Pin_A4 not working reliably

jelodavid · Joined: 03 Apr 2005 Posts: 22 Location: Laguna Philippines

Hello All,

I observed with one of my design the PIN_A4 is not working reliably...
Sometimes the system hangs-up after several days.

mkuang · Joined: 14 Dec 2007 Posts: 257

I have a slow internet connection so I cannot check for you but you should check the datasheet to see whether RA4 is a Schmitt trigger input, it often is on some devices. That would mean you would not see a high unless the level is greater than 0.8Vcc and a low of 0.2Vcc.

Other members of this forum have pointed this out in many previous posts.

newguy · Joined: 24 Jun 2004 Posts: 1908

Pin A4, on some processors, is open drain. Not sure about the 16F876 - check the data sheet.

edwardelric · Guest

yup, the pin is open drain.just put a pull-up resistor tied to positive supply.

John P · Joined: 17 Sep 2003 Posts: 331

Whether the pin is open drain or not only affects its use as an output. It shouldn't affect it as an input.

Are you sure that somehow, some way, it's not getting reset to be an output? That would sure prevent it from reading the input state properly! Or getting set for analog input?

If you used a pin with an internal pullup (usually Port B) you could eliminate R28.

Ttelmah · Guest

A4, on this chip, is 'unusual', in that first, it is open drain, and second, it has no high level protection diode on the input. It is actually specified to allow inputs up to 12v!. However if you go over this, it really can cause fatal problems.
Not enough of the circuit is shown to know whether there is anythng likely to cause problems, but one has to ask if there may be first, a software problem with the code involved in looking at this pin. There is nothing in what is shown, but (for example), is it possible that if this pin changed state, after being tested, could the code get into a hung state?.
Then one has to ask how you know that A4, is the pin involved? Presumably by testing with it disconnected?. Have you considered whether this affects anything else?. How is the signal to which R28 attaches, actually connected to the PIC?. What drives X5-1?. Have you actually put a scope on RA4, and checked how high it is going when the opto is off?. The Schmitt nature of this input, does seem to be the most likely problem, though this should not cause a 'hang', unless there are code problems....

Best Wishes