TFT display problem

#1

I have a no name (literally!) chinese 2.8 TFT touch screen shield that I’m having trouble with.

The display works with as expected using the adafruit TFT script, but only when I connect it using jumper wires. If I plug the shield into the Uno, I just get the back light.

Problem appears to be the reset line for the screen - connecting to the reset pin works, but connecting to A4 doesn’t (even though its defined in the sketch).

I havent been able to get anything out of the touch screen sketches (again back light only), but I might end up just using it as a display anyway so that may not matter

Be interested to hear if anyone has any thoughts on this though

many thanks

S

#2

@Target, sounds interesting.

Got a schematic of the touchscreen shield? Looking at the Adafruit variants, there’s a reset generator chip that will send a reset to the TFT controller chip and the Uno.

That it works one way and not another is handy. I suggest mapping the differences between the two wirings; with jumpers or plugged in, which you have probably done already, and then look at the timing of the reset pin relative to the rise of the power supply pins.

Have you captured anything from the graphicstest program using the console? If not, I’d also use an oscilloscope or other instrument to capture the SPI transactions between the Uno and the controller.

#3

no supporting materials unfortunately (seems to be a common thread - I’ve looked at dozens of posts about these displays) but the attached images should give you an idea of what I’m working with

WP_20161129_06_20_55_Pro.jpg769×1027 340 KB

FWIW when I run the graphics test it reports the driver as an ILI9341, but all the other sketches fail to recognise it at all

I’m also curious about the touch screen - it appears to have a touchscreen overlay, but there doesn’t appear to be any pinouts for it??

#4

and here’s the front

WP_20161207_21_19_55_Pro_(2)[1].jpg2160×3840 2.11 MB

#5

Okay, it seems to be a parallel bus display, not SPI or I2C; note the number of LCD data pins, and the read (RD) and write (WR) pins. So discount any library you find that talks about SPI, MOSI, MISO, and so on. (Your board does have SPI for the microSD card slot though, but the signals don’t seem to be shared with the LCD controller).

Good chance the touchscreen overlay is hooked into the LCD controller.

Got any legends on the chips U2 and U3? Can the screen be disconnected from the board to look at chips under it?

I’m not sure I can trust the graphicstest now; perhaps that string is generated by the test without communication happening.

#6

Those chips are labled thus

HC245
HZ11117
TXD540E

I assumed the control lines are shared but clearly there’s nothing meaningful (to me) to go on

I could probably peel the display away from the board but I’m a bit wary about doing it cos I have no idea what’s behind it

That literal is dependent on the script reading one of the registers so I’m going to assume its correct

#7

Thanks. The 74HC245 chips are octal bus transceivers being used as level shifters, and I’ve found a few schematics that look close to what your photographs show. The controller chip itself is on the back of the display, between the display and the PCB.

Without getting the board into my lab, my best guess is you have been inflicted with process variance on the reset timing, which two other people have reported against the library;

Try bringing something like that into your code.

The change of wiring can easily change the timing; because the wires add capacitance.

#8

sorry, real life intrudes for the moment :frowning:

Thanks for the tip though, I shall have a game with this on the weekend and report back

#9

wonderful!! added a delay and it works as expected !!

many thanks - but now for the touchscreen…

#10

Oh, that’s great. Just an educated guess. :grin: