methods not just the precision estimates on the data collector.
Do you have any experience of using them around canopy/buildings etc.?
@lukenz Experience Under Canopy and near buildings, yes. Buildings first: The North side of tall buildings is challenging. East, west and south result in good repeatability and offer no major problems.
For canopy: One story only to save time, I once had to set a point under a very large umbrella pine tree. My S631 held a fixed solution with about 15 birds and sent me to the point. Being a non-trusting individual, I simply set a 60d Nail, then wet out into the open and set three control points.
Set up the total station and checked into the control points very well. 0.02' then swung over to the 60d nail and actually hit the head of the nail. I realize they are large nails, however, it was a big grinning "YES" ! moment. Needless to say, I still don't trust anything and always double check.
Another quick story, (sorry) I had to locate a monument that was shaded by large ash trees. My Leica owner buddy was along to do a live job comparison. He went "under cover" first but could not get a fix after five minutes. My S631 rover lost a fix but regained it after about 30 seconds.
It guided me to the "way point". set paint dot, Then we dug and found the monument My Leica Buddy was totally amazed.
I am very happy with the Hemisphere S 631 units. As with any units, care must always be taken. Also, watch out for Geo Storms, they can dissolve precision.
Thank you (^_^)
I have been using a Base Rover pair of BRX7s for a couple years now, and have had great repeatability under 0.07' in pretty dense canopy. Come back a different day and still get the same answer. The SureFix rocks. Love hearing how others are using the tech. Thanks, all for sharing.
Come back a different day and still get the same answer.
Is this checking against total station work or just GNSS repeatability? Have had GNSS give the same wrong result twice before with same multipath etc.
The North side of tall buildings is challenging
Be the opposite for me I suspect down here in NZ, we orientate our houses to face north for the sun here. Makes sense with the concentration of satellites in mid latitudes.
good deal mate ! yes, your south side is the trouble zone. Decades ago, I ran single frequency Trimble receivers. They did well as long as they were reading 6-8 satellites. the early Trimble single frequency units were subject to a "jump" occasionally. For all control, they suggested taking the reading, dump the unit then initialize on that same point. If it failed then we'd have to return to a previous good point, re-initialize then proceed. Those heavy units were great in the wide open. I staked many subdivisions in dense fog. Contractors were amazed to discover that stakes were in the ground and marked as they arrived after the fog lifted. The men were so grateful to be working on a foggy day, that they'd account for the vertical jumps these units would output. (lasers etc) Reflecting back, I think those old single frequency days instilled solid checking methods that may be less required with todays 7th generation GNSS units. I really must hand it to Javad Ashjaee for pioneering the use of multiple constellations for receivers. [RIP]
Have a good one Luke