Some of my students have access to hardware for their projects and experiments, including various Raspberry Pi-s and alternative ¹ operating systems and accessories.

Unfortunately, given the way the Pi interacts with HATs ² and other similar devices via GPIO ³ pins, there is always the possibility that 5V will go where it shouldn’t and damage will be done.

In the case of the official Raspberry Pi Touchscreen Display, the device can be wired up to either receive or provide power to the Pi via jumper cables through the GPIO pins or provide power to the Pi via an included USB A port in the more “traditional” way.

When it comes to hobbyist hardware (and software!) there is an impetus to err on the side of giving the user as many options as possible.

When it comes to custom wiring, I think Murphy’s Law ⁴ should take precedence over hobbyist convenience. In other words, don’t even give us the option to power it via a method that will release magic smoke if done wrong.

There is some value in allowing users to power the Pi using the pins – and indeed this appears to be encouraged, as the enclosure that ships with the display only provides access to the Pi usb port.

At any rate, multiple options are available and inevitably, one of my students has configured one of the options that puts power where power should not go. As a result, neither the Pi nor the display are giving me any joy now when wired correctly.

It would seem that the Pi is beyond redemption – there is no display via HDMI and the SD card reader is unable to read cards.

The display is a happier story – there is no possibility to push power to or from it via the pins, but it seems perfectly happy to power on and pass power through via the USB port.

So just a quick note to anyone in a similar position – try your “dead” display with another (known working) Pi using the USB ports to provide power to both and you might find the display still has life yet.

1. read: Linux/unix (non-OSX)
2. “Hardware Attached on Top” – cute
3. “General Purpose Input/Output”
4. What you think is Murphy’s Law probably isn’t. To paraphrase: “If something can be done wrong, it will be done wrong”. This is very different to, “Anything that can go wrong will go wrong.” The latter statement implies some kind of (malevolent) magical agency to the universe at large; the former statement is a treatise on interface design. It says: “Design it so that every way you can do it is right, or so that it can only be done one way”. Early IDE cables were not built with consideration for Murphy’s Law – it could be connected two ways and if you wired it the wrong way, it would fail. Early USB ports can some consideration – it is hard to plug them in the wrong way. USB C is, in my opinion, the best design with Murphy’s Law in mind – any way the user connects it is correct.