I think ideally a PoE ethernet serial server (5v TTL version) for the f9p, it can send UDP nmea and receive UDP rtcm (ntrip from AgIO). Then another ethernet serial server (rs232 version) at the 20/20 to convert the UDP nmea to rs232.
A tcp232-t2 works well but would need a PoE splitter/midspan at the f9p and the other one at the 20/20 would need a ttl-rs232 adapter. There are ready made PoE devices too. A PoE panda PCB could also add in IMU roll in place of the 1st tcp232.
This gets me thinking about the AiO Combo PCB I designed this winter. It costs about the same as the v4.5 Micro but is smaller, still supports dual Std or Micro F9P, gm982, rs232, triple can bus, esp32 wifi socket, 8 machine/section outputs, on-board Cytron & BNO socket. The idea is that with different firmwares it could be a jack-of-all-trades PCB, youād only need to stock/build one PCB design for many applications & only carry one as a backup. It doesnāt have provision for PoE thoughā¦ Testing & firmware coding is ongoing.
If your GPS antenna is mounted on the bar itself or on the lid of a row unit, roll shouldnāt be an issue as itās low enough to really not matter. It would be far more accurate, laterally, than roll-corrected tractor GPS feeding the 20/20 and having it estimate the trailer position.
But that is a totally different deal than sending from GPS_Out.
Maybe a solution would be Arduino code to convert Panda/paogi strings to gga/vtg. Then, that could either go on a teensy Panda board on the Implement, or on its own board on the tractor network, or on whatever aio board youāre using
If your antenna is 2.5ā off the ground, a 15 degree roll is about 6-8" of lateral offset. Pretty sure trailering estimation error and draft will be a lot more than that if you used corrected GPS from the tractor.
Also at that slope, your planter bar is also 4% narrower, at least if youāre driving top-down AB lines (which all our lines are). An 8 row 30" planter would be 8" narrower than on flat ground. hope that makes sense.
Yes that makes sense, but roll corrected tractor GPS works good enough for me, and an implement receiver is not worth the added complexity, unless I can gain implement guidance. If Iām using tractor GPS, my antenna height will be 10 feet, not 2.5 feet, and if Iām on my combine, my antenna height is something like 13.5 ft. with no trailing implement error involved.
Maybe Iām missing something, but Iām using a fairly old dual board that I believe does all roll calcs on board. I set my tractors antenna height at zero. So maybe GPS out would send roll corrected position already?? Just thought of it this morning.
Pretty sure dual GPS outputs the raw GPS position of one antenna, and also the heading and roll angle of the of the other antenna. AOG still has to process that roll angle and translate the GPS position down to the axle.
Hereās a copy of AOG from the uturn branch (as of today) modified to output corrected position along with a (slightly modified) copy of GPS_Out.exe from the RollCorrected branch. Note that the GGA/RMC output timestamp is currently just using the computerās Windows time but maybe in the future we can get the actually GPS timestamp from the original nmea sentence coming into AgIO. I donāt have a good way to test this corrected position output, maybe some of you āsouthern neighborsā can give it at try. Iām not sure how to test it other then trying it on a field.
We got the planter outside today and tried GPS Out. I didnāt have any problems getting it to talk to the 20/20, but Iām showing a difference in the degrees of roll between AOG and GPS out. Any ideas whatās going on?
Testing & firmware coding is ongoing.