Ok so i think figuring this out is on the list for winter projects. I want to see if i can pull off this trick for a 24hr period.
So RTK Extend apparently gives you RTK accuracy for two weeks with no corrections after only a 20min RTK session. I know it sounds like some heavy voodoo.
So the first step to solving this is i have to figure out how to store and replay old corrections, I know this is not the solution but i think its part of it.
Next i think i need to see if there are any generalized patterns in the corrections that can be used to calculate and modify the stored corrections as the originals degrade over time. This is under the assumption that the ionospheric changes happen like waves and head in a generalized direction.
Also I have a trimble that can pull off a moving base, I will see if the deere system has to do with comparative error from a second tractor mounted “moving base” receiver.
Then mow my lawn in the spring and see how bad my math is.
So does anyone else have any guesses or theories how you can pull off a two week disconnect and maintain RTK accuracy?
Yes that’s how it works , when we are working in a field with no cellular reception the rtk extend kicks in and we still have straight rows . How they do it is beyond me.
Does the gps unit need to remain powered or can you restart the tractor and still be in RTK extend?
I have thought of another method as well. They may be recording the RTK correction differential vs SBAS (waas or sf3) then those differentials are used to augment the live SBAS corrections. But sf3 has convergence so I am unsure this is it either.
I did some digging and now it makes a lot more sense. After reading the first and second generation Deere RTK extend manuals i have a pretty good handle on whats going on.
Fancy marketing.
RTK extend is fallback to SF3, much in the same way XFill is fallback to RTX. Thats why the gps unit always stays powered so you do not lose your convergence.
Deere is just hoping farmers do not hop out of the cab to measure. Really any guidance under 2” pass to pass seems really straight enough for 98% of farming. SF3 is seasonally repeatable so for most operations you would not notice much difference between RTK and fully converged SF3.
The only part i think may have some special magic going on now is the shorter 15min window before sf3 fallback. I think this may have the SF3 signal tweaked to live conditions compared to the local RTK corrections. This is cool if its happening, but I have no way to test without sbas support.
The M2’s float is scarily just about as good as waas, without any sbas support.
Seeing as though RTK Extend is less science and more marketing ploy, I think its not worth the energy to duplicate. The M2 gets massive signal, and does not drop fix easily.
