nRF52DK + 4 wire resistive touch interface

RE: nRF52DK + 4 wire resistive touch interface

MartinBL — Fri, 05 Jan 2018 13:00:28 GMT

I have never used a resistive touch screen before, so I don't really know what to expect.

Would it be possible to use a pot meter instead of the touch screen to measure that the ADC readings actually work? That way we could rule out the ADC and blame the touch screen (or some other external factor).
Is it somehow possible to measure the resistance and/or voltage across the resistive touch screen with a multimeter and compare the results with the ADC readings?

RE: nRF52DK + 4 wire resistive touch interface

sumit5ue — Fri, 05 Jan 2018 07:26:16 GMT

Also please note that we are connecting the touch panel directly to the GPIO - touch panel resistance is 300 Ohms between x+ and x- and 510 Ohms between y+ and y- on the panel. Not sure if that could be creating problems due to source current ...

RE: nRF52DK + 4 wire resistive touch interface

sumit5ue — Thu, 04 Jan 2018 18:50:58 GMT

Hi Martin - I am working with Maddy and our readings are as below. As you can see, if the touch was along x axis the readings were reasonable whereas if the touch was along y axis the readings were going up and down whereas the expected reading was for it to just increase.

Also, at the edges of the touch panel, we were expecting readings close to zero for both x axis and y axis but they were around 200 for x axis and 400 for y axis.

Testing along X axis X Y 220 400 225 403 243 398 288 410 325 398

Results as expected - small variations in Y readings are due to variations in finger touching the pad

Testing along Y axis X Y 220 425 215 435 223 398 217 470 210 436

Results NOT as expected - as Y readings are going up and down even though the touch was only in direction of increasing Y

RE: nRF52DK + 4 wire resistive touch interface

MartinBL — Thu, 04 Jan 2018 14:28:56 GMT

How much is it off? What you are seeing and what do you expect to see?

RE: nRF52DK + 4 wire resistive touch interface

maddy5075 — Thu, 04 Jan 2018 06:59:53 GMT

Martin, the Y position value is somehow not right. The pin handler disables the event to release the x plus from GPIOTE module and make it work like a plain GPIO. It's pulled high as digital input and if the user has touched the screen, it configures the x plus to digital out and drives high to calculate x position and then configures the x plus pin as analog input AIN1 to calculate the Y direction. When it does so the Y position value is not as we expect it to be. Could you think of any possible reason as to why the value is not right ? We tried increasing the settling delay for the pins, yet it did not make any difference.

RE: nRF52DK + 4 wire resistive touch interface

maddy5075 — Thu, 04 Jan 2018 06:59:02 GMT

main.c

We changed the pins to P0.03 (x plus), P0.04(y plus) to act as digital I/O and analog pins and P0.28(x minus) and P0.29(y minus) and this seems not to cause the loop forever. Also I have disabled the analog input's explicitly like we discussed. Thanks for the valuable suggestion Martin. We owe you one!

RE: nRF52DK + 4 wire resistive touch interface

MartinBL — Wed, 03 Jan 2018 13:25:23 GMT

Hi,

This sounds somewhat related to this issue that we are still investigating. But as you suggest, you should explicitly disable the input. Otherwise the SAADC will still be "attached" to the GPIO pad internally which may affect the logic levels. Can you try and see what happens?

Note that by default, P0.05 on the nRF52 DK is used for UART communication with the onboard Segger MCU. Unless you have cut solder bridge SB13 this could also interfere with your measurements.