Controlling floating pins

Gabriel · Joined: 03 Aug 2009 Posts: 1067 Location: Panama

hello all,

I'm working on a small project, where the only requirements are to turn on and off 2 LEDs for debugging and communicate with a PC through RS232

I'm using a 16F877A... as you can see I am only using 4 pins on the chip.
2 LEDs and RX&TX

I'm having alot of problems with noise and static and I suspect most of my problems (resets and so on) are caused by floating pins.

I have set all my other pins as Outputs, using Set_Tris except the Usart input, in order to avoid my pins floating.

Is this Ok?

Thank you in advance.
_________________
CCS PCM 5.078 & CCS PCH 5.093

mkuang · Joined: 14 Dec 2007 Posts: 257

Gabriel · Joined: 03 Aug 2009 Posts: 1067 Location: Panama

If you consider my living room a lab and my dinner table a work bench... I guess I don't have any static problems hahahaha.

Its winter here... cold=coats=layers of coats= MASSIVE STATIC BUILDUP.

My code and my hardware work fine... but only when they want to.... I believe its because of noise... either from my power line.. or me being a freaking HV source of doom.

I am not using ICSP so I do have to touch my pic constantly....thus I think, the source of my problems.

I don't think I need fast io if I am only setting them as outputs...to avoid floating pins? Do I?

Will setting them high or low make any difference?

I am assuming setting them as outputs is better than inputs... to prevent noise and erratic behavior that is.

Thanks for your help

g
_________________
CCS PCM 5.078 & CCS PCH 5.093

mkuang · Joined: 14 Dec 2007 Posts: 257

You can use output_high(your_favorite_PIN) or output_float() and then use a 50K resistor to GND. Or you can use output_low() or output_float() and use a 40K resistor to +5V.

I personally have not had issues setting unused pins as inputs (without any pullup or pulldown resistors) but since it is a practice that is frowned upon I will go with the book so to speak and suggest you do not use this method.

Gabriel · Joined: 03 Aug 2009 Posts: 1067 Location: Panama

hmm...

I'm trying to avoid the resistors... because... hahaha i don't have any!
and by that i mean having to tie everything high or low...

I got very limited resources.... stores don't open after 7pm... i get off work at 7 pm... they don't open before 7 am... i go to work at 7 am....

and electronic shops... are not very fond of opening Saturday afternoon...
which i cant go to ethier..... cause I'm working....

I can jump all my pins to ground.... i do have jumpers hahahhaha

How about .. setting them all as inputs and tying them all to ground?

... I sent there like a tristate option.... that does exactly what i want?

g
_________________
CCS PCM 5.078 & CCS PCH 5.093

asmboy · Joined: 20 Nov 2007 Posts: 2128 Location: albany ny

got a PDF of the schematic - and info on your POWER supply?

using a MAX232 / HIN202 / similar RS 232 IFC ?

what is receiving your rs232 ? A pc ?

knowing that - i may better help u

mkuang · Joined: 14 Dec 2007 Posts: 257

gs · Joined: 22 Aug 2005 Posts: 30 Location: Ioannina - Greece

Do you have any capacitors on the circuit? MAX232 produces some noise that needs filtering.
_________________
www.hlektronika.gr

Gabriel · Joined: 03 Aug 2009 Posts: 1067 Location: Panama

hello,

so what is the diference between just setting them as outputs Vs. output_float?

Guest · Posted: Mon Aug 10, 2009 11:09 am

Hi,

There were tons of suggestions/questions in this thread that you basically ignored. An answer is not necessarily 'definitive' just because it is what you want to hear.....

Dave

mkuang · Joined: 14 Dec 2007 Posts: 257

Gabriel · Joined: 03 Aug 2009 Posts: 1067 Location: Panama

i didnt ignore anything...
and i am thankful for everybodys help... thats why i joined... to get helped. ignoring peoples suggestions would be against my own benefit.

i was just complimenting the user. his answer i thought was the most acurate for my situation... mabey he worded it better...

in anycase.. thank you all once again

g
_________________
CCS PCM 5.078 & CCS PCH 5.093

Gabriel · Joined: 03 Aug 2009 Posts: 1067 Location: Panama

Mkuang:

so, i should avoid setting them as inputs....
but i should set them as output_float, which makes them inputs? hahahaha

i think i get it...

output_high / low... set them as outputs... common outputs...source/sink.

output_float ... is the high impedance "third" state if you may... (this answers a question on a previous post)...and they are basically inputs because inputs are also high impedance...

and setting them as inputs using tris... well.. they are inputs, and noise will kill me...

thats what you meant right?

using output_float, even though they are in their high impendance state i still have to tie them high/low?
i was under the impression that by setting them in this state... the pin basically became an open circuit... dead pin, N/C sort of thing...

thank you for you help

g
_________________
CCS PCM 5.078 & CCS PCH 5.093

mkuang · Joined: 14 Dec 2007 Posts: 257

I am sorry I should have worded it better. What I meant was you should avoid setting them as input as much as possible. HOWEVER, since all you have are jumpers and no resistors available, you can set them as inputs by jumpering them all to GND or 5V.

The two ways of setting them as inputs that I know of are:

1) using the output_float() function (yes the name sucks but that is what CCS calls it).

2) using the fast_io and set_tris functions.

Using either method the pin is an input.

A preferred way is to set them as outputs and use pull-up or pulldown resistors. But remember that never drive a pin high if you use a pullup and never drive a pin low if you use a pulldown. But since you don't have any resistors this is a moot point.

Also let me reiterate that I personally have not had problems setting unused pins as inputs and floating them. But that's just me and the kind of work environment I am in. Your experience may vary.

FvM · Joined: 27 Aug 2008 Posts: 2337 Location: Germany

Seems like it has finally been clarified how to set unused pins to output.

It would be interesting to know, if this changes anything regarding your reported issues? Although I completely agree with the suggestion to avoid floating pins, I don't expect, that they cause any problem except additional current consumption (some mA).

Particularly, why should we expect problems with static electricity related to floating pins? Floating pins aren't connected to anything, it should be difficult for electrostatic discharges to enter a floating pin directly. If they would, either the existing on-chip protection diodes are capable to absorb it or they can't. But don't assume, that enabling the low or high driver of the respective I/O pin changes the path of the several 10 A discharge current. Without feeding considerable currents to a floating pin, the level always stays within the limits set by the chip's substrate diodes, supply voltage +/- 0.6 V.

In most cases, where electrical interferences or "noise" are affecting a single chip processors's operation, they are causing voltage drops or peaks at the supply pins or possibly hitting the reset input directly.