# Voltage transition sense - rapid raise or fall

Which means do we have to sense (relatively) fast voltage transition from one value to another. The exact point of transition start, transition end and the intermediate state. Which schematic gives us indication of these points.
And the special case: how do ICs sense voltage raise or fall?

Edit: Last edit turned question to the wrong direction, so I rolled it back and rephrased.

• Do you mean sense an edge transition or the supply voltage? Dec 17, 2010 at 18:04
• Yes, I mean the edge - transition from low to high and vice versa. Dec 17, 2010 at 19:16
• @PF4Public, The title "Raise/Fall sense" is a little vague. Perhaps incorporate some of [ Fake Name's previous edits](electronics.stackexchange.com/posts/8015/revisions) into the current version. Different MCUs (microcontrollers) will detect edges in various ways, not excluding any of those used in other ICs. You won't be able to copy their methods exactly using discrete components, but you can come close (or mimic). Dec 19, 2010 at 0:40
• But "Integrated Circuit" - is just like usual circuit, except it's extremely compact - "Integrated". So it employs usual elements with slightly different parameters. Well, until you're looking deep inside at microscopic effects. I agree, the title is not just "little vague" - it ambiguous :( I'm thinking about better replace. Dec 19, 2010 at 0:48
• You rolled back all my clarifications: MCUs are a type of IC, edge-trigging is used in all of them. Furthermore, sense is a very inaccurate description. The term "Edge-Triggering" specifically refers to exactly what you are describing. It is the correct term for what you are asking about. Dec 19, 2010 at 5:01

See Schmitt Trigger - http://en.wikipedia.org/wiki/Schmidt_trigger basically it's an amplifier circuit with positive feedback.

• by basically, you mean it is an amplifier circuit with positive feedback. Dec 17, 2010 at 17:25
• @Kortuk yes, basically :p Dec 17, 2010 at 17:28
• your response just made my day. Dec 17, 2010 at 18:21

This really depends how far down you want to go. Keep in mind that every MCU that you will use is composed entirely of simple elements. Specifically, you will find transistors, capacitors, resistors, and diodes in every microcontroller and microprocessor since they stopped using vacuum tubes.

Without getting into digital logic fundamentals, a MCU can tell if an input changes by storing the digital value of a pin (1 or 0), and then comparing that stored value to the current pin value. The logical comparison is typically an exclusive-or.

You could make a very simple edge-detection circuit by using a resistor and capacitor in series to store the input voltage. Then measure the voltage difference between the capacitor and the input - if the difference is close to zero, then there has not been a change. If the difference is large, then the input has just changed.

• Here;now we've gotten into digital logic fundamentals. ;) Dec 18, 2010 at 0:14

I don't understand the insistence on using "MCUs" instead of "ICs". MCUs are ICs, and many ICs also have digital inputs. Furthermore, a MCU is basically just a specialty IC, and you can actually make a MCU out of a bunch of ICs.

The behavior of the inputs and outputs of a Schmitt trigger are thus:

Both axes are in voltage. (Shamelessly pillaged from wikipedia)

As you can see, the Schmitt-Trigger is specifically designed to not have an "intermediate state".

The "transition start, transition end and the intermediate state" are all part dependent, and there is no blanket answer. Read the datasheets for your part for the answer.

Here is a simulation of how edge triggering works at the logic-level:

http://www.play-hookey.com/digital/rs_nand_flip-flop.html

vicatcu has already covered the analog underpinning to building a schmitt trigger.

Interestingly enough, for logic devices without buffered/schmitt-trigger inputs, if you wrap negative feedback around the device, you can actually use it in an "analog mode", kind of like an amplifier.

It's not a good idea, though, and it may damage some parts. Also, there may be wide variations between parts from the same manufacturer, since analog performance is not specified or controlled.

• not have an "intermediate state". My question demands to feel that state, notice that at last. Dec 19, 2010 at 6:47
• I would call an MCU the opposite of specialty IC. You make an IC that can be programmed to do many different things so that the company can fabricate one chip for many many different applications. Dec 19, 2010 at 16:55
• A MCU is a specialty IC which implements a very complicated, specialty logic system which, when fed specific sequences of bits, can do many different things. It's still very specialized, and very much an IC. The applications may be general in nature, but the hardware is specialized. Dec 20, 2010 at 8:24