Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
06/21/05 11:09
Read: times


 
#95451 - I understand what Mehdi is saying.....
Responding to: ???'s previous message
I think that I understand what Mehdi is trying to say here. He is desiring how to automatically detect if a logic signal to the microcontroller is in the no-mans land between Vil and Vih. Unfortunately there is NO reliable way to detect this situation without additional hardware. The reason of this is due to the fact that the 1/0 state that you can read from the port input bit does not provide any information about the signal state if it is in the no-mans land. If the logic of the port bit was re-done so that it could report as a tri-level input as opposed to a binary state then maybe it could be that the 3rd level relates to the signal in the no-mans zone.

However there is no designed in predictability in the chip to assure that binary port bit is inside the no-mans land or not!!

Here are some things that could be considered to enhance a system design so that inputs in the no-mans land are detectable:

1) Use a microcontroller that has an A/D converter with a digital / analogue MUX on the bank of port pins that want to be checked. The A/D can read out and represent the pin levels as a byte value (in case of 8-bit A/D).

2) Add an external circuit consisting of some comparators configured as a window detector that can ascertain if the signal is inside the no-mans zone or outside of it.

3) Add an AC bias to each input signal via a low cost capacitor from an AC source. Choose the capacitor so that it can couple the AC signal to the port pin when the pin is in its faulty (not properly connected) state. At the same time fix it so that the normally connected driver that would connect to the input pin has a low enough output impedance that it can successfully drive the input pin to valid levels despite the presence of the capacitor coupled AC signal. This coupled AC signal can push the faulted input port pin to a semi-predictable state above Vih and below Vil at the frequency of the AC signal. Lets say you used a 50 Hz AC source. You could then write and provide some software to detect a 50 Hz behavior on the port pin(s) to determine if they are not properly connected.

Hope that helps
Michael Karas


List of 49 messages in thread
TopicAuthorDate
Invalid input voltage?            01/01/70 00:00      
   read the specs            01/01/70 00:00      
      Re:read the specs            01/01/70 00:00      
         Noise margins            01/01/70 00:00      
            Re:Noise margins            01/01/70 00:00      
            MAKE the signal determinate            01/01/70 00:00      
   0 or 1            01/01/70 00:00      
      Fuzzy logic...?!            01/01/70 00:00      
         zone between VIL and VIH            01/01/70 00:00      
         zone between VIL and VIH            01/01/70 00:00      
            The 'grey' zone            01/01/70 00:00      
               Illogical signals help us to having...            01/01/70 00:00      
                  Well, I'm lost!            01/01/70 00:00      
                     Re:Well, I'm lost!            01/01/70 00:00      
                        Scopes            01/01/70 00:00      
                           Re:Scopes            01/01/70 00:00      
                     7 or 9 value logic            01/01/70 00:00      
                  I understand what Mehdi is saying.....            01/01/70 00:00      
                     Ah            01/01/70 00:00      
                     excellent, Michael            01/01/70 00:00      
                        triad            01/01/70 00:00      
                     wow, Mike            01/01/70 00:00      
                     There's the rub            01/01/70 00:00      
                     Detect What?            01/01/70 00:00      
                     Absolutely but!            01/01/70 00:00      
                        a question based on a guess            01/01/70 00:00      
               Not debounce. Hysteresis/schmitt            01/01/70 00:00      
                  Re:Not debounce. Hysteresis/schmitt            01/01/70 00:00      
                     Well, then ADC.            01/01/70 00:00      
         It Could be            01/01/70 00:00      
            "undefined operation"            01/01/70 00:00      
               Undefined            01/01/70 00:00      
                  Exactly            01/01/70 00:00      
   Thanks...            01/01/70 00:00      
      based on the above            01/01/70 00:00      
         Erik:Based on mehdi's emphasis            01/01/70 00:00      
            Thanks            01/01/70 00:00      
            Wow.            01/01/70 00:00      
               I describe it!            01/01/70 00:00      
                  Service            01/01/70 00:00      
                  I will never            01/01/70 00:00      
                     I will always            01/01/70 00:00      
            the missing if            01/01/70 00:00      
   Port 0 output when no power            01/01/70 00:00      
      Port 0 to Port 0.            01/01/70 00:00      
         Voltage Divder of 10k            01/01/70 00:00      
            nope            01/01/70 00:00      
               2 line solution            01/01/70 00:00      
                  I do somethimes my thumbs have another o            01/01/70 00:00      

Back to Subject List