Any Electronics Gurus Out There?

[npalen]

I've done a lot of business process automation. The thing I like to tell people is you have to be able to do it manually before you can automate it. So I would work on getting it manual before trying to automate it. Here's what I mean: try putting an electric control on the blade height adjustment with two buttons, one for up and one for down. Can you come up with a simple set of rules based on just the lights from the level that allows you to do the job? An example of simple rules might be, if the "high" light comes on, push the "down" button until the light goes off and if the "low" light comes on push the "up" button until the light goes off. A more complicated set of rules might be if the "high" light comes on, push the "down" button for one second and then check the "high" light.

Does that set of rules give you the result you want, with all terrain, soil conditions and speeds? Because if you can't come up with a set of rules that allows you to get good results with you acting as the computer, the actual computer is never going to be smarter than you. If it turns out that you need additional information -- speed, soil type -- you need either additional inputs (speed sensor) or a way for the user to input settings into the computer.

This approach also allows you to attack the problem in a divide-and-conquer way which allows you to work methodically. You'll never get something like this working if you try to do everything at once. I could see a process with the following steps:
1. Implement electric control of blade height with manual switches.
2. Devise rules for controlling blade height based on height sensor (and other sensors and inputs as necessary) using manual switches.
3. Devise way for reading height sensor from computer (and other sensors as necessary).
4. Devise way of activating blade control from computer.
5. Program computer to follow rules, read sensors and control blade.

I agree 100% with what you're saying here and been thinking about this while giving the valve lever a few quick "hits" to bring the light back to green. (on grade) I have a relatively large bore cylinder at 3 1/2" x 8" stroke so it takes a few good hits with about 3-4 GPM (GallonsPerMinute) with engine running not even half throttle.

Another control issue is what happens when the laser calls for "blade down" and the ground is so hard or the box is so full that it doesn't go down but continues to "ride high". From what I've seen on videos of operators running the automatic units is that they keep their finger on the manual/auto switch at all times ready to take charge of this situation. Part of the solution can be pre-tillage to make sure enough loose dirt is available.

I normally run my box blade on the 3PH but decided to put the tongue on this afternoon and did some grading on this practice field that is going to be planted to Buffalo in a couple months. Its a contoured terrain so not able to use the laser here. There are, however, laser levels with "high in the center" capability. I can see where GPS contouring would have its place but oh the dollars.

Its interesting to compare the pros and cons of running 3PH verses tongue pull. Maneuverability is better in some cases with the tongue as can turn shorter since I don't have castering gauge wheels on the box blade. Backing up into tight places is, of course, easier with the 3PH mode.

 

[MossRoad]

If I was going to try and level a large area, knowing what I know now, I'd never us a box blade on a 3pt hitch again, due to the way it moves up when the tractor nose goes down and digs in when the tractor nose goes up. I'd use something like you show with wheels on the back and tow it. That eliminates the problem with that. Then you just have to control the wheel height on the implement with a lever while you drive. You don't have to worry so much about the tractor following the oscillation in the terrain wagging the box blade up and down.

 

[npalen]

If I was going to try and level a large area, knowing what I know now, I'd never us a box blade on a 3pt hitch again, due to the way it moves up when the tractor nose goes down and digs in when the tractor nose goes up. I'd use something like you show with wheels on the back and tow it. That eliminates the problem with that. Then you just have to control the wheel height on the implement with a lever while you drive. You don't have to worry so much about the tractor following the oscillation in the terrain wagging the box blade up and down.

Are you running gauge wheels on the back of the box? They make all the difference in the world but you have to drop the 3PH all the way down so its in float mode. Then the tractor can go up and down but the box blade tries to maintain grade level because the wheels are on graded ground. With that said, I tend to agree that the long tongue does a little better job.

Depending on what type of smoothing you're trying to accomplish, the distance of the wheels behind the blade certainly comes into play. One extreme is right behind the blade, like in the picture above, and the other would be about twenty feet behind similar to the big Eversman land planes used to level fields for flood irrigation.

eversman land plane - Yahoo Image Search Results

Guess we're going to have to design a box blade with gauge wheels adjustable for distance behind the blade.

 

[npalen]

Are you running gauge wheels on the back of the box? They make all the difference in the world but you have to drop the 3PH all the way down so its in float mode. Then the tractor can go up and down but the box blade tries to maintain grade level because the wheels are on graded ground. With that said, I tend to agree that the long tongue does a little better job.

Depending on what type of smoothing you're trying to accomplish, the distance of the wheels behind the blade certainly comes into play. One extreme is right behind the blade, like in the picture above, and the other would be about twenty feet behind similar to the big Eversman land planes used to level fields for flood irrigation.

eversman land plane - Yahoo Image Search Results

Guess we're going to have to design a box blade with gauge wheels adjustable for distance behind the blade.

Silly question about whether you are running gauge wheels after rereading your post.

Also, my apologies to all for inadvertently taking this electronics thread off course. Hopefully can get back on track with additional thoughts on automating the laser control.

 

[kenmbz]

Picking up an LED during daylight is full of issues. Picking up current going to an LED makes more sense.
I do like the process idea, think through each "scenario" this would go through when you do it manually. It's important to make sure you have stops to make sure you don;t go to far up/down (limit) . People react to a lot of inputs that machines don't understand.
And a kill switch if the system enters an unknown state.

 

[quicksandfarmer]

I've done a lot of business process automation. The thing I like to tell people is you have to be able to do it manually before you can automate it. So I would work on getting it manual before trying to automate it. Here's what I mean: try putting an electric control on the blade height adjustment with two buttons, one for up and one for down. Can you come up with a simple set of rules based on just the lights from the level that allows you to do the job? An example of simple rules might be, if the "high" light comes on, push the "down" button until the light goes off and if the "low" light comes on push the "up" button until the light goes off. A more complicated set of rules might be if the "high" light comes on, push the "down" button for one second and then check the "high" light.

Does that set of rules give you the result you want, with all terrain, soil conditions and speeds? Because if you can't come up with a set of rules that allows you to get good results with you acting as the computer, the actual computer is never going to be smarter than you. If it turns out that you need additional information -- speed, soil type -- you need either additional inputs (speed sensor) or a way for the user to input settings into the computer.

This approach also allows you to attack the problem in a divide-and-conquer way which allows you to work methodically. You'll never get something like this working if you try to do everything at once. I could see a process with the following steps:
1. Implement electric control of blade height with manual switches.
2. Devise rules for controlling blade height based on height sensor (and other sensors and inputs as necessary) using manual switches.
3. Devise way for reading height sensor from computer (and other sensors as necessary).
4. Devise way of activating blade control from computer.
5. Program computer to follow rules, read sensors and control blade.

Let me add that while all of the above steps are necessary, #2 is the only one that is hard. All of the other ones are things that are commonly done.

 

[npalen]

Found this info on detecting a red laser that might/might not be of interest:

Detecting red laser with an IR receiver/photodiode?
Arduino Laser Tripwire - YouTube

This is a little different that picking up a beam from a rotating laser at, say, 1500 feet distance but some of the same principles probably apply.

 

[MossRoad]

Amazon.com: Johnson Level and Tool 4-6792 Remote Display: Home Improvement

This remote is available to plug into the johnson receiver shown in the first post of this thread. It can be plugged into the machine mounted receiver and placed near the operator for easier viewing.

Well, between the receiver and this remote, now you've got $1000 wrapped up in it before you start. Do you have a rotating laser unit yet?

You can rent a laser grader box for a skid steer for $250 for a day.
LASER GRADER BOX Rentals Evansville IN, Where to Rent LASER GRADER BOX in Evansville Indiana, Newburgh, Chandler, Darmstadt, Boonville IN, Henderson KY

 

[MossRoad]

Here's a grader blade for $200 a day.

Rent this 218 Bobcat 84 in. Grader in Dekalb, IL

 

[MossRoad]

However, if your desire is to experiment, well, then, have fun.

I'd want to experiment without destroying the sensor unit.

So, I'd try to find a photodiode or phototransistor that can detect light from the LED's on that sensor first. Once I found that, I'd make a board to mount the photodiodes onto so that I could place that board directly on the receiver unit's face, where they'd face the LEDs. Then I'd fasten it with tape to seal out any light from the sides, but not block any of the receiving sensors on the unit, so that it'll still work 360 with the laser.

The unit looks like it has 5 LED banks:

1 NEUTRAL
1 HIGH
1 MORE HIGH
1 LOW
1 MORE LOW

I'd make a circuit with 5 photo diodes to correspond with those.

Neutral is just that. Neutral.
Neutral = no hydraulic movement.

HIGH = lower hydraulics
MORE HIGH = lower hydraulics faster
LOW = raise hydraulics
MORE LOW = raise hydraulics faster

Then you'd have to work out logic in a PLC to say something like:

If I1 = 1 then O1 = 5V
If I1 = 1 and I2 = 1 then O1 = 10V

Where I = inputs and O = outputs
And assuming changing voltage = more valve movement....

Stuff like that.

However, if it's strictly an off and on valve and is not variable in speed with voltage, you're going to have to vary by time instead of speed, so you'd do something like:

If I1 = 1 then O1 = on for 1 second
If I1 = 1 and I2 = 1 then O1 = on for 2 seconds

Stuff like that.

Then you have to account for varying soil conditions, speed of tractor, etc... but it could be doable and a fun project.

I haven't done this kind of stuff for 35 years, but will enjoy following your project. Good luck. :thumbsup:

 

[npalen]

Well, between the receiver and this remote, now you've got $1000 wrapped up in it before you start. Do you have a rotating laser unit yet?

You can rent a laser grader box for a skid steer for $250 for a day.
LASER GRADER BOX Rentals Evansville IN, Where to Rent LASER GRADER BOX in Evansville Indiana, Newburgh, Chandler, Darmstadt, Boonville IN, Henderson KY

I envy you guys who can go out and rent this kind of stuff. We're 60 miles from the nearest rental that would have anything even close to this and I seriously doubt they have any laser graders.

 

[npalen]

However, if your desire is to experiment, well, then, have fun.

I'd want to experiment without destroying the sensor unit.

So, I'd try to find a photodiode or phototransistor that can detect light from the LED's on that sensor first. Once I found that, I'd make a board to mount the photodiodes onto so that I could place that board directly on the receiver unit's face, where they'd face the LEDs. Then I'd fasten it with tape to seal out any light from the sides, but not block any of the receiving sensors on the unit, so that it'll still work 360 with the laser.

The unit looks like it has 5 LED banks:

1 NEUTRAL
1 HIGH
1 MORE HIGH
1 LOW
1 MORE LOW

I'd make a circuit with 5 photo diodes to correspond with those.

Neutral is just that. Neutral.
Neutral = no hydraulic movement.

HIGH = lower hydraulics
MORE HIGH = lower hydraulics faster
LOW = raise hydraulics
MORE LOW = raise hydraulics faster

Then you'd have to work out logic in a PLC to say something like:

If I1 = 1 then O1 = 5V
If I1 = 1 and I2 = 1 then O1 = 10V

Where I = inputs and O = outputs
And assuming changing voltage = more valve movement....

Stuff like that.

However, if it's strictly an off and on valve and is not variable in speed with voltage, you're going to have to vary by time instead of speed, so you'd do something like:

If I1 = 1 then O1 = on for 1 second
If I1 = 1 and I2 = 1 then O1 = on for 2 seconds

Stuff like that.

Then you have to account for varying soil conditions, speed of tractor, etc... but it could be doable and a fun project.

I haven't done this kind of stuff for 35 years, but will enjoy following your project. Good luck. :thumbsup:

Yes, at my stage in life its all about having fun!

Might be able to leave about half of the input "screen" on the receiver open and blank out the other half (side shield in between maybe) for the photo diodes. Would be nice to retain a visible reference on the receiver.

Or maybe leave the entire screen open and use some fiber optics to fit over just one LED of the output at I1 and I2. Heck even a little black tube between the diode and the LED lens might work.

Thanks again for the input. Keeps me thinking.

 

[MossRoad]

I envy you guys who can go out and rent this kind of stuff. We're 60 miles from the nearest rental that would have anything even close to this and I seriously doubt they have any laser graders.

Bummer.

 

[MossRoad]

Yes, at my stage in life its all about having fun!

Might be able to leave about half of the input "screen" on the receiver open and blank out the other half (side shield in between maybe) for the photo diodes. Would be nice to retain a visible reference on the receiver.

Or maybe leave the entire screen open and use some fiber optics to fit over just one LED of the output at I1 and I2. Heck even a little black tube between the diode and the LED lens might work.

Thanks again for the input. Keeps me thinking.

Yes, it would be nice to be able to still see the unit for manual operation.

 

[Industrial Toys]

Amazing how little used and broken laser grading Junk is on E-Bay. Discouraging. Just as with my Rotating Level, it was a better bet to buy new from Tiger Supplies in NJ. They have a Topcon 270 degree stand alone receiver for $500.00. That is just about the cheapest thing I can find, new or used for a machine receiver. You can get a cool wireless remote, that costs substantially more than the receiver.

If I had a use for it and not four hundred other unfinished projects, I am confident that for the $500.00, a bit of other electronic stuff and some hydraulic fittings I could make this work. with my luck, though, the receiver would probably be totally potted! That actually would not surprise me, given it's meant to be stuck to an excavaor stick.

 

[wvgca]

Amazing how little used and broken laser grading Junk is on E-Bay. Discouraging. I am confident that for the $500.00, a bit of other electronic stuff and some hydraulic fittings I could make this work. with my luck, though, the receiver would probably be totally potted! That actually would not surprise me, given it's meant to be stuck to an excavaor stick.

The receiver [machine mount] is usually set to a height of five or six feet above ground when the attachment is at depth. They usually don't suffer too much, except for some vibration. I have a Spectra [non slope] laser with low end receiver, good enough for 500 yards anyways ..

 

[npalen]

Can you post a link?

I was just on the auction site before posting ... look it up ..
This works very well with a box blade or level sweep

Still waiting for an ebay link.

 

[wvgca]

Still waiting for an ebay link.

I decided not to post one .. if you want, go look ..

 

[npalen]

Can you post a link?

I decided not to post one .. if you want, go look ..

You decided not to post one because there are none.

 

[npalen]

A couple new developments just to make things more complicated and less complicated all at the same time:

I don't believe that I've mentioned that the LEDS flash and they also flash at different speeds depending on the distance from grade. They also change speeds when going thru a couple of zones and I think one of them is when going past the 2" departure from grade. I need to do some testing to see exactly where the speed changes occur. The owner's manual says nothing about all this.

Something that just occurred to me this evening is that we have pinouts on the receiver that go to the remote when attached via the cable. There are seven of them and one appears to be ground and one for the 12 volt charger and/or the 12 volt battery input. (The receiver has an onboard rechargeable battery or can also be powered by the tractor battery.) That leaves five pins for the five sets of LED's. Yes I feel very stupid for not realizing the connections for automation are right there in plain sight.

It gets interesting when one thinks about utilizing the different flashing speeds to change on/off times for the solenoid valve.

Edit: Added pic.