Teensy3.2 + ILI9341 display not working

[schrandt]

I have used the teensy3.6 for a lot of my projects and never have had any issues working with the Color 320x240 TFT Touchscreen display sold by PJRC. This time, I have a teensy3.2 and am only trying to run the example file "graphicstest" from the "ILI9341_t3" library. I have made the hardware connections as detailed in the display's page on the PJRC site. I have 2x of the teensy3.2 and LCD module so I made one up on a breadboard and another with wires directly soldered. The wires in both cases are ~4" in length.

The code uploads happily and I can see feedback through the serial monitor. However, the screen remains white most of the time. Sometimes it will display some black background with the text mirrored over the X-axis and inverted in color (appears cyan instead of yellow, so i guess not an inversion?) and white streaks will cascade through the screen until eventually the screen seems to go out of sync and I have no display anymore except the white background. I will also note that occasionally the screen appears black with yellow text (as coded) and in the right orientation but this usually fades quickly to white again. I had assumed this was maybe due to the wires on the breadboard being bad connections which is why I soldered the second one, but am seeing the same results with the soldered set as well.

This has led me to believe the wiring is correct (display sometimes shows proper screen), but I don't know what else to try next.

I'm new to this thread so don't know how to post a picture of my connections or the display.

Thank you!

 

[Theremingenieur]

Easiest way to diagnose the problem with your setup is taking a logic analyzer and an oscilloscope to check the signal quality arriving at the TFT in general, making sure that there is neither ringing nor slow signal ramps, an then the assertion timing of the CS and D/C Pins, the correct reset pulse at startup, the correct clock speed, and finally the MOSI data traffic.
Just do what every technician or engineer would do...

There are many working projects with the T3.2 and that display out there, so it must be either your wiring or your code. But you did show none of both.

 

[schrandt]

I don't really have easy access to those tools unfortunately. I know I have seen many working examples of this same combination of hardware, which is why I'm so puzzled.

I found that adding in a short delay() after any graphics call prevents any problems. I don't believe that this should be necessary but it is an acceptable solution for what I am doing in this project.