5 Wire Touch Screen Schematic

Thread Starter

Peter David Grant

Joined Apr 4, 2019
2
Hello,

First post here, but hopefully I've managed to cover everything!

Some quick background: I'm very much a dabbler in electronics. Trained as a mechanical engineer, and now working as a software programmer (LabVIEW). We're building a demonstrator project (a Stewart Platform), which will start off with 2DOF balance ball on the top. I'm using a National Instruments cRIO for the I/O.

I came across All About Circuits while searching for a 5 wire touch screen schematic. I did however managed to find a schematic for a 4 wire via this thread: https://forum.allaboutcircuits.com/threads/touch-screen-controller.13424/

I'm looking to make my own controller for the 5-wire touch screen for reading the ball position, as I want to see how quick we can read position feedback through the FPGA (i.e. I don't want to use a controller outputting position via i2c or some other communication method).

My initial thought was to use 8x n-channel Logic Level MOSFETs (I have spare TTL channels for control) to produce a voltage across the touch screen, and then read voltage back through an AI channel. Each corner would be attached to a MOSFET pair as per the attached schematic.

I'd drive the MOSFET gates with a delay between any change, to ensure we don't get any overlap (programmatically though the FPGA), but thought this would mean at some points in time it would be floating (?), so added a pull down resistor, to allow it to be connected to ground.

So to my actual questions:
  1. Am I totally barking up the wrong tree, or is the basic schematic attached how I should go about doing this?
  2. Should I just switch to a 4 wire, and use the schematic above? I picked a 5 wire, after reading they are typically more robust.
  3. Is there an alternative schematic I should be using instead?
Kind regards,
Peter
 

Attachments

John P

Joined Oct 14, 2008
2,025
Years ago I had to use a 4-wire touch screen for a microcontroller project, and I designed my own reading system. It wasn't very difficult.

I found this explanation of using a 5-wire touch screen:
https://www.dmccoltd.com/english/museum/touchscreens/technologies/5-wire.asp

It seems to me that you don't really need any external components, though it depends on what the resistance of the screen is versus how much current the pins of your FPGA can supply. If a booster is necessary, a CMOS logic chip should be able to do it. You'd have 2 diagonally opposite corners of the screen connected to +V and Gnd: the other 2 corners would also be +V and Gnd, but you'd need to be able to switch them from one to the other. Measure the voltage on the single output under both conditions, and you'd have a position reading.
 

Thread Starter

Peter David Grant

Joined Apr 4, 2019
2
Hi John,

Thank you kindly for the reply. Ahh thanks for the link, I've come across that link before, and it helped my understanding.

It seems to me that you don't really need any external components, though it depends on what the resistance of the screen is versus how much current the pins of your FPGA can supply. If a booster is necessary, a CMOS logic chip should be able to do it. You'd have 2 diagonally opposite corners of the screen connected to +V and Gnd: the other 2 corners would also be +V and Gnd, but you'd need to be able to switch them from one to the other. Measure the voltage on the single output under both conditions, and you'd have a position reading.
Ahh yes, I hadn't thought about connecting two diagonally opposites permanently to +V and Gnd, that makes a lot of sense.

I hadn't considered using a CMOS Invertor, but having researched it, it seems like a very good choice for the problem at hand. I believe that looking at the following data sheet, it would work, with the gate coming from my 5V ttl source, where Vcc is 5V?

https://docs-emea.rs-online.com/webdocs/0826/0900766b808265ab.pdf

Thanks again,
Peter
 

John P

Joined Oct 14, 2008
2,025
Yes, a simple 74HC14 would be as good as any other logic chip to drive the panel; you just need to know how much current will flow. Most likely the resistance is well within the range that the output can drive without the voltage dropping significantly. It would be best though, to use a power supply that also powers the A/D converter, assuming that the converter's range is the power supply. If that's the way it's set up, then any inaccuracy in the supply voltage will be canceled out in the A/D.
 
Top