The Combine Forum banner

3721 - 3740 of 4139 Posts

·
Registered
Joined
·
45 Posts
My steering orbital is a closed centre load sense (ccls). I can tell that as it has 5 hydraulic hoses attached - pump, tank, left, right and load sense.

The hydraulic valve is one which suits this. If you look at the AgOpenGPS wiki under hydraulic steering this information should be there.
 

·
Registered
Joined
·
376 Posts
Not quite sure what you are looking for. AOG should be able to parse the section/switch data from serial = Usb or Bluetooth and UDP. At least it did when I checked in February. The sentence is the same.
I'm looking for AoG to send the section control settings sentence to the Arduino via UDP. So far, I'm only receiving the section control data. I made the change and now I also receive the settings sentence.
 

·
Registered
Joined
·
5,856 Posts
Discussion Starter #3,723 (Edited)
I've been having a test of StanLee. Very smooth. I think I've found a bug though? That, or there are settings I am completely missing.
StanLee working very smooth in 'steer axle ahead' config, but when you set up a vehicle for a 'steer axle behind' like a header, the correction doesn't kick in until the rear of the vehicle has reached the desired path, by which time, the front/implement is way off course. Is there a way that StanLee can 'lead from the front' even when the machine is rear axle steered?

Correct it does not work. I tried this a few days ago, and it doesn't work in pure pursuit either. You are also correct that the header will be considered the steering axle and then steering then would be set to turn the opposite way. I'm going to have to figure this one out - especially now there are 2 methods of guidance and several instances.

Edit: I've set it up so steer behind, antenna behind, rigid tool, tool ahead, hitch at 250cm. So just like a combine, but if your antenna is ahead of the drive pivot wheels, it should be ahead. Anyway all I had to do is set the steerangle setpoint times -1. I'll send an update on github this evening. It will work on Curve, AB Line, Contour - but only in pure pursuit. I'm going to have to figure out how to do this in StanLee. Maybe it can't be done, I hope it can.
 

Attachments

·
Registered
Joined
·
1,340 Posts
Is it possible to get more then 250cm for a front mounted sprayer? Any more and it gives me an error.
That should be in inches, I think its 270 inches from my pivot axle to the boom.
 

·
Registered
Joined
·
31 Posts
Correct it does not work. I tried this a few days ago, and it doesn't work in pure pursuit either. You are also correct that the header will be considered the steering axle and then steering then would be set to turn the opposite way. I'm going to have to figure this one out - especially now there are 2 methods of guidance and several instances.

Edit: I've set it up so steer behind, antenna behind, rigid tool, tool ahead, hitch at 250cm. So just like a combine, but if your antenna is ahead of the drive pivot wheels, it should be ahead. Anyway all I had to do is set the steerangle setpoint times -1. I'll send an update on github this evening. It will work on Curve, AB Line, Contour - but only in pure pursuit. I'm going to have to figure out how to do this in StanLee. Maybe it can't be done, I hope it can.
Gotcha. Well, I'm finally doing a build on a combine in the next couple of weeks so I'll keep testing the devs on it. So that's flipped the antenna position around, correct? I'm thinking about placing the antenna more towards the engine block so it'll pick up changes quicker being closer to the steering. But I would have to put 'pivot axle point is behind the antenna', even though the drive/pivot point would be in front of the antenna?
 

·
Registered
Joined
·
87 Posts
Just a little post to say I greatly miss the headland feature... it was my visual tool to know where to on/off sections and do the outer round(s) afterwards when spraying and seeding green manure crops and cereals. Which I will be doing a lot this summer and autumn.
I understand it is not so easy to re-implement because of the way the inner and outer boundaries are now handled in the code?
 

·
Registered
Joined
·
5,856 Posts
Discussion Starter #3,728
Gotcha. Well, I'm finally doing a build on a combine in the next couple of weeks so I'll keep testing the devs on it. So that's flipped the antenna position around, correct? I'm thinking about placing the antenna more towards the engine block so it'll pick up changes quicker being closer to the steering. But I would have to put 'pivot axle point is behind the antenna', even though the drive/pivot point would be in front of the antenna?

I tried it in the tractor yesterday by just driving backwards and it seems to work really well. All i set was steering at the back and antenna behind pivot - code unchanged. The thing with rear steer i found was the to move the look ahead point way longer and turn down the counts per degree, like if it was 30, to like 20.



Kinda makes sense when you think about driving in reverse how much more twitchy it is, but i think a combine steer geometry has the ackerman angles backwards.
 

·
Registered
Joined
·
5,856 Posts
Discussion Starter #3,729
Just a little post to say I greatly miss the headland feature... it was my visual tool to know where to on/off sections and do the outer round(s) afterwards when spraying and seeding green manure crops and cereals. Which I will be doing a lot this summer and autumn.
I understand it is not so easy to re-implement because of the way the inner and outer boundaries are now handled in the code?

Do you think I should just add it back as a line? Skip the inner boundaries etc. After all the uturn line is really just a headland except closer to the outer boundary. it could work like the very outer boundary where the sections just turn off outside the outer boundary.
 

·
Registered
Joined
·
87 Posts
Do you think I should just add it back as a line? Skip the inner boundaries etc. After all the uturn line is really just a headland except closer to the outer boundary. it could work like the very outer boundary where the sections just turn off outside the outer boundary.
I for one would be very happy this way. If it works with the width setting screen with "x.x times implement width" like before, it is all I need :)
 

·
Registered
Joined
·
376 Posts
So I've been stumbling through AoG's code & the shared Arduino sketches trying to get a handle on what the different PGN bytes/bits all do. I've been editing the pgn-dev.xlsx from about a week ago. Is there another online document or any collaboration going on with creating an API or PGN document or something similar?
 

·
Registered
Joined
·
376 Posts
Brian, I'd like to install pressure sensors on my sprayer so that it can be operated manually but AoG will still map the coverage. Any pointers on how to modify AoG so that it doesn't assume coverage is the same as commanded sections but rather based on a reported byte from the Arduino?
I think I have things working except for in the future, I expect we would like to move towards solenoid valves and AutoSection control WITH pressure sensor feedback/mapping. Currently, both AutoSteer(7FFE) and Rate(7FFA) section control relay data to the Arduino is controlled by my pressure switches. Instead, I'd like to have the option of commanding the section relays according to AutoSection/Manual control but draw the coverage map according to the Rate Module(7FF9) relay/sectSW data from the Arduino. This way AoG gets confirmation that the booms at least have pressure instead of assuming the solenoid valves are working and also allows for automatic mapping with a simple manual control.

Any thoughts on how to go about that? So far there are no solenoid valves and I don't expect any will get installed any time soon.
 

·
Registered
Joined
·
5,856 Posts
Discussion Starter #3,734
Gotcha. Well, I'm finally doing a build on a combine in the next couple of weeks so I'll keep testing the devs on it. So that's flipped the antenna position around, correct? I'm thinking about placing the antenna more towards the engine block so it'll pick up changes quicker being closer to the steering. But I would have to put 'pivot axle point is behind the antenna', even though the drive/pivot point would be in front of the antenna?

Well, here ya go! It was a driving disaster lol.



 

·
Registered
Joined
·
20 Posts
That would be a really fun algorithm to write and figure out - to back up a trailer. Maybe fun isn't the right word :)
Interesting problem though. Need to deal with the pivot points and wheel placements.
The Ford trucks use the fish-eye backup camera and a couple of stick-on targets on the trailer to orientate the system.
 

·
Registered
Joined
·
4,005 Posts
For two ackerman steering trailers, I'm sure the math is doable. But reality would probably prove an order of magnitude more difficult (at least) because slop in the pivot points means there are actually many more directions of motion possible. Increases the complexity of the problem dramatically.

Backing up one trailer with a steerable axle (which is the equivalent of two trailers) is not that hard. Backing up three trailers or more? I doubt anyone, including Volvo could do it for very long.
 

·
Registered
Joined
·
9 Posts
Today's tests:

https://youtu.be/vaWTCXng5EU


Driving in the vineyards:

https://youtu.be/uM4W7o_stao


The software:

https://github.com/mtz8302/AG_GPS_ROOF_UNIT_ESP32

In AOG add to CNMEA line ~435:

//MTZ8302 june 2019
double.TryParse(words[13], NumberStyles.Float, CultureInfo.InvariantCulture, out headingTrue);
if (headingTrue > 0.00001)
{ headingHDT = headingTrue; }
else//keep old value
{ headingTrue = headingHDT; }



Greetings Matthias

Golly Matthias, that is some trust in the technology. Very cool (and scary) to watch!

This sounds like a silly question but could you post some closeup photos of your steering motor mount setup? It looks very good and smooth.
 
3721 - 3740 of 4139 Posts
Top