Despide the bad weather which prevents all work in the fields, I made a lot of tests on the road today to find the good settings. First time with the definitive tablet for the projet replacing my labtop that I used before for the development.
I think I am closed to succes: Video Fiat 115-90
The displayed difference betwwen position and line was 95% of the time under +/-4cm and 5% up to 5 or 6cm
After that, I made some test in a little grass field: all line was spaced of 4m, the tool width I program in AGO.
Now, we are waiting for good wather to go back in field and doing real work !
For thoses test, BNO was activated (very stable for me) and MMA unactiveted.
I notacied that when I activated MMA, the accuracy falls to someting like 95% under 6 or 7 cm and 5% up to 10 cm.
Any idea why and on which parameter I can adjust to improve accuracy with MMA activated ?
As all our fiel hillside, I need to activate it.
My configuration:
Surface Go with USB hub
L1/EGNOS TeeJet Rx370p Antena
Delphi ER10031
Adafruit BNO055 Adafruit MMA8451 Adafruit ADS1115
IBT2 and Phidget motor
Homemade PCB cards based on the old wiki schematics, with the following arrangement:
* Rooting adaped the actuel AGO INO pin assignment
* Work & steer switchs throught opto, like on PCBv2
Thank to @BrianTee_Admin and alls the developer for the software and to the more advanced user their help.
Also special thanl to @Tooki57 for his help .
The antena is placed on the top of the roof. As I got pretty good results like this, I haven’t try with the antenna on the front axle.
Do you think it can improve the accuracy when the MMA is activated ?
I think that the further the antenna is from the pivot point, the bigger the gps error. And therefore AOG will correct it earlier than with the error on the roof, and maybe improve the accuracy a little more.
I don t use mma or bn0, and positioning the antenna on the front of the tractor has improved my system.
Logically, I think that if the antenna is lower, the roll effect is less significant on the GNSS signal (especially for ground vibrations).
So finally, after being confined 2 months, I could spend a week in the farm last week.
The occasion to install a new steering controller box with PCBv2 and makes more proper wiring in the tractor.
Also trying the new box in the field. Unfortunately all cultures are established, so the only possible test was to follow the sprayer lines.
Work pretty well à 7-8km/h in PP. A little bit more oscillating arround 10km/h. Don’t know if I should move to standley at 10km/h ? Or perhaps the frenquency of the antena, blocked to 5 Hz, isn’t fast enought for hight speed ?
Vidéo:
Now, working on design and installation of a pinion and ring gear system:
Usually I prefer contactless sensor like inductive sensor. They are more robust and with higher number of switching cycles. It’s why I use a inductive sensor for steer switch.
But yes for the work switch on the hitch, those mechanical swith are really handy and easier to instal.
The issue with inductive / hall sensors is the relatively close tolerances required. Some mechanical linkages are just too sloppy for reliable operation.
Hello math,
Yesterday i did some mowing with the f130 and it worked pretty well but not perfect. I took 24 degrees for my max steering angle but i didnt measure it and im starting to think i made a big mistake here, what is the max steering angle of your f100?
Edit:
Sorry for the stupid question, just figured out the max angle is actually 40 degrees.
Unfortunately i cant get v5 working properly on my panasonic fz-g1, gps randomly jumps about 20cm sometimes but v4.3 no problems. Got the correct angle now and works better than i could have imagined.
When experimenting with settings in v5+ it can be a good idea to set antenna hight to zero, that way temporarily eliminate influence from roll.
Are your cmps fastened so it can’t vibrate?
I fixed the cmps in a aluminum box with hotglue and the box is fixed on the cab floor next to the seat, tractor has no cab suspension and roll seems stable. Read somewhere that v5 gives problem with the fz-g1 so for now im happy with v4.
.
