Any definite news on causes? I’m still wary to even plug my replacement inkyphat in (playing with the Unicorn Hat HD instead at the moment) but at some point I’ll want to, not least because if this replacement fails we’ll have to go around again.
Are the returned boards actually damaged, is it something on the boards themselves, or is it a software problem, or something the Pi or software is doing to damage them?
Nothing definite yet, I’m afraid. We’re still conducting some multi-day tests to try and re-create the problem. It’s proven extremely difficult to make them fail on demand. It’s a going concern!
I received my replacement inkyphat yesterday.
I plugged it on the same pizero on which the previous inkyphat failed
I am running the weather.py example using the up to date library.
The example run periodically each minute for 14 hours without any problem so far.
We seem to have narrowed down what’s causing the issue with Inky pHAT. That is to say we have managed to damage multiple brand-new displays in the same fashion, and they’re progressively getting worse exactly as you’ve reported. To counter this, we’ve also managed to run multiple displays for the same length of time, under the same conditions and have them not display any faults at all.
Sorry this process has taken so long. Our enemy in these tests has been time- we have to balance getting this solved swiftly, with how long it takes for issues to manifest.
Right now we need to explore in some detail why it’s happening and then come up with a fix, report our findings to your good selves and get replacements to those who need them.
I think it goes without saying ( at least I hope it does ) that we’ll do what we can to make this right.
Or even heat from a soldering iron interacting with the photo-luminescent chemicals in the display, although the delayed error would be a mystery this way.
So many variables! 0_0
So far my replacement is holding up, on the same Pi Zero, same Python code, same schedule (weather.py patched to turn it upside-down, run hourly), but with the updated driver.
I await information with interest though as I’m still treating this with some caution before thinking it’s OK.
Slightly later than Monday for an update, but we had a bit of a plot twist in that our tests eventually disproved our leading hypothesis. Science hurts! D’oh.
We’re now running another test over this weekend. Stay tuned!
OK, add me to the ‘streakers’ club. On 5 minute updates, ran for about a week, then streaks started to appear. More streaks appearing, but it’s not unreadable (yet).
What news Pimoroni? How are the investigations going? Do you want some of these defective units back for testing?
Yes, we stopped shipping/producing them so I’m not surprised the seller was unable to replace it. If you could please hang in there until we’ve sorted the problem, we’ll make sure we get your replacement.