General • Re: RP2350 Errata E9 and PIO

Somehow RPI has not been able to replicate this yet, but at this point it seems pretty obvious that there is a serious problem.

That does feel strange when multiple people say they have demonstrated latch-up, trivially easily, with just a resistor.

It's not clear to me if Raspberry Pi are unable to demonstrate the latch-up others are seeing, or can, but don't consider it a problem, or believe such latch-ups are covered by RP2350-E9 so, as far as they see it, there's nothing more to say or add.

If they can't replicate the latch-up there could be some, so far unknown, factor which causes different results.

Could it be environmental; having a domestic rather than commercial mains supply, using LED lighting rather than incandescents, halogen, fluorescent or CFL, encountering Wi-Fi or being shielded from it, living closer to or further from a radio or TV transmitter, being in the middle of nowhere compared to an urban environment, living in a Faraday cage or not ?

Or could it just be the testing environment; Raspberry Pi undoubtedly has kitted out R&D and test labs where mere mortals will be testing on their desktops, in a kitchen, shed, or other random place.

Or maybe it's the testing regime, the test program or its reporting mechanism; perhaps whether connected via USB using a terminal emulator or a Debug Probe ? Are they using random USB supplies pulled from the junk drawer, plugged into their PC, or using a quality calibrated bench supply ? Are they powered by VUSB or other means ?

I'll admit there's a lot of clutching at straws there but I can't see what creates such a divergent view; that latch-up can be easily demonstrated and it's a major problem against it not being reproducible or it not being a problem when it happens.

And let's not forget there seems to be plenty out there who haven't joined the "Me Too" throng.

Are they still getting to grips with their new purchase, haven't done things which could provoke latch-up, have somehow avoided the issue ?

Maybe the easiest way to put this matter to rest is for Raspberry Pi to invite someone willing and able to head to Pi Towers and do a Show and Tell with hardware they can demonstrate causes latches-up ?

Or for them to arrange a home visit to see what's going on for someone experiences sees latch-up.

I guess it's relatively easy not to encounter the issue if you only connect signals that are of low impedance (to ground, at least) to a Pico. However, how you could perform e.g. leakage current testing (specified in the datasheet) on a sample that exhibits the latching issue, and not notice it, is beyond me.

#include <stdio.h>#include "pico/stdlib.h"int main(){    stdio_init_all();    for(int buc = 0; buc < 16; buc++)    {        gpio_init(buc);        gpio_set_pulls(buc, false, false);    }    while (true) {                sleep_ms(1000);    }}

Statistics: Posted by Dr.Gusman — Sun Sep 01, 2024 12:03 am