I recently bought a HXD8357D 3.5" TFT and got the Adafruit software test running on an UNO. BUT I want to use a Wemos D1 mini. I made the following changes:
WIRING
display Wemos
CS => D8 (GIPO15)
D/C=> D1 (GIPO5)
RST=> D3 (GIPO0)
and the 5 standards.
SOFTWARE
I MODIFied graphicstest to reflect above:
// These are 'flexible' lines that can be changed #define TFT_CS 15 #define TFT_DC 5 #define TFT_RST 0 // RST can be set to -1 if you tie it to Arduino's reset
OUTPUT:
8HX8357D Test!
Display Power Mode: 0xFF
MADCTL Mode: 0xFF
Pixel Format: 0xFF
Image Format: 0xFF
Self Diagnostic: 0xFF
Benchmark Time (microseconds)
Text 611562
Lines
Soft WDT reset
Do you see anything on screen while this is happening?
Please post the code you are running, or a link to it. IMPORTANT: read the forum guide to find out how to post code and links. EDIT: is this it?
My guess is that because the code is performing speed tests, it is running a series of tests in tight loops and causing the esp's watchdog to be triggered. Hopefully it's just a case of putting some "yield()" calls into the code at points which won't affect the timing of the tests, to keep the watchdog quiet.
It is always wise to start with the Bit-Bang constructor.
Make sure that you connect each display pin to the correct Wemos D1 Mini pin
And that you specify D1 Mini as your "Board"
If the HX8357-D is wired correctly, the MADCTL, Pixel Format, ... should show real values e.g. PIXFMT as 0x05 or 0x06
Yes, the Watchdog is kicking off. Insert yield() in strategic places.
Personally, I would start with Bodmer's TFT_eSPI library and examples. It is designed for ESP8266.
Thanks for the thoughtful advice.
Paul, The display shows nothing but backlight, and you're right about the origin of the code.
Since the first thing the sketch does is read the driver configuration from the TFT, which fails, I don't think this is a "yield" issue.
David, the software constructor fails.
THIS SHOULDN'T BE SO HARD!
Thanks again.