Coffee Space


Listen:

Linux PDA Part 2

Preview Image

Preview Image

This is a follow-up article from the previous article on the Linux PDA project. In this article I will discuss the planning process for the device.

PineCube

Previously we asked two questions regarding the possibility of using the PineCube hardware as a base for a PDA device:

  1. The current state of the PineCube project overall and ensuring they are still in stock.

I've had some discussion on IRC and it appears some people are still active there. There appears to be some active and ongoing work to build a motion detection activated recording for the device. As long as the hardware remains in stock, it should be possible to use this as a base device. There appears to even be updated Linux kernels for the device.

  1. The current state of the LCD driver. Writing an LCD driver for the Linux kernel from scratch is not something I've done before, so it will be a lot of pain if I end up having to do that.

Apparently the display that is compatible is a RB043H40T03A-IPS, that can be found here for about $10 USD each (without shipping and GST). Apparently it is designed for use with a 16 bit STM32 microcontroller.

You can also download the video here.

As you can see in this video shared publicly via IRC, Gaimee has been able to get the LCD working. Apparently:

<@gamiee> B[] : it's not touch. The source code from this preview is in PineCube BSP

Some more investigation is required, after all Pine64 wiki says:

LCD touch screen panel information:

It must support touch right?

Touch Screen

Firstly I found out that for a similar touch display, they use pins 37, 38, 39 and 40 for the touch display:

LCD pins for touch

LCD pins for touch

So far so good. I'm not entirely sure how this electrode setup works, so I check further in the same document:

LCD electrodes

LCD electrodes

As you can see, we appear to be measuring the changing 'resistance' between two points on the display to measure each axis. This is confirmed when we look further:

LCD linear response

LCD linear response

Awesome! So now we just need to look at the 40 pin connector on the main board...

Main Board

Looking at the main board schematics, we can see our 40 pin header named J10:

Mainboard LCD connector

Mainboard LCD connector

Weird, our pins are called NC, i.e. "No Connection" - but they are clearly connected to something. We see that pin 37 is pulled up to 3.3V, a good start. We then see pin 40 pulled down to ground, not so bad either.

Where things get weird is that pins 38 and 39 are attached to JTAG-DI and JTAG-DO. Maybe they can read analogue data and all will be good?

S3 ARM CPU

Reading the CPU datasheet, on page 31 we see that PB12 and PB13, the JTAG pins, are digital I/O only:

JTAG pin capability

JTAG pin capability

There is just one last hope... If PB12 and PB13 are exposed externally, we could look to jump them over to an ADC pin on the CPU!

PineCube GPIO pinout

PineCube GPIO pinout

Nope.

PineCube Conclusion

This hardware was already a challenge to work with. I was already not happy with the form factor and the PCB stacking, but I was willing to overlook it.

The real problem here is that the hardware would need a re-work in order to get touchscreen capability working. Accessing PB12 and PB13 on a ball-mount CPU is going to be impossible, and I really don't want to solder the sort of pitch that 40 pin connector will be.

This hardware will simply be too difficult to work with. We can likely find something much more appropriate for the task.

Alternatives

So, it looks like we need a new base-board. There are a few interesting options out there:

I'm literally not happy with any of the alternatives on offer. Perhaps the lack of base-device hardware is what ultimately kills this pipe dream?