TN had service from TDOT, but you need to be one of their contractors doing road work. And my farm is right in a gap anyway so wouldn’t help me. I appreciate your offer to build but I think the more I investigate I think paying the TORCH subscription $8/month might be my best option. Thanks for sharing your build. I bet a lot of people would like to know more about it, as more devices need accurate location (Self driving cars, our home help Robots, [Rosie from the Jetsons]) Please keep sharing as you learn and decide next steps, it’s been very helpful to me.
I’m just a humble surveyor trying not to pay copious amounts of money when doing side work and not having the nice 30k rovers I use at my city job. My boss is jelly because he bought a used topcon gps for 8k and the postcard beat it 
2 Likes
@blackmanops I see this notion about tying in and getting great coordinates but the question I keep scratching my head with on this forum is why? Typically we want survey coordinates in state plane for gis. If you want to survey having the base be within even 5 feet for its coordinate is fine. Your rover measurements being relative to it is more important and that’s why you don’t want to venture too far from a base (greater than 5km) that is what most of you seem to be trying do get good results from. GIS technicians would love if everything could be tied to coordinates. But do I care if the property I’m recovering monuments on and doing retracement of checks to benchmarks within inches or do I care the pins are measured relative to each other and that is tight? Same with topo measurements.
The other point I would like to say is a good method of measuring on your property, farm, whatever is do what contractors often to drive an object of some kind into the ground and set your base on it so that you can repeat it each time and it’s very sturdy. Then your origin of your survey is always correct. Understand geoids and local datum planes or low distortion projections because for instance we have low distortion projections in Oregon that if working in the zone grid always equals ground distance. There’s a lot more than turning on a base and rover and producing dynamite results. Sparkfun devices can produce great results but you only will get out what you put in.
You are correct. I care more about relative accuracy for the local grid I am surveying and not exactly what the earth model Lat/Long coordinates are. My understanding is that the GNSS based accuracy is this relative accuracy, but based on what you say I may be mistaken?
Let’s say for example your base is 10000,10000 but the “true” coordinates based on the reference frame is 10005,10005 if you’re using a base rover and you set that base on your point 10000,10000 everytime it doesn’t really matter that you are off positionally because all those other measurements are vectors off the base and relative. One thing that does get computed based on these coordinates is orthometric heights which is ellipsoid height + h from the geoid. This is determinate on your lat long and the model. Heights and models of heights are why even if all your measurements done by gps claim accuracy of let’s say .1’ they might not be due to inaccuracies in the geoid.
If verticals aren’t super important then gps can probably do what you need it to. If verticals are important then you’ll need to use a level or other tools to get accuracy.
This is why in surveying we topo say an open field but not a roadway with gps.
I happen to be using ORGN as a reference as well, just with a slightly different setup based on um980 (it’s a little more predictable than “Postcards”, even if you buy um980 board from Aliexpress for less than $100) with a very good surplus triple-band antenna. Getting horizontal accuracy of 0.02m and better (that number is not what the receiver reports, that’s the actual result of measurements taken at the survey marker).
What you essentially get when using RTCM corrections is a precise and accurate vector (distance and direction) from the base to the rover. So your positional accuracy depends on how accurately the other end of this vector is located - how accurately the coordinates of the reference station were determined.
However, there’s a trick. Since the American continent isn’t sitting in one spot, geographic coordinate systems here are “nailed” to its surface, not to the Earth as an entire planet. That’s essentially what “NAD83” stands for, unlike the WGS84 that you hear about in the context of GPS measurements the most.
So when I hook up my receiver to OGRN NTRIP, the closest reference station tells my receiver its own coordinates in NAD83. So everything I measure with my rover receiver is referenced to it. And if I want to, say, upload my points to Google Earth, I need to either explicitly tell it that it’s NAD83 (so it doesn’t assume it’s in WGS84) or translate it to WGS84. Otherwise, there’s a shift of several feet. So when doing any kind of RTK measurements, you must be aware of how the base station coordinates were determined and what they are referenced to.
Here’s a little more technical description of how translation between different CRS works Dzertanoj's Diary | An infamous "NAD83 to WGS84" affair | OpenStreetMap - I didn’t write it as a tutorial, but it gives you an idea.
We’re spoiled that ODOT gives us ORGN for free. RTN they offer now is even better than plain old RTK. We’ve calibrated it to 83/91 and it checks solid. The postcard vs facet helped me with heights more than anything.
Well, if there were no ORGN access (which seems pretty justified to have, just like with data that NASA and USGS offer because it’s paid for with taxes), I would have invested in setting up my own reference station (possibly the same way as some of OGRN stations are built - at some local airport). I’m not a surveyor but a former ECM/EW specialist, so “good antenna makes more than a half of a good receiver” is something I follow strictly, even though my stuff is all surplus (very reputable despite being cheap, like WL Gore, General Dynamics, Amphenol) from eBay.
I tend to be a little skeptical about Quectel LG290P because of how many instances of various firmware-related discrepancies are reported by people who use it.
I have only found the need for firmware updates because RMS used to be broke. I’ve checked numerous NGS and lane county benchmarks and been accurate and repeatable. It is just young in its nature.
Great Conversations !
I own all (4) of the SparkFun RTK products mentioned Plus the other (2) which are the RTK Surveyor and RTK Express. Of the (4) mentioned in this Thread- each one of them are better at certain particular applications.
The Surveyor and Express were how I began my journey into the U-blox F9P chipset.
The F9P was what convinced me that inexpensive hardware could complete with the performance of my expensive Professional Equipment.
The obvious progression was the Facet with a cleaner form factor, and later the Facet L-Band.
Then comes the Torch, which isn’t related at all to the previous SparkFun hardware. It uses the UM980 chip, a much smaller helical antenna, and adds the L5 frequency.
And now the most recent, the PostCard with the new LG290P chip.
But it’s actually impossible to rank the (4) devices in any order.
The $700 Facet is a proven winner. Coupled with an appropriate correction source (just like any RTK), it’s a beast.
The Facet L-Band is like an Easy Button for GIS work, we don’t need to teach GIS technicians anything about RTK, corrections, etc.
The Torch has a very slick form factor and adds L5 that is beneficial around tree canopy…
And the recent PostCard is impressive, considering the price-point.
It’s brand new, and the LG290P FW from Quectel needs a little bit of time to mature.
To muddy the waters even more, several new Products are now in the Source Code 
I own a lot of SparkFun GNSS devices and I don’t regret purchasing any of them. They are tools for the toolbox. There isn’t a perfect device for every application, but I can use any of the 4 models.
Just remember, you cant outperform your antenna or your correction source.
If I’m going to a Job Site where a bunch of folks will be, I’m generally taking the Torch.
If I need to PPP a new Control Point in a new area, a helical antenna (Torch) isn’t my first choice.
But if the Government banned all RTK GNSS and came to repossess mine, I’d hide a handful of PostCards in my pocket and be an Outlaw like Jesse James
1 Like