Missing "hysteresis voltage between Schmitt trigger switching levels" in datasheet

[VolkerKugler]

In the datasheet of the STM32F765ZI and many other datasheets I have read, the specification value of hysteresis voltage between Schmitt trigger switching levels is missing. There is a comment (8) under the table (here table 66 I/O static charasteritics)  which should be the right comment, but the (8) was used in the row "I/O pin capacitence" where a capacity was specified and not the hysteresis voltage i search for.

[Peter BENSCH]

Hi, Volker,

that might be a typo, as footnote (8) relates to Vhys, 4th row of table 66, currently on page 158 (rev 6).

Does it answer your question?

Regards

/Peter

[VolkerKugler]

Thanks,

and this has no tolerance? We have measured the hysteresis on two pins.

One pin has 340 mV, that was nearly the specified value of 331mV (10% VDD), but the other pin has only a hysteresis of about 96 mV.

Best regards

Volker

[Peter BENSCH]

Please do not expect the values from the built-in Schmitt trigger that are made possible by specialized, precise comparators. The Schmitt trigger itself is only used to restore signal edges and to avoid oscillations if the input signal is between the voltage limits.

Where is your input signal coming from that requires an exact value for the hysteresis?

Your question probably relates to this one, right?

It is certainly a better approach, e.g. use an ADC channel instead of trying to use digital inputs as comparators.

Regards

/Peter

[VolkerKugler]

I tried to measure directly at the pins PF0 to PF7 of a nucleo STM32F767ZI to get a better idea of the real situation.

What I can see is, that PF0 to PF5 toggle extremly when starting at 0V and turn up the voltage slowly. They start toggeling at 1,50V. From High to Low they start toggeling at 1,55V. PF6 and PF7 don`t toggle and they have a hysteresis which is not lower than 100mV.

Here the measured values:

 

As you see, the 10%VDD Hysteresis (VDD = 3,13V) does not exist on 6 of the 8 pins.

Best regards

Volker

[VolkerKugler]

Can I believe on the levels for VIL and VIH given in the datasheet with "guaranteed by design", when the hysteresis (guaranteed by design) is not exactly like defined in the datasheet?

[Peter BENSCH · Accepted solution]

Well, at least you've found some hysteresis, which is the intention of the Schmitt trigger. Please refer to Table 66 and the Parameter column that indicates for Vhys: FT, TTa and NRST I/O input hysteresis. It looks like you want to use this for standard GPIO, i.e. not for the 5V tolerant, 3.3V tolerant ADC or NRST pins.

However, please do not stick to the use of these simple Schmitt triggers for analog measurement functions, as a measurement with an ADC channel should be the better approach. You could maybe trigger the ADC by a timer and use the analog watchdog function of the ADC to get an interrupt when the input voltage goes

beyond the user-defined, higher or lower thresholds.

Regards

/Peter