Noise from floating CMOS inputs.

[dnb]

We all should know that if a CMOS IC input pin is left floating, it acts like an aeriel, picking up whatever noise it likes.

I noticed I had a problem with my LCD switching into one of the table edit modes some time ago.  What I hadn't thought about was that these problems started when I removed the PS2 keyboard from the car when I no longer needed it.

It seems the PS2 keyboard input has been left unterminated on my VEMS - is this because the pull-up (or pull-down) resistors in my ECU have died (they're probably on-chip, so nothing can be done) or is there a bug in the firmware which causes them to be turned off? 

I seem to have fixed the problem by making up a PS2 plug which permanantly grounds the data line.  It's been a couple of days now and I have had no table-mode LCD funnies.  Prior to this, I could rely on at least one a day happening.

I also wonder if this mechanism would allow table corruption in severe cases - if the noise just happened to provide the right signals to edit something...

[rob@vems.co.uk]

I remember this question from something that came up from a question I had about the serial port, and as I understand it, there is a form of locking to ensure that this eventuality never happens.

I can see no earthly reason for having an LCD and keyboard on an ECU.

[dnb]

I can see no earthly reason for having an LCD and keyboard on an ECU.

That's rather my point - removing it caused the problem!  Of course, if it wasn't allowed to be there in the first place then there wouldn't have been a problem in the first place...

I quite like the LCD though, and since there aren't any inputs from it, the ECU should remain tolerant to noise.

Not sure what you mean by "locking"?  Unless you're refering to the table corruption thing - Kind of like a "cat walking on keyboard" detector...  (I need one of those!)