I’ve got an RTK Express (with the L1/L2 antenna) and it works great, the only downside is that every time I move it, I’ve got to leave it in place for 3-5 minutes before the signal “settles down” and I can get 1 foot accuracy or better (which is plenty for my needs). The area where I’m using it (Allenspark, Colorado) has -zero- cell phone coverage and -zero- WiFi. It’s high altitude and not very tree-dense so there shouldn’t be any problem getting the satellite signals.
(1) is that amount of time expected, given my setup?
(2) if I get a second RTK Express and use them in a base/rover configuration, will that speed things up? There’s an official survey point where I can put the base unit and make all my measurements can be relative to that point.
(3) how accurate are the elevation measurements I’m getting? Again, I could use the elevation at the official survey point as a known value and only use the rover for elevation values relative to the base.
Right now, you use your RTK Express in the middle of nowhere, without differential corrections. Correct? What dynamic do you have selected? Are you in rover mode or base mode? What do you mean by the signal “settles down”? What kind of fix do you get? What is the accuracy announced by the RTK Express?
(1) is that amount of time expected, given my setup?
Yes. The device is resolving many unknowns without external corrections.
(2) if I get a second RTK Express and use them in a base/rover configuration, will that speed things up? There’s an official survey point where I can put the base unit and make all my measurements can be relative to that point.
Tremendously. You would see an RTK Fix within a few seconds of receiving corrections from the Base. Your accuracy would be excellent (sub 100mm or better). But note that official survey marks often require a bit of work to get their modern lat/lon. See our work here using an NGS mark.
(3) how accurate are the elevation measurements I’m getting? Again, I could use the elevation at the official survey point as a known value and only use the rover for elevation values relative to the base.
I believe there are some other better posts by @rftop on the forum diving into the science, but elevation is the worst of the three (lat/lon/elev). As a rule of thumb I double the reported HPA. So if you’ve got RTK Fix with a reported 14mm accuracy (this is achievable with an RTK Express) then the elevation accuracy would be ~28mm.
does SparkFun even sell a radio anymore that would allow the base and rover to communicate?
Depends on the distance you have between the Rover/Base. I recommend starting by using the internal ESP-NOW radios to get your setup working. Serial radios are the next step up. After that, LoRaSerials work but require more configuration work than ESP-NOW or telemetry radios.
Assuming that the survey point has a datasheet (adjusted control), this would be good head-start for occupying that point with a Base. The RTK Rover Solutions would inherit the same Reference Frame, Epoch, etc as the control if you use the published coordinates at your Base.
However, I don’t see many good choices for the area though on the NGS Data Explorer
If you’re not talking about 1 of the 3 Triangulation Stations, then you might as well Survey-In a Base and overwrite the Elevation from the datasheet for a “vertical-only” tie basically.
Considering your location/situation, you also might consider the StarLink thread for RTK via NTRIP ?
Some followup - first of all, as I was thinking more about the situation I’m working on, my most pressing need was to just grab some relative elevations for a driveway, and for that I just needed a laser level.
But then next spring I’ll need to do some more serious “surveying” (for which I don’t care about elevation) and getting a second RTK Express and a pair of serial radios sounds like it would be good. And that conclusion of course leads to more questions -
(1) What makes that F 326 point a not good choice? I know that NGS data points exist (I’ve used Tessar which is about 3 miles from the SparkFun HQ), and I know how to convert between reference frames, but that’s about the extent of my knowledge there.
(2) There is a surveyor rod with official cap showing the corner of my neighbor’s land, I had been planning on using that as my ground truth. Is there any reason to try to use F 326 instead of, or in addition to that rod? My thought is that all the neighbors agree on that point being valid, so even if it’s “wrong” using F 326 as truth, I have no reason to fight it.
Thanks!
I had to confirm with Google, but Colorado uses the Public Land Survey System (PLSS) for land description. In the 1800’s the State was divided into Townships & Ranges, with Sections being carved out of each. Most of the official marks were likely piles of stone with the top rock having a chiseled “X”. Individual Land Parcels (Deeds) use Metes and bounds descriptions, but it will start at a PLSS corner first. So yes, a multi-million dollar parcel/lot is legally tied to a pile of rocks that some guy placed in the 1800’s measuring distances with a 66’ long chain.
You are correct to consider the Surveyed Property Corner as the Ground Truth because it marks/monuments the physical location in the PLSS, but that property corner and the Benchmark F 326 (NGS monuments in general) have no relation to each other.
OK so trying to figure out how to take the deed and turn it into GPS lat/lon (a tricky venture, I know). The deed says the east edge of my property starts at such-and-such a point and “thence north, 953.60 feet”. In a professional site plan I had done in 2023, that edge is listed as 953.6 feet long (so far so good) with heading N02°15’52”E so I feel like that’s telling me due to continental drift and some guy in the 1880s with a long chain, I need to take all the headings in the deed and add 02°15’52” to them. But then I still can’t use those headings to calculate lat/lon since I need to account for magnetic declination (right)?
BTW, my goal here is to measure the lat/lon of the property corner using my RTK and then use math to calculate the lat/lon of the other points.
In this situation, I’d normally just draw the parcel in AutoCad according the to the metes and bounds calls. You can start at 0,0. Then RTK the easy pins/corners that you know where they are. Import those points into CAD and shift & rotate the parcel to the real-world ground points (localization). Now you have decent lat/long for the entire parcel, in order to physically recover and RTK all pre-existing iron in the ground.
