Maker Pro
Maker Pro

Building a manually-controlled touch screen tapper for controlling farm tractor GPS

jac09

Apr 3, 2022
10
Joined
Apr 3, 2022
Messages
10
Hello, first post here. I apologize if this isn't in the correct forum and for what will be a long post. Also, the photos of products are for reference only, that's why I included photos instead of product links. Didn't want anyone to think I'm trying to spam you. I'm basically trying to create one of these, but one that I can manually control instead of having it tap non-stop:
upload_2022-4-2_21-39-49.png


Aside from engineering, I also farm. Our tractors can be driven via GPS though our Agleader Incommand 1200 displays. This is the product link if you're curious to know what we're using: https://www.agleader.com/farm-management/displays/incommand/?locale=en. They do this by following a set guidance line and then automatically follow that line through the field. There are some touch-screen buttons on the display that allow the driver to manually nudge the tractor left or right from that guidance line if needed. See pic.
upload_2022-4-2_21-44-37.png



My problem is that the most convenient place to mount the display in the tractor is frequently not very convenient to be reaching over and manually touching the buttons on the screen. Therefore, I want to make a simple wired remote control that will simulate touch the screen whenever I push a physical button.

From what I've read, these displays are capacitive touch screens. I experimented today with taping a piece of aluminum foil to the screen and then running wire leads from the foil to a switch, and then from the switch to a spool of heavier gauge wire. (no electricity actively involved in the circuit) I discovered that so long as the initial lead from the foil to the switch was short and used a light-gauge wire, then the screen did not sense the foil and the initial lead. However, once I closed the switch, it would sense touch due to the longer, heavier gauge wire on the backside of the switch.

This was great, except that the initial lead had to be so short such that I really wasn't gaining anything to use the switch instead of the actual touch screen button. So, I experimented with placing a relay that we had laying around into the circuit. This relay is a fairly heavy-duty 12vdc that I've seen called a starting relay. Theoretically, using the relay would allow me to place my switch farther away from the screen while still keeping the initial wire lead from the foil to the relay short.

When I did this, the screen did not sense the foil when I had it hooked up to the relay with the switch off. I also tested and there was no continuity between the relay terminals that I was using to connect the initial lead to the heavy wire spool. Perfect. However, when I close the switch, the relay clicked, but the screen did not sense the foil. I probed the two relay terminals, and there was now continuity between them. I also tried ground the wire spool to the tractor and also to the monitor itself via a usb cable that I stripped down and plugged into a USB port on the side of the monitor. I connected the wire spool to the ground lead in the usb cable. Neither of these things helped. For some reason, adding the relay into the circuit is preventing the screen from sensing the foil, and I cannot figure out why.

I'm a mechanical engineer. I typically leave stuff like this to you sparkies because I usually end up letting the smoke out of everything. I got a nice whiff of burning wires today from a cab fan I supposedly fixed. So, if anyone has any suggestions on how to improve my circuit, or even how to possibly hijack the tapping probes used in the product in the amazon link above, I'd be very interested in hearing your ideas. Of, if anyone knows what the chip is in the product below and sourcing and using a similar chip isn't difficult, I'd be willing to look into that as well. thanks in advance.
upload_2022-4-2_21-41-59.png
 

Bluejets

Oct 5, 2014
5,985
Joined
Oct 5, 2014
Messages
5,985
I'd be getting in touch with the manufacturers as they may have some ability to serial link either through hardwired or wireless.
 

kellys_eye

Jun 25, 2010
5,095
Joined
Jun 25, 2010
Messages
5,095
If the device has a USB port anywhere try putting a mouse into it - this usually brings up a screen cursor. If this is the case then you can also get a trackerball that could be mounted somewhere convenient that would work even better.
 

jac09

Apr 3, 2022
10
Joined
Apr 3, 2022
Messages
10
I'd be getting in touch with the manufacturers as they may have some ability to serial link either through hardwired or wireless.
I've tried. They have no interest in developing such functionality. Most farmers don't use the nudge function as much as we do. We farm organically, and our mechanical weed control needs to be controlled more tightly than what the GPS can manage sometimes. Most farmers don't need to be as precise, so they don't want to waste resources on a niche market.
 

jac09

Apr 3, 2022
10
Joined
Apr 3, 2022
Messages
10
If the device has a USB port anywhere try putting a mouse into it - this usually brings up a screen cursor. If this is the case then you can also get a trackerball that could be mounted somewhere convenient that would work even better.
I can try, but I'd be shocked if it worked. Plus, using a mouse in a bumpy, dusty tractor cab isn't going to be the most tactile tool for what I need to do. It'd be interesting to see what happened though. Just need to find a wired mouse now.
 

kellys_eye

Jun 25, 2010
5,095
Joined
Jun 25, 2010
Messages
5,095
Plus, using a mouse in a bumpy, dusty tractor cab isn't going to be the most tactile tool for what I need to do.
Which is why I suggested you use a trackerball........

Trackerballs are used in all sorts of applications where a mouse cannot 'roam' - even the military use them.

Amazon has a gazillion examples....
 

jac09

Apr 3, 2022
10
Joined
Apr 3, 2022
Messages
10
Which is why I suggested you use a trackerball........

Trackerballs are used in all sorts of applications where a mouse cannot 'roam' - even the military use them.

Amazon has a gazillion examples....

I appreciate the suggestion, but I have used them before. They would not work well for what we need it to do, even if a cursor does appear. I don't want to be watching the screen to make sure we're clicking the correct button; instead, we are watching camera monitors on a different screen to make sure our tools don't get too close to the crop. Moving back and forth between monitors plus moving the mouse around will lead to errors. We need something that is fixed on the screen. If the mouse input could somehow be set to click the same button every time then we might have a solution, but I highly doubt that this monitor supports macros.
 

kellys_eye

Jun 25, 2010
5,095
Joined
Jun 25, 2010
Messages
5,095
A pair of simple solenoids with touchpen pads (the soft rubber stuff you get on pen pointers for screens) on lever arms would work.

Two pushbuttons each generating a short pulse (556 dual monostable circuit) to 'kick' the lever arm onto the screen might simulate the necessary touch effect.

The difficult part is fabricating the lever arms but a length of copper-clad pcb material (glass fibre) will accept a soldered-on tube to accept an axle for pivoting? Some experimentation with the position of the pivot point to get the necessary movement from a small deflection offered by the solenoid. I can picture this in my mind but can't get it on paper for you.... sorry!
 

jac09

Apr 3, 2022
10
Joined
Apr 3, 2022
Messages
10
A pair of simple solenoids with touchpen pads (the soft rubber stuff you get on pen pointers for screens) on lever arms would work.

Two pushbuttons each generating a short pulse (556 dual monostable circuit) to 'kick' the lever arm onto the screen might simulate the necessary touch effect.

The difficult part is fabricating the lever arms but a length of copper-clad pcb material (glass fibre) will accept a soldered-on tube to accept an axle for pivoting? Some experimentation with the position of the pivot point to get the necessary movement from a small deflection offered by the solenoid. I can picture this in my mind but can't get it on paper for you.... sorry!

This was my first thought, I actually have something modelled up in CAD and ready to be printed. I was a bit concerned about the noise of the solenoids clicking repeatedly throughout the day and also about having something punch my monitor repeatedly. If I didn't get the distance set right or it happened to move while in use I could damage the screen. Then I came across these electronic, actuator-free tappers shown in my OP and was hoping I could cobble together a similar system.

As of now, I've ordered a set of these because I noticed that they used a simple 3.5mm headphone jack: https://www.amazon.com/gp/product/B09H1DSVXH/ref=ppx_yo_dt_b_asin_title_o01_s00?ie=UTF8&th=1. I could be completely wrong, but my thought is that if they're using a headphone jack instead of a usb port, the signal going to the tapping nodes is liking a simple on/off analog signal and therefore easier to hack than the usb-based ones in my OP. (I realize the usb could also be used with analog, and I realize the headphone jack could be used with digital). It is a 3-banded headphone jack, so I'm assuming they are using the 3 bands for ground, constant-power, and signal power.

I also ordered a variable voltage regulator. My hope is that I'll be able to start low on the voltage and dial the voltage up until I start activating the tappers. If I can reliably activate the tappers, then I'll model up and 3d print a mounting arm to hold several of them in the correct spot. I'm still curious about the mouse so I'll try it just to see what happens.
 

kellys_eye

Jun 25, 2010
5,095
Joined
Jun 25, 2010
Messages
5,095
It could well operate on the piezo principle, using sound waves to create an acoustic 'tap'. There are videos of levitation devices using piezo sounders so the 'physical' aspect is certainly applicable.
 

jac09

Apr 3, 2022
10
Joined
Apr 3, 2022
Messages
10
Well the clicker came in the mail today and it works to active the touch screen when using the supplied auto-clicking module. Using my multi-meter, it appears it is activated using a 1v signal, but I'll play with it a bit once I get my voltage regulator up and running. I haven't had a chance to do that yet, so don't know if I'll be able to turn it on/off manually yet, but I was excited that the tapper worked. Its tough to get a good voltage reading with the multimeter because the signal is constantly clicking on and off.
 

jac09

Apr 3, 2022
10
Joined
Apr 3, 2022
Messages
10
It works! 2v into a 3.5mm headphone jack adapter with a momentary on switch is enough to tap the screen as fast or slow as I want. Now I just need to design and 3d printer the holder for the actuators and also a holder for the switches. I'm stoked, this will be amazing for us and relatively cheap.
 

jac09

Apr 3, 2022
10
Joined
Apr 3, 2022
Messages
10
I do have another question. The gadget I used to find the required voltage was a variable buck converter. It did want I want, but I want to find a fixed 12v to 2v DC converter. Are the buck converters I see online suitable for mobile electronics like this? I don't know what a "buck" converter actually does so want to make sure I'm using the correct tool. If there's a different cheap converter I should use please point me that direction. There is less than 3a draw.
 

crutschow

May 7, 2021
511
Joined
May 7, 2021
Messages
511
A buck regulator is a switching type regulator to get high conversion efficiency (>80%).
A 12V to 2V converter would only be about 15% efficient (the rest generating heat).

On what basis do you think it needs near 3A?
Did you measure the current?

Since I assume this current is only for a short pulse, a simpler, cheaper, linear regulator may be appropriate, even though they would be much less efficient, since the average power would be low.
A very common linear regulator for 1.5A is an LM317.
 

jac09

Apr 3, 2022
10
Joined
Apr 3, 2022
Messages
10
A buck regulator is a switching type regulator to get high conversion efficiency (>80%).
A 12V to 2V converter would only be about 15% efficient (the rest generating heat).

On what basis do you think it needs near 3A?
Did you measure the current?

Since I assume this current is only for a short pulse, a simpler, cheaper, linear regulator may be appropriate, even though they would be much less efficient, since the average power would be low.
A very common linear regulator for 1.5A is an LM317.

I have not measured current. Don't have enough hands. The tapper module that I stole the actuators from is normally USB-powered. If a USB typically only provides up to 0.5A @ 5v, then I think we can safely assume that actual current at 2V is less than 1.25A? The 3A listed above was just a guess based off the limitations set on the buck converter I was using. Plus, like you said, it is only for brief moments, and there will be periods of rest in between.

Are you talking about something like this: https://www.mouser.com/ProductDetail/STMicroelectronics/LM317P?qs=swDD%2BF%2Bps7e6xZSe3iJsFg==. Do they get hot enough in brief useage that I'd need a heat sync?

In your data sheet link, they show a circuit with multiple resistors for charging a battery. Am I understanding correctly that I'd need to size some external resistors in order to get my desired 2v output voltage? What exactly is the LM317 doing in the circuit? I tried working through the examples to figure out the resistors I'd need but that seems to be beyond my skills. Even though buck converters are literally 10x more expensive than linear, I still might go that route anyways if they are fully self-contained and otherwise suitable for the project. Looks like I can find them for $4-$16, which is tolerable.

upload_2022-4-6_12-45-57.png
 

crutschow

May 7, 2021
511
Joined
May 7, 2021
Messages
511
The LM317 uses feedback from the output to maintain a constant output voltage, independent of the input voltage and output current.
It adjusts the output voltage to maintain 1.25V across R2 (between the OUT and REF terminals).
For the circuit below, this generates a current of 1.25V /121Ω = 10.33mA through R2.
The ADJ input takes no significant current, so this current must also flow through R1, generating a voltage drop of 10.33mA * 71.5Ω = 0.738V.
The nominal output voltage is then 1.25V + 0.738v = 1.989V.

Below is the simulation of the circuit.
(Rload is just generating a varying load for the simulation).
The output (green trace) stays at 2V, until the output current (red trace) reaches the LM317 internal current limit of about 1.4A.

The LM317 from Mouser you referenced will work for this.
It should need no heat-sink for your intermittent usage.

The resistors are any 1/8 or 1/4 watt leaded resistors.
The 0.1µF cap is a ceramic type, and the 1.0µF can be either an electrolytic or ceramic, both leaded.

upload_2022-4-6_12-39-58.png
 
Last edited:

jac09

Apr 3, 2022
10
Joined
Apr 3, 2022
Messages
10
The LM317 uses feedback from the output to maintain a constant output voltage, independent of the input voltage and output current.
It adjusts the output voltage to maintain 1.25V across R2 (between the OUT and REF terminals).
For the circuit below, this generates a current of 1.25V /121Ω = 10.33mA through R2.
The ADJ input takes no significant current, so this current must also flow through R1, generating a voltage drop of 10.33mA * 71.5Ω = 0.738V.
The nominal output voltage is then 1.25V + 0.738v = 1.989V.

Below is the simulation of the circuit.
(Rload is just generating a varying load for the simulation).
The output (green trace) stays at 2V, until the output current (red trace) reaches the LM317 internal current limit of about 1.4A.

The LM317 from Mouser you referenced will work for this.
It should need no heat-sink for your intermittent usage.

The resistors are any 1/8 or 1/4 watt leaded resistors.
The 0.1µF cap is a ceramic type, and the 1.0µF can be either an electrolytic or ceramic, both leaded.

View attachment 54804

Thank you for working that out for me. I'll have to dig into it some more.
 
Top