From: theom+news@chiark.greenend.org.uk
The Natural Philosopher wrote:
> On 06/05/2025 14:19, Theo wrote:
> > Andy Burns wrote:
> >> the BusPirate folks had issues with their BP5XL boards and needed to
> >> junk a batch, I see that redesigned as the BP6 it's now on sale ...
> >
> > Yes, that's the kind of application where it's a problem. You attach GPIOs
> > to random stuff and intend to probe what's there. That means you flip
> > through the internal settings for input/output/pullup/pulldown and listen
> > out to how the input changes.
> >
> > The leakage affects that kind of application badly (input pins should be
> > high impedance but they aren't here) but if you are putting the chip in a
> > board where you already know what's on the other end of the pin, you can
> > design the circuit appropriately.
> >
> > RPis are more likely to be used in situations where you take the MCU and
> > plug in and out random things onto the pins which are affected by it,
> > compared with other MCUs where they only ever get put on vendor PCBs where
> > everything is predetermined in the schematic. So it's really a hobbyist
> > focused problem rather than a wider problem.
> >
> > Theo
> All my Pis end up on my design of PCB where stuff is either disabled or
> its strictly controlled
>
> I cant see why anyone would enable an input pin and leave it floating.
> Its simply bad design
If you're making a logic analyser, that's what you do. That's effectively
what the Bus Pirate is. (It will also generate some of the protocols it
analyses, but that's a different feature)
Anyhow they didn't leave the pins floating - they enabled the onboard
pulldown resistor. The problem is that the pulldown has unexpected side
effects.
Theo
--- SoupGate-DOS v1.05
* Origin: you cannot sedate... all the things you hate (1:229/2)
|
