CASE MAXXUM Can Bus steering issues

Hello,
I was using AOG well with my previous tractor (JD6620) with phidget motor. RTK base station at home.
I changed tractor and now have a CASE MAXXUM 125 autosteer ready.
I built Tony’s board with Teensy and connected everything well to both can busses on the case. Set up in service manager for Case. AOG 5.7.1, latest ino from tony’s github. UDP without router
I can get all infos from tractor : hitch height, wheel angles, autosteer button for activation, … It’s integrated in the tractor as an official commercial system which is really great !

But…, I can’t get it to follow the line.
When starting on the line (AB straight) it’s making wider and wider S around the line.
It overshoots and then corrects sharply to the other side, then overshoots and corrects sharper, …
It should turn wheels the other way when getting to the line but it keeps crossing the line and correct too late.

AOG 5.7.1 :
PP, Acquire look ahead : 20, hold look ahead 2.4, look ahead speed gain 1.2, integral 15
Motor : proportional gain 25, max limit 164, minimum to move 5
Rtk fix, WAS zero adjusted, counts per degree as well. Tried with Ackerman 100 and adjusted to 84
All tractor and antenna parameters are set. Antenna on the front hood; rtk fix.

Tried with 5.6.15 which is working for Emmanuel, with his adjustments and all the same, not working for me.

When adjusting manually wheel angle, it’s fast, precise and reapetable.
tried to adjust PP values but still not good…

Looking at curves, bothe curves are very close. It looks like tractor is following precisely the target angle but target angle looks sleeping…

I don’t know what else to check or to adjust, has someone an idea ?

Thanks for your help

2 Likes

Hi Julien,
you should look after the counts per degree, in my case (Massey Ferguson) the wheel angle was not correct causing overshoot. I drive with stanley. Works good so far.

1 Like

Hi,
I have the feeling counts per degrees are correct.
I used the « record » function to measure the real angle and set the cpd to match the real angle.
I did it the other side and adjusted the Ackerman to come again to the correct wheel angle between measured and displayed on AOG.

I tried several things, including reseting Tractor WAS and Spoold valves.
I kind of managed to reproduce the same issue with simulator (even if it is maybe not that).
I have the antenna on the bonnet (323cm from rear axle) and if I tried to set Hold Look Ahead to 3.5. It was like if the reference point of the tractor was the antenna instead of the rear axle, and the antenna was so close to the target point (look ahead) that even a small deviation was causing a huge turning and going always worse and further away from the line.
As my max steering angle is small (27°) it can not short enough to catch back the line.
By changing the antena position in tractor parameter to 50cm (on simulator) it was much better and worked correctly.
Has anyone already faced the same ?
Is it possible that I have somewhere on the laptop an “old” config file with wrong parameters stored ?
(5.7.2 with latest ino from tony for teensy and CAN)

What happens when you set antenna on the roof and 30 cm infront of rear axle??

I have posted it before, and dont remember where I saw brian write the info years ago, that SIM expect you to set antenna above or up to 10 cm in front of the pivot point.

What happens when you set hold look ahead to 6, 7 or 10?

I tried this morning, it is worse with antenna on the roof at 30cm in front of the rear axle. Better with antenna on the bonnet.

Cannot set it higher than 5, but the goal would be to decrease it not to increase.
Further front it is, less reactive it is

Did some other tests this morning.

With look ahead = 5, it is not bad but not reactive at all, it takes a lot of time to be on the line and cannot follow uturns.
With smaller look ahead it cannot stay on the line.
What is strange is that real wheel angle is the same as the target. It follows very close with less than 0.1° error. So the tractors does what AOG tells.
But when looking at the “XTE” curves, as soon as I reduce the look ahead, it begins to move around the target and is never correct.

Hi Julien,
Did you try with different CPD settings (even if not real)?
The tractor seems to turn sharper than it thinks he does, so smaller CPD should help steer less (and less over correction)

Here is an explanation to understand cpd, by brian. How Counts Per Degree work

1 Like

I tried to reduce and increase CPD settings and to tweak wheel base. It is a bit better but I cannot reduce look ahead, it continues to move around the line.

I have the impression that AOG is just sleeping… waiting to cross the line to correct instead of anticipating when comming to the line
I checked my F9P receiver configuration, it is correct, I have RTK fix, less than 1km from base.

I don’t know where to look now ?

It is solved. it was an issue with GPS position i guess
I updated the F9P and put the latest configuration file and it works well now. I don’t know exactly what was wrong, but I guess the position was not the correct one from F9P and because of that the steering was wrong.

2 Likes