TheBackShed.com

Forum Index : Microcontroller and PC projects : Forget the ILI9488, The ST7796S is the new kid in town

Page 2 of 2

Author

Message

disco4now

Guru

Joined: 18/12/2014
Location: Australia
Posts: 983

Posted: 05:26am 19 May 2025

Copy link to clipboard

Print this post

The printout results should match below.
The print hex$(pixel(1,1)) are giving incorrect results, one of the symptoms of it not working 100% even if screen output looks OK.

Do you see D1 on your board?

st7796S Test
F80000 RED F80000
F80000
FC00 GREEN FC00
FC00
------------
F8 BLUE F8
F8
F8
F8 BLUE F8
F8
F8
F8 BLUE F8
F8
F8
----------------
FC00
FC00
FC00
FC00
F8
WRITE TIME: 1.268
F80000
F80000
BLIT TIME: 12.134
CLS TIME: 0.518

Latest F4 Latest H7 FotS

phil99

Guru

Joined: 11/02/2018
Location: Australia
Posts: 2449

Posted: 05:44am 19 May 2025

Copy link to clipboard

Print this post

Bypassed the diode D1 in the CS line but no change. BLIT and reading a pixel cause the display to freeze, accepts no new data. Interpreter still running but requires 'cpu restart' to get the display to respond.

Now here is an odd thing, I think the display memory still takes data but it doesn't always go to the screen.

> cpu restart
> cls 255 'blue
> ? pixel(9,9)
248
> ? pixel(9,9)
248
> ? pixel(19,19)
248
> cls 255<<8 'this normally is green but remains blue
> ? pixel(19,19)
1630208 'returns value for green but display is blue
>

Edited 2025-05-19 16:03 by phil99

disco4now

Guru

Joined: 18/12/2014
Location: Australia
Posts: 983

Posted: 06:14am 19 May 2025

Copy link to clipboard

Print this post

Mine works with Peter's latest code,both in landscape and portrait mode.
Did you bridge J1 for 3.3v

Latest F4 Latest H7 FotS

phil99

Guru

Joined: 11/02/2018
Location: Australia
Posts: 2449

Posted: 06:24am 19 May 2025

Copy link to clipboard

Print this post

Found the problem, old eyes - diode bypass hadn't stuck at both ends. All good now.

PicoMite MMBasic RP2040 Edition V6.00.02RC23
Copyright 2011-2025 Geoff Graham
Copyright 2016-2025 Peter Mather

> RUN
ST7796S Test by disco4now 19-05-2025
F80000 RED F80000
F80000
FC00 GREEN FC00
FC00
------------
F8 BLUE F8
F8
F8
F8 BLUE F8
F8
F8
F8 BLUE F8
F8
F8
----------------
FC00
FC00
FC00
FC00
F8
WRITE TIME: 0.692
F80000
F80000
BLIT TIME: 16.933
CLS TIME: 0.091
> Blit 0,100,100,200,40,40
>

Footnote added 2025-05-19 20:40 by phil99

homa said Which pins are best to use with the Pico? Do you have any recommendations

It has the same connector as ILI9341, ILI9488 etc. so the pin setup described in the manual for them will work.
The PicoMite is very flexible, the only requirement is that System SPI uses appropriate SPI pins.

matherp
Guru

Joined: 11/12/2012
Location: United Kingdom
Posts: 10088

Posted: 06:49am 19 May 2025

Copy link to clipboard

Print this post

Great news on the diode, brilliant diagnosis

- I'll mod mine and assuming all is well remove the delay and all the related code. I'll put a note in the manual update about the diode.

phil99

Guru

Joined: 11/02/2018
Location: Australia
Posts: 2449

Posted: 01:12pm 19 May 2025

Copy link to clipboard

Print this post

With RC24

> RUN
ST7796S Test by disco4now 19-05-2025
F80000 RED F80000
F80000
FC00 GREEN FC00
FC00
------------
F8 BLUE F8
F8
F8
F8 BLUE F8
F8
F8
F8 BLUE F8
F8
F8
----------------
FC00
FC00
FC00
FC00
F8
WRITE TIME: 0.702
F80000
F80000
BLIT TIME: 8.971
CLS TIME: 0.107
>

An odd thing.
If left for an hour or so mine becomes unresponsive. Sometimes it comes good after 'CPU Restart' command but sometimes requires a full reset.

phil99

Guru

Joined: 11/02/2018
Location: Australia
Posts: 2449

Posted: 11:21pm 19 May 2025

Copy link to clipboard

Print this post

The odd behavior continues this morning.
Left 'gui test lcdpanel' running until it happened again, the screen went black then white. Ran Gerry's test program and as you can see it shows nothing wrong yet the screen remains white throughout the test.

Supply to the panel is 4.83V, output from it's reg. is 3.27V.

Edit.
Probed all connections from Pico to panel and found no problem.
Pulled Pico and panel out of their sockets, put hem back in and so far no problem.
Its annoying when a fault isn't found, it may return any time.
.
Edited 2025-05-20 22:48 by phil99

homa

Guru

Joined: 05/11/2021
Location: Germany
Posts: 462

Posted: 06:42pm 20 May 2025

Copy link to clipboard

Print this post

So my display works right away thanks to the tips on J1 and the diode.

However, I have noticed that ghost dots appear when I touch the screen. Circled by me in the picture. But I only painted on the right side of the touch. Except afterwards the circle for marking ;-)
Does anyone else have the same problems?

Matthias

phil99

Guru

Joined: 11/02/2018
Location: Australia
Posts: 2449

Posted: 09:25pm 20 May 2025

Copy link to clipboard

Print this post

Many screens do that with GUI TEST TOUCH, especially when touched near the edges of the screen.
However they give the correct coordinates with Touch(x) and Touch(y) functions.

Adding a delay reduces the number of random dots.

> cls:Do:circle touch(x),touch(y),2,1,,,rgb(white):pause 1:loop

Edited 2025-05-21 09:21 by phil99

homa

Guru

Joined: 05/11/2021
Location: Germany
Posts: 462

Posted: 04:29pm 21 May 2025

Copy link to clipboard

Print this post

Unfortunately, this is not the solution. The ghost dots appear if you don't press hard enough with the pen. Sliding lightly over it causes many of these dots to appear. With relatively firm pressure, it only happens when releasing (like a key bounce, only with offset).

Page 2 of 2

Print this page

To reply to this topic, you need to log in.