Hi, have a standard device I put together regularly, which is an Atmel 328P (arduino) controller connected to a Digole Serial Display (2.6" IPS TFT).

The previous batch of displays purchased were marked V2 (version 2) on the PCB, and worked with no issues.

The new batch have just been received, and are marked V5 (version 5) on the PCB. When connected exactly the same way, with the same software, the controller starts continuous reset as soon as either SDA, SDL (or both) wires are connected to the controller. The controller and display are directly connected with 15cm wires, using arduino internal pullups from Vcc to SDL & SDA.

I notice that the V5 displays have different components and layouts compared to the V2, notably including resistors that seem to be connected to the Comm traces.

Do you have any idea what could cause this issue?

The display backlight comes on, but nothing else...no splash screen, nothing.

 Did you short the I2C jumper on the display? if you disconnect the SDA and SCL, can you see any message on the screen when power on?

 On the V5 PCB, it removed few 0 ohms resistors (which were going to used for another kind of display) and replace it with copper trace, and also fixed a reset problem for LCD panel when power on, the performance was improved according to our test.

If your controller keeps reset, the problem might be the watch dog, on the old V2 PCB, the display didn't use clock stretch technical when you reading data from display, and this will cause I2C read problem on most of controller, but on the new V5 PCB, the clock stretch used, so, if you are reading a touch screen click, you need to add clear watch dog function when your sketch waiting for the data returning from the display. Or just disable the watch dog.

 Thanks for your response Victor, yes indeed I have configured the display for I2C per usual with a dob of solder. After much testing and scratching my head, I think this is simply a DOA display. After connecting different displays from the same batch you sent to exactly the same microcontroller running the same code, they work fine. Only this display causes the microcontroller to reset indefinitely (which is a surprise I've personally never seen an I2C slave do that).

I will email you offline to discuss replacement policy, thanks for your response!