GPIO input external pullup question

[TLin.5]

Testing a board with STM32U5A5.

VDD is 3.3V

I have configured one GPIO as input with an external pullup to 3.3V through a 470k ohm resistor.
The input reads high, but when I measure the input with a scope, I read 2.2V on the input, which is below 0.7VDD which is guaranteed as V_IH, but above 0.5VDD+0.2 which is specified by design as V_IH.

Is 470k ohm too high of a pullup resistance?

[mƎALLEm]

Hello,

Yes 470k is too high. It is may be seen as almost Hi-Z from the input side.

The range between 4.7k to 47k is acceptable. But consider the power consumption. The lowest pull-up value the more the current consumption you get.

[TDK]

470 kOhm should allow it to be pulled up. You may have something else going on there.

What happens when you use the internal ~40 kOhm pullup on the pin?

[Ozone]

> 470 kOhm should allow it to be pulled up. You may have something else going on there.

I would agree. Without any load, this 470k resistor should safely pull it up to Vdd level.

The 0.7V is what one would expect from a divider between this external 470k and an internal pull-down of about 50k.
Seems you have pull-down enabled on this pin.

[Danish1 · Accepted solution]

It sounds like you are using a x1 oscilloscope probe.

These have 1 Megohm input impedance, so 1 MΩ to ground.

Combined with 470k to +3.3V, I would expect about 2.2V on the pin.

Remove the probe and the voltage will go back up to 3.3V (but you won't see it!)

Many 'scope probes are switchable x1 / x10. If you switch it to x10 you should see perhaps 3.0 V (but might have to tell the 'scope that you're using x10, or scale what you see on the screen).

Imho 470k is a little high for comfort, particularly if the track to the pin is fairly long, as it might pick up electrical interference. Where I have had to use high value pull-up resistors for extremely low-power operation where e.g. a toggle-switch might pull the pin low (perhaps 10 MΩ) I might add a 1nF or 10nF capacitor to ground to kill any electrical interference. It killed the speed of response but was acceptable for that job.

[TLin.5]

Shame on me after looking into further...

The oscilloscope probe itself affected the measurement at 1x as it's only 1Mohm at that point...

At 10x (10Mohm) I get a better measurement.

[TLin.5]

I suppose the oscilloscope probe itself turned out to be a pulldown in this case...
D'oh.

[TDK]

Maybe change the accepted solution since 470 kOhm is fine for a pullup.

[TLin.5]

Changed solution.