PIC – RB0 Interrupt pin Schmitt Trigger input

[elexhobby]

I am referring to Peatman’s book on PIC uCs. He says that when the RB0/INT pin is used as an interrupt input, it is automatically configured as a Schmitt trigger input, triggering on the input edge regardless of the rise (or fall) time. What do you mean by this?
Even among the RA0 – RA4 pins, RA0 – RA3 are TTL while RA4 is Schmitt trigger. What’s the difference b/w them? RB0 is infact TTL / ST ?!
Please help

 

[Nigel Goodwin]

I am referring to Peatman’s book on PIC uCs. He says that when the RB0/INT pin is used as an interrupt input, it is automatically configured as a Schmitt trigger input, triggering on the input edge regardless of the rise (or fall) time. What do you mean by this?
Even among the RA0 – RA4 pins, RA0 – RA3 are TTL while RA4 is Schmitt trigger. What’s the difference b/w them? RB0 is infact TTL / ST ?!
Please help

Try doing a google search for 'schmitt trigger', it will explain everything you want to know!. But basically, a schmitt trigger switches more 'reliably' than a conventional input - if it's fed from a slowly changing source, rather than a proper logic level.

 

[elexhobby]

WHy wudnt it be more reliable, with the i/p changing slowly, if it is configured as TTL. As far as I know, Schmitt trigger helps when there are many fluctuations in the i/p as it crosses near the threshold. But if the i/p is changing monotonically, TTL & Schmitt trigger make no difference. Let me know if I'm wrong...

 

[Nigel Goodwin]

WHy wudnt it be more reliable, with the i/p changing slowly, if it is configured as TTL. As far as I know, Schmitt trigger helps when there are many fluctuations in the i/p as it crosses near the threshold. But if the i/p is changing monotonically, TTL & Schmitt trigger make no difference. Let me know if I'm wrong...

A standard logic input will switch at a specific point - for the sake of this thread, let's presume it's 2.5V (half way for TTL logic).

So, if you have an input of 2.5V - what is the output?, high or low?, and is it 100% guaranteed to ALWAYS be exactly the same. In fact, the switching point is so vague that you can use CMOS gates as an analogue amplifier!.

For a schmitt trigger input, and I'm just assuming values again!, the pin doesn't read HIGH until the input reaches say 3V, then it actually SNAPS to be high, rather than a vague change. If the pin then drifts downwards, it won't go LOW until say 2V, when again it again SNAPS low. The difference between the two switching points is called hysteresis.

As I said before, there's plenty on information on the net, try http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/schmitt.html for one example.

 

[elexhobby]

Thanks a lot, Nigel. I got your point. So if rise or fall time is too long, had the pin been configured as TTL, there would be an ambiguity regarding the precise timing of the interrupt when the i/p is between 0.8V & 3.5V (VIL & VIH).

However I didn't get this point of yours.

In fact, the switching point is so vague that you can use CMOS gates as an analogue amplifier!.

It won't behave as a linear amplifier, am I right?

 

[Nigel Goodwin]

It won't behave as a linear amplifier, am I right?

Yes it will, it's a common usage, there's even a link in a recent thread on these forums to a MicroChip application note for X10 (power line data communication) that uses them in this way.

However, there's different degrees of 'linear', you wouldn't want to use them for a high quality audio preamp?.