Perimeter issues after implementing improvements

[K!NG]

Hi Folks,

After implementing improvements last winter and using the Ardumower several times; I am now seeing some issues with the perimeter wire signal. The Ardumower is mowing in it's area but sometimes catches "glitches" which results a detection as being outside of the area.

Last week I have monitored this behavior by plotting the perimeterMag signal. The glitches are clearly visible on a consecutive way. See also printscreen in attachment.


The improvements which I have done last winter:
- Mounted raspberry pi 3 on PCB 1.3 reserved area/slot;
- Added DC/DC convertor to supply the RPI3. DC/DC is located > 10 cm from the coil;
- Mounted the receiver + coil on a more professional way by using a 3D printed housing;
- Changed firmware from Azurit to AzuritBer (incl. commits from Apr 21, 2022)

I have crosschecked the signal code pattern. Both Sender and Ardumower are using the signal Code 1 - which is default. So this looks OK.

Is someone recognizing this issue ? Can it be a defect to the coil/receiver ? Bad contact ? OR is related to some disturbances coming from the DC/DC convertor (maybe from the power management chip of the RPI3?)

 

[bernard]

FIRST SAFETY , SO REMOVE THE BLADE
Into test odometry menu you can find 2 value "Peri Mag/Noise" the noise need to be < periMag /2 when mower is in the middle of area.

So to adjust location of cable DC/DC etc inside the chassis you can put the mower in the middle of the area without blade and the wheel not on the floor (free rotation)
Start the 5 turns wheel and activate mow motor , now move some cable inside the chassis and check the noise value to find the lower possible number

The issue also occur some time at the begin of the season with high and wet grass What append if you deactivate the mow motor into setting mow

 

[K!NG]

@bernard, thanks for your quick reply.

I have now disconnected the RPI3 solution. So back to default hardware without additional DC/DC convertor. While running it looks improved.

The only issue which I have now is that the Ardumower stops way before the perimeter wire. An average of ~0,5 meter away from the wire. Who recognizes this symptom ?

It's hard now to monitor as I don't have a dashboard anymore. In the coming weeks I will implement the ESP32 solution.

 

[bernard]

Into Arduremote Console .

 

[K!NG]

Hi Bernard and others,

Today I have checked the complete perimeter solution (Sender + Loop + Receiver).

During visual check of the receiver; I came to the conclusion that LM386 is dirty with most likely a faulty potentiometer due to the dirt. I have replaced the complete combination (LM386 + Coil) - and it's now fully up and running again including all the improvements which where implemented last winter.

The combination (LM386 + Coil) is in use for ~1,5 season. It's quite fast that this became faulty and I am doubting about the housing. I have kept the LM386 in a 3D printed plastic housing with the coil sticking out. Also the last couple of grass cuts' where really wet here in NL - which might caused a lot of dirt.

How do you guys keep the combination away from dirt/dust/wett grass ? Are there other robust solutions to keep the receiver clean during harsh conditions ?