Hacker News Reader

Comment by 34679 | Hacker News Reader

I feel like the flaws of the Pico 2/RP2350 should be advertised better. The Pico is great. I waisted half of an extremely frustrating day with a Pico 2 before suspecting the board itself was the problem, and confirming it with very specific searches that brought up threads about the issue. The internal pull-downs don't work.

Maybe it's my fault for not making it to page 1357 (!) of the datasheet, where the issue is described as "RP2350-E9".

https://datasheets.raspberrypi.com/rp2350/rp2350-datasheet.p...

View on HN

> The internal pull-downs don't work.

They don't work at all? How the heck did something that important get past testing?

Guess I'm not moving on from the RP2040 anytime soon...

Yeah, it's an A0 flaw. Word on the streets is that it was from modifying the pads to be 5V tolerant.

Do later steppings have this issue fixed? Or will they not fix it for backward compatibility?

> They don't work at all? How the heck did something that important get past testing?

Because that's how all hardware is; I complained about this sort of thing a few days ago: https://news.ycombinator.com/item?id=43202090

> Guess I'm not moving on from the RP2040 anytime soon...

Doesn't matter what you move to, there's still going to be 2000 pages of datasheets+errata, and one line in the middle of all of that will tell you "This does not work".

That's why for hobbyists it's best to stick to devices with a large community around them, who surface niche problems in community forums.

However, with everyone moving to Discord, this will no longer be useful too...

This is unfortunately not emphasized with many breakout boards. It pays off to skim errata sections at the start of a project. All hardware has errata, and it ranges from incorrect details and minor malfunctions all the way through to broken peripherals and all manner of critical malfunctions.

To be clear, this isn't a problem with the breakout board. The issue is inside the microcontroller itself, which, iiuc, is why there is such a long lead time on fixing it.

While I agree that reading the errata absolutely should be on the "mandatory TODO before you start" list for any project (esp with new/unfamiliar hardware), one should be equally aware that manufacturers aren't always transparent about the defects in their products, sometimes failing to note that entire subsystems, frankly, don't work.

Something is just not right in the Raspberry Pi design team if at this point ESP32 have less unpleasant surprises. New RP micros always turn out have disappointing bugs that make them just shy of considering using for actual embedded projects.

What, "Increased leakage current on Bank 0 GPIO when pad input is enabled"? How does that relate to the pulldowns?

If you scroll down to the next page, it gives more details, but one of the criticisms I've seen is their obtuse language in the errata.

Oh wow, OK, I was wondering why that page was so short, thanks.

Reply on news.ycombinator.com