AZURIT FULL ODOMETRY MOUVEMENT CONTROL

[bernard]

Hi.Roland.
I use the green one (Paranello) as dev Platform because all is very easy to access.
This one, yellow(DENNA L600 ) is the first i use when i start this project and have until now the old PCB1.2 Inside with DUE.
2 or 3 weeks ago i change it for the PCB1.3 and now i have Dev and working Platform with same board.

But with the new PCB1.3 i don't understand why but in a particular position (not in the center) into my area the mag become >0 and the mower make an IN OUT transition.
Also when i upload a sketch the Buzzer ring and not on the second board ????
I have change the DC/DC for the 5A one and put it on the rear of the mower so now no noise on the coil but the problem is always here ?????.
I have made some video of the noise with the 2 DC/DC at multiple distance and the result show that the original DC/DC is very noisely.
Certainly i test the green one with your sender signal into this area to see if there is a difference.

In all case the By lane work very well on small distance so for me a good alternative to ROS and full mapping can be a parcelling of the mowing area.

Concerning your Raindancer soft i am very happy of the motor management and normaly the Behaviour Tree (If i can understand all but i have time).

Something very cool with raindancer will be:
Add the GYRO MPU6050 for perfect straight lane.I think the ODOmetry is already accurate for distance mesurement.
A PI to manage the GPS,define a (5 Meter accurate Mapping) and send command to mower to start to mow at parcelling Area.
2 Months ago i put RFID tag under the ground on my perimeter too see if i can use it as beacon during the Tracking (like your escape solution but you can know exactly(+or- 5Cm) the position of the mower).All my test on table work a 7 CM ,but in the groung and 2 or 3 years later ????

So time to work on this!
By.

 

[bernard]

Hi.Roland.
I use the green one (Paranello) as dev Platform because all is very easy to access.
This one, yellow(DENNA L600 ) is the first i use when i start this project and have until now the old PCB1.2 Inside with DUE.
2 or 3 weeks ago i change it for the PCB1.3 and now i have Dev and working Platform with same board.

But with the new PCB1.3 i don't understand why but in a particular position (not in the center) into my area the mag become >0 and the mower make an IN OUT transition.
Also when i upload a sketch the Buzzer ring and not on the second board ????
I have change the DC/DC for the 5A one and put it on the rear of the mower so now no noise on the coil but the problem is always here ?????.
I have made some video of the noise with the 2 DC/DC at multiple distance and the result show that the original DC/DC is very noisely.
Certainly i test the green mower with your sender signal into this area to see if there is a difference.

In all case the By lane work very well on small distance so for me a good alternative to ROS and full mapping can be a parcelling of the mowing area.

Concerning your Raindancer soft i am very happy of the motor management and normaly the Behaviour Tree (If i can understand all but i have time).

Something very cool with raindancer will be:
Add the GYRO MPU6050 for perfect straight lane.I think the ODOmetry is already accurate for distance mesurement.
A PI to manage the GPS,define a (5 Meter accurate Mapping) and send command to mower to start to mow at parcelling Area.
2 Months ago i put RFID tag under the ground on my perimeter too see if i can use it as beacon during the Tracking (like your escape solution but you can know exactly(+or- 5Cm) the position of the mower).All my test on table work a 7 CM ,but in the ground and 2 or 3 years later ????

So time to work on this!
By.

 

[roland]

Hi Bernard,

"But with the new PCB1.3 i don't understand why but in a particular position (not in the center) into my area the mag become >0 and the mower make an IN OUT transition."
With the PCB1.3 I have also fight with noise which I didn't had with the Nucleoboards an Sabertooh driver. This noise comes up when I start the drive motors, which are now 40cm away from the coils.
I will first redesign my chassis and put out the metal and put the receiver direct on the coils and then will see what happens. But after some investigations, I think there could also be a crosstalk on the PCB1.3. If this happens between the left and right coil signal, then it is important, that both coils are connected with the same polarity. If this happens with running motors, then there is no easy soulution. I am not sure about this at the moment. I have to order another PCB1.3 to investigate this. Have you errors on the perimeter signal, when your drivemotors are running?

"Also when i upload a sketch the Buzzer ring and not on the second board ????"
The buzzer also rings on my board while programming.


"...So time to work on this!"
Yes, very nice. If you want to do changes to the code, we shoud find a way (maybe with github fork), that you can put in your changes and use also my changes.
I will support you where I can. At the moment after finishing escaping obstacles on the perimeter, I have to do some improvement to the behaviour. Because in the last two month I realized two minor problems with this. This I have to fix first.