I'm building a project (using a PIC16F627) and have reached the stage where
I'm ready to design the board layout, so need to decide what I'm going to do
with the unused port lines. As you'll be able to see, I don't have a lot of
experience.
RA0,1,2,3 are used as O/P's and RA4 unused . so can I leave RA4 as an output
and leave it unconnected .. or should it be terminated in some way?
RB0 is used for an EXT Int . RB1,2,3,4 as I/P's .. and RB5,6,7 unused.
I'm assuming I should leave RB5,6,7 as I/P's and just tie them directly to
Vdd.
I'm aiming for a tiny pcb with no wasted space, so am trying to avoid any
components I don't absolutely need to have.
There are two schools of thought on this and there are valid reasons for
both.
1. Pins to input and tied to a supply line.
2. Pins to output with dummy data on them.
If the pins are inputs and tied then when you read the port as a whole you
will get some absolute data that can be checked against a reference so
therefore validating the input.
If the pins are outputs then there is always room for expansion / debugging
connections to other 'Stuff' and can be used on s future enhancement of the
product for zero cost.
Roger Weichert wrote:
> I'm building a project (using a PIC16F627) and have reached the stage
> where I'm ready to design the board layout, so need to decide what
> I'm going to do with the unused port lines. ...
>
> I'm aiming for a tiny pcb with no wasted space, so am trying to avoid
> any components I don't absolutely need to have.
In that case make them outputs and make sure the firmware drives them. If
any are input-only pins, then tie them to Vss or Vdd directly or thru a
resistor.
If you have a little extra space, bring the unused pins out to pads and
still make sure they are driven. If you have a little more extra space you
can add 100Kohm pullup or pulldown resistor pads. You don't actually have
to load the parts, so it's only a board area issue. Any changes to make use
of the pins would require rework and firmware change anyway. The parts can
be added to the pads and the firmware changed at that time to make pins
inputs if that's what's needed.
If you were able to fit the resistor pads then the next step up is to
install at least a few of the resistors and make the pins inputs. Unless
this is a very high volume project, a few cents for the extra resistors
won't matter. The advantage here is the same firmware can then auto-detect
different hardware versions. In future versions or reworked versions of
this board the input lines would be pulled differently.
This topic comes up occasionally, so from experience I expect you'll hear
voodoo electronics about how lines should always have resistors on them, not
driven by the micro in case of short, etc. All these come down to
protecting the hardware from firmware bugs. That may make sense for the
development prototype, but is just silly superstition for volume production.
> From: Olin Lathrop <.....olin_piclistKILLspam.....embedinc.com>
> Subject: Re: [PIC] Unused port lines question
> To: "Microcontroller discussion list - Public." <EraseMEpiclistspam_OUTTakeThisOuTmit.edu>
> Date: Sunday, September 13, 2009, 4:34 PM
> Roger Weichert wrote:
> > I'm building a project (using a PIC16F627) and have
> reached the stage
> > where I'm ready to design the board layout, so need to
> decide what
> > I'm going to do with the unused port lines. ...
> >
> > I'm aiming for a tiny pcb with no wasted space, so am
> trying to avoid
> > any components I don't absolutely need to have.
>
> In that case make them outputs and make sure the firmware
> drives them. If
> any are input-only pins, then tie them to Vss or Vdd
> directly or thru a
> resistor.
>
> If you have a little extra space, bring the unused pins out
> to pads and
> still make sure they are driven. If you have a little
> more extra space you
> can add 100Kohm pullup or pulldown resistor pads. You
> don't actually have
> to load the parts, so it's only a board area issue.
> Any changes to make use
> of the pins would require rework and firmware change
> anyway. The parts can
> be added to the pads and the firmware changed at that time
> to make pins
> inputs if that's what's needed.
>
> If you were able to fit the resistor pads then the next
> step up is to
> install at least a few of the resistors and make the pins
> inputs. Unless
> this is a very high volume project, a few cents for the
> extra resistors
> won't matter. The advantage here is the same firmware
> can then auto-detect
> different hardware versions. In future versions or
> reworked versions of
> this board the input lines would be pulled differently.
>
> This topic comes up occasionally, so from experience I
> expect you'll hear
> voodoo electronics about how lines should always have
> resistors on them, not
> driven by the micro in case of short, etc. All these
> come down to
> protecting the hardware from firmware bugs. That may
> make sense for the
> development prototype, but is just silly superstition for
> volume production.
>
>
> ********************************************************************
> Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
> (978) 742-9014. Gold level PIC consultants since
> 2000.
John Chung wrote:
> I do understand the importance of not floating the pins but I am
> not sure of your suggestion:
>
>> If you have a little extra space, bring the unused pins out
>> to pads and
>> still make sure they are driven.
>
> Would the pin be tied to ground? If it's just a pad by itself I am not
> sure why would this be any advantage. Any suggestion?
The pad is for later when you wish you had another line hooked up to some
thingy. It's a lot easier to rework the board if a pad is there. It can
also be useful as a test point during continued development and debugging.
You could hook a LED up to it, or watch it on a scope. Holding a scope
probe on a PIC pin can be tricky, but with a thru hole pad, you just stick
the tip of the probe thru the hole.
> Roger Weichert wrote:
>> I'm building a project (using a PIC16F627) and have reached the stage
>> where I'm ready to design the board layout, so need to decide what
>> I'm going to do with the unused port lines. ...
>>
>> I'm aiming for a tiny pcb with no wasted space, so am trying to avoid
>> any components I don't absolutely need to have.
>
> In that case make them outputs and make sure the firmware drives them.
So if I make sure (right at the start of my program) to put in the code
to set the unused pins as outputs, there should be no conflict in leaving
them
open and I should be able to forget about them from there.
> If any are input-only pins, then tie them to Vss or Vdd directly or thru a
> resistor.
and same as above? As long as I address their function by setting them up in
my initialisation,
they can happily survive regular resetting and restarting for whatever
reason ...
without potentially causing problems down the track.
> If you have a little extra space, bring the unused pins out to pads and
> still make sure they are driven. If you have a little more extra space
> you
> can add 100Kohm pullup or pulldown resistor pads. You don't actually have
> to load the parts, so it's only a board area issue. Any changes to make
> use
> of the pins would require rework and firmware change anyway. The parts
> can
> be added to the pads and the firmware changed at that time to make pins
> inputs if that's what's needed.
OK, I can understand the sense in doing this ... in fact I probably can
afford to
take the space on this board to do it. (probably cant afford NOT to)
I have one thing to add. RA4 is open collector when used as an output. I
am sure one could debate about what is acceptable. But I usually tie it
to ground if I'm not using it and don't want to waste a resistor on it.
> From: "Olin Lathrop" <@spam@olin_piclistKILLspamembedinc.com>
>
>> Roger Weichert wrote:
>>> I'm building a project (using a PIC16F627) and have reached the stage
>>> where I'm ready to design the board layout, so need to decide what
>>> I'm going to do with the unused port lines. ...
>>>
>>> I'm aiming for a tiny pcb with no wasted space, so am trying to avoid
>>> any components I don't absolutely need to have.
>> In that case make them outputs and make sure the firmware drives them.
>
>
> So if I make sure (right at the start of my program) to put in the code
> to set the unused pins as outputs, there should be no conflict in leaving
> them
> open and I should be able to forget about them from there.
>
>
>> If any are input-only pins, then tie them to Vss or Vdd directly or thru a
>> resistor.
>
>
> and same as above? As long as I address their function by setting them up in
> my initialisation,
> they can happily survive regular resetting and restarting for whatever
> reason ...
> without potentially causing problems down the track.
>
>
>> If you have a little extra space, bring the unused pins out to pads and
>> still make sure they are driven. If you have a little more extra space
>> you
>> can add 100Kohm pullup or pulldown resistor pads. You don't actually have
>> to load the parts, so it's only a board area issue. Any changes to make
>> use
>> of the pins would require rework and firmware change anyway. The parts
>> can
>> be added to the pads and the firmware changed at that time to make pins
>> inputs if that's what's needed.
>
>
> OK, I can understand the sense in doing this ... in fact I probably can
> afford to
> take the space on this board to do it. (probably cant afford NOT to)
>
> Thanks for your input Olin, it has helped a lot.
>
> Regards, Roger
>
>
>
Some PICs have the high-side FET on RA4 so it can drive high
rather than "float" high. It caught me out when a circuit based on
an F88 (has the FET) was ported to an F628 (doesn't). So you
have to watch the RA4 voltage too. Open collector RA4 spec is
a lot higher, 14V vs Vdd + 0.3V
> > Open collector RA4 spec is a lot higher, 14V vs Vdd + 0.3V
>
> Only on the really old PICs. It's something like 7 volts on anything
> current or was-recently-current.
For most PICs I use (including 18F) it's 8.5V, which is > Vdd max.
In the vast majority of applications only 5V is used but I have seen
RA4 used as a high-voltage pin (sensing and logic conversion for
example). That specs have been reduced could cause problems
with legacy products which can cope with 14V. But, as always,
check the datasheet. Things change and what might be a functional
drop-in might not be so wrt electrical specs
Roger Weichert wrote:
> So if I make sure (right at the start of my program) to put in the
> code
> to set the unused pins as outputs, there should be no conflict in
> leaving them
> open and I should be able to forget about them from there.
Yes. In fact, that is what my PIC development environment does
automatically. You specify all the pins you use with /INBIT and /OUTBIT
preprocessor directives, and the remaining pins are automatically set to
outputs and driven low in the standard PORT module.
>> If any are input-only pins, then tie them to Vss or Vdd directly or
>> thru a resistor.
>
> and same as above? As long as I address their function by setting
> them up in my initialisation,
> they can happily survive regular resetting and restarting for whatever
> reason ...
> without potentially causing problems down the track.
I'm not sure what you are asking about "regular resetting", but note that
high impedence is the power state of all pins that can be high impedence.
Just don't ever tied unused Vdd pins to ground ;-)
Bob Blick wrote:
> I have one thing to add. RA4 is open collector when used as an
> output. I
> am sure one could debate about what is acceptable. But I usually tie
> it
> to ground if I'm not using it and don't want to waste a resistor on
> it.
Setting it as a output and driving it low works for this pin too if unused.
That way the single output transistor will be on and the pin won't float.
This is really no different than a totem pole output because the missing
high side driver would be off anyway.
>I have one thing to add. RA4 is open collector when used as an output. I
> am sure one could debate about what is acceptable. But I usually tie it
> to ground if I'm not using it and don't want to waste a resistor on it.
>
> Cheerful regards,
>
> Bob
> > Just don't ever tied unused Vdd pins to ground ;-)
> >
> LOL! What's wrong with that? :-)
At the weekend I re-wired a breadboard from a 12F675 to an
F88 but didn't remember to swap the analyser ground from the
right to the left of the breadboard. For a good 15 minutes I couldn't
figure out why the analyser wasn't working. Then smelled "heat".
The analyser's flimsy E-Z Hook ground was connected to the
5V rail. Its wire was hot and the tip all melty and she not so good
looking no more. Luckily it was the PSU taking the strain, not the
USB
Vicent Colomar Prats wrote:
> Anyway, it is always better to use protecting resistors connected in
> serial to all unused pins.
Here we go with the voodoo electronics, and a absolute statement for bonus.
So using a "protecting" (whatever that really means) resistor is *always*
better than any other alternative?
I'll have to go back and tell my customer with the high volume
cost-sensitive electronic badge to stop making them immediately because the
design is incorrect. I'll need to make the PC board a little bigger to fit
all the resistors for the unused PIC pins (we used a 44 pin QFN package even
though a 28 pin PIC was sufficient because the 44 QFN package was smaller
than any 28 pin package for that PIC type), the board will be a bit more
expensive because I'll need another layer or two to route the lines out from
the PIC, and they'll have to make the case bigger too. I'm sure the
customers will be understanding when the badge is bigger, heavier, and costs
more. After all, it *always* has to be done this way. It's not clear
though what exactly these resistor are supposed to be in series with. A
small piece of dead fish, I presume? Can you get those in 0402 or 0603
packages?
> At the weekend I re-wired a breadboard from a 12F675 to an
> F88 but didn't remember to swap the analyser ground from the
> right to the left of the breadboard. For a good 15 minutes I couldn't
> figure out why the analyser wasn't working. Then smelled "heat".
> The analyser's flimsy E-Z Hook ground was connected to the
> 5V rail. Its wire was hot and the tip all melty and she not so good
> looking no more. Luckily it was the PSU taking the strain, not the
> USB
>
Oh dear! Maybe it is cheaper to turn a bit down that current limiter on your
PSU :-)
>After all, it *always* has to be done this way. It's not clear
>though what exactly these resistor are supposed to be in series with.
If you go back to TTL (pre LSTTL) abs max supply voltage is 7v, abs max
input voltage is 5.5v (at least that's what the data sheets say). You would
pull inputs high through a current limiting resistor to provide more safety
margin for excess supply voltage. No resistor needed to pull them low.
Another reason would be for automatic test allowing you to force unused
inputs, the argument being you can't test a chip is good without testing
all of it including the inputs and bits not actually used in the circuit.
Popular with test departments because regardless of what circuit it was in
they could just tell the ATE that is a 7432 - test it.
Olin, you can speak about voodoo, ufos, politics or whatever you want... but
that has nothing to do with this list. It's all about PIC nothing else.
I do not understand why you ask me what I mean. You know it perfectly, you
are not a 20 years old man, but sometimes you seem as of your responses. You
are not in your first day working so, you know perfectly what I mean.
It's "always" better to have a protection on each non-used pin than having
nothing? True.
Is it mandatory? Not. If you are experienced, mistakes like setting a pin
connected to Vcc as an output, will probably not to occur. Not even updating
a customer's board firmware.
Are the resistors the only way to protect from a firmware mistake? No, there
are many others.
No need to be sarcastic about your customers boards. I never spoke about
them nor told you to change them. Have you ever used a dead fish in
electronics??? What sense has that to be with the post and the question from
the original PO?
If you consider my response is not clear enought, tell me. I will clarify.
Respond with arguments silly, sarcastic and arrogant, just wasting our time,
me and the listers.
Sorry about that, Roger Weichert and all PIC-listers.
> Here we go with the voodoo electronics, and a absolute statement for bonus.
> So using a "protecting" (whatever that really means) resistor is *always*
> better than any other alternative?
>
> I'll have to go back and tell my customer with the high volume
> cost-sensitive electronic badge to stop making them immediately because the
> design is incorrect. I'll need to make the PC board a little bigger to fit
> all the resistors for the unused PIC pins (we used a 44 pin QFN package
> even
> though a 28 pin PIC was sufficient because the 44 QFN package was smaller
> than any 28 pin package for that PIC type), the board will be a bit more
> expensive because I'll need another layer or two to route the lines out
> from
> the PIC, and they'll have to make the case bigger too. I'm sure the
> customers will be understanding when the badge is bigger, heavier, and
> costs
> more. After all, it *always* has to be done this way. It's not clear
> though what exactly these resistor are supposed to be in series with. A
> small piece of dead fish, I presume? Can you get those in 0402 or 0603
> packages?
>
>
> ********************************************************************
> Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
> (978) 742-9014. Gold level PIC consultants since 2000.
Vicent Colomar Prats wrote:
> It's "always" better to have a protection on each non-used pin than
> having nothing? True.
It's absolute statements like this that are voodoo electronics. They are
certainly not real engineering. In real engineering there is most often a
tradeoff, as in this case.
Protection on pins can sometimes be the right tradeoff. But saying it is
*always* the right tradeoff is just silly since that ignores the costs. In
this case the costs include components, board space for the components, and
board space for routing the lines. These costs have to be weighed against
the benefits, which are miniscule at best compared to not even routing
traces to the pins. Adding the tracks and extra parts may actually make
things worse due to high susceptibility to static discharge or RF pickup.
So if you want absolute certainty, go wave a dead fish over your project.
That always works, although you have to use the right species of fish during
the right phase of the monne and your thoughts have to be pure. Meanwhile
for the rest of us there are real engineering tradeoffs to consider.
> No need to be sarcastic about your customers boards. I never spoke about
> them nor told you to change them.
Actually you did since they fall into your "always" case.
> It's "always" better to have a protection on each non-used pin than having
> nothing? True.
If you still really think it is always better to have protection on each
non-used pin you have a very narrow interpretation of "better". Perhaps
you should read Olin's reply again. and again.
> Sorry about that, Roger Weichert and all PIC-listers.
I do not recall giving you permission to speak for me, so please don't.
Wouter van Ooijen
(always speaking (only) for my self)
-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu
>
> It's "always" better to have a protection on each non-used pin than having
> nothing? True.
Well, that's a new one on me. It seems like overkill based on the way the
PIC port works. Once in a while, I GND a PIC pin to keep out noise. Is
that what you meant here?
Olin, I never spoke about price, size, etc. I just refered to protection. So
maybe, I see now, my "always" could create a confussion to you, or others.
So I asume I missused the word "always". I referred to electrical protection
not pice, size, etc.
You prefer to leave them unconnected, stacked to Vcc or to Gnd, is all up to
you. I prefer to not. But, of course, I do not complain about more size and
the extra components. Usually you (me, somebody) would chosse a
microcontroller that best fits the project, not only peripherals, speed,
memory, also pin count is taken in consideration. So, the number of
unconnected pins is not going to be very large, and so tradeoff. But that's
my opinion with my very personal conditions, and yours should be different.
You bring again the death fish and mix with real engineering... May be I do
not know sufficient english to really understand it, but it seems to me
offensive, correct me if it's not. So I'm not going to respond that way.
> Vicent Colomar Prats wrote:
> > It's "always" better to have a protection on each non-used pin than
> > having nothing? True.
>
> It's absolute statements like this that are voodoo electronics. They are
> certainly not real engineering. In real engineering there is most often a
> tradeoff, as in this case.
>
> Protection on pins can sometimes be the right tradeoff. But saying it is
> *always* the right tradeoff is just silly since that ignores the costs. In
> this case the costs include components, board space for the components, and
> board space for routing the lines. These costs have to be weighed against
> the benefits, which are miniscule at best compared to not even routing
> traces to the pins. Adding the tracks and extra parts may actually make
> things worse due to high susceptibility to static discharge or RF pickup.
>
> So if you want absolute certainty, go wave a dead fish over your project.
> That always works, although you have to use the right species of fish
> during
> the right phase of the monne and your thoughts have to be pure. Meanwhile
> for the rest of us there are real engineering tradeoffs to consider.
>
> > No need to be sarcastic about your customers boards. I never spoke about
> > them nor told you to change them.
>
> Actually you did since they fall into your "always" case.
>
>
> ********************************************************************
> Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
> (978) 742-9014. Gold level PIC consultants since 2000.
> >
> > It's "always" better to have a protection on each non-used pin than
> having
> > nothing? True.
>
> Well, that's a new one on me. It seems like overkill based on the way the
> PIC port works. Once in a while, I GND a PIC pin to keep out noise. Is
> that what you meant here?
>
> --Bob
>
>
>
> You bring again the death fish and mix with real engineering... May be I do
> not know sufficient english to really understand it, but it seems to me
> offensive, correct me if it's not. So I'm not going to respond that way.
It is not offensive, at least not more offensive than stating that you
are wrong (which might be offensive but IMHO correct, because of the way
you used the word 'always').
Waving a dead fish (at a certain phase of the moon) is piclist speak for
any measures that have absolutely nothing to do with the problem at
hand. Or to state it in one word, for sillyness.
--
Wouter van Ooijen
-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu
> What I mean is to do it thru a resistor, not directly. That can save your
> pic from an unfortunate firmware mistake.
We all agree that such resistors serve a purpose. The question is
whether that purpose is balanced by the cost (cost in a very broad
sense). Anyone who gives a simple cover-all-cases 'yes' or 'no' answer
to such a question is IMHO not an engineer. The good answer(s) are about
the cases in which the answer should be yes, the cases in which the
answer should be no, and in particular about the interesting gray land
in between.
--
Wouter van Ooijen
-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu
As I said my english write and understanding is not fantastic. The way Olin
spoke seemed to me offensive, but I can mistake, and I hope so.
For the "allways" word, I already said, I could made I mistake not
precissing it was only applied to electric protection, not about size, not
about price, not ... But I maintain that I prefer to protec unused pins thru
limiting resistors against leaving them unconnected or tied to Vcc/Gnd. I do
not remember the PO has specified he had size or price restrictions. A few
resistors and a very small extra space not allways is going to limit your
design, and can be (once in a time or never) usefull. That's my opinion and
tried to express that way to Roger. Sorry to everybody wich I make
confussion, but tried to speak allways correct without discualifying nobody,
so I expect the same to me.
Well, no more discussion from my part, I'm here to learn and try to help
when I can (not so often), and not to loose my time or make all other loose
them.
Vicent Colomar Prats wrote:
> But I maintain that I prefer to protec
> unused pins thru limiting resistors against leaving them unconnected
> or tied to Vcc/Gnd.
This makes no sense. Protect against what? What are the resistors supposed
to limit, considering these are unconnected pins in the first place? And
it's not clear what is supposed to be on the other end of those resistors
from the PIC pin. If you are leaving them open, then all you've done is add
antennas to the unused pins, with a little extra capacitance to unknown
parts of the circuit.
There are some legitimate reasons for pullup or pulldown resistors on unused
pins as I also mentioned in my first post on this thread, but "protection"
and "limiting" aren't included in these reasons. Setting unused pins to
outputs and driving them low essentially connects them to ground via a few
10s of ohms. Given that, a otherwise unconnected PIC pin is going to be
about as safe as it gets. If something can get in there to hurt it, the
rest of the circuit is in serious trouble.
Olin, do you really ask me that? I do not think so. I think you are
experienced engineer so why to make a question you know the answer, just to
create more confussion? I think this only ads more offtopic to this post.
> What are the resistors supposed
> to limit, considering these are unconnected pins in the first place? And
> it's not clear what is supposed to be on the other end of those resistors
> from the PIC pin.
>
I think Olin makes a valid point. If you have unused pins, set them as
outputs and leave them unconnected. You were suggesting that it is
necessary to tie all unused pins high or low through resistors, which is
extremely defensive treatment, arguably too defensive. But if you have
an application where setting unused pins as outputs is not enough,
describe it.
> Olin, do you really ask me that? I do not think so. I think you are
> experienced engineer so why to make a question you know the answer, just to
> create more confussion? I think this only ads more offtopic to this post.
>
> 2009/9/18 Olin Lathrop <RemoveMEolin_piclistTakeThisOuTspamembedinc.com>
>
>> What are the resistors supposed
>> to limit, considering these are unconnected pins in the first place? And
>> it's not clear what is supposed to be on the other end of those resistors
>> from the PIC pin.
>>
Vicent Colomar Prats wrote:
> Olin, do you really ask me that? I do not think so. I think you are
> experienced engineer so why to make a question you know the answer,
> just to create more confussion? I think this only ads more offtopic
> to this post.
I was trying to ask questions that would provoke a little thought on your
end and reveal to yourself the silliness of your statements. Apparently
that didn't work, so I guess I have to go thru this in little baby steps.
Your statement was:
> But I maintain that I prefer to protec
> unused pins thru limiting resistors against leaving them unconnected
> or tied to Vcc/Gnd.
So I asked:
> What are the resistors supposed
> to limit, considering these are unconnected pins in the first place?
Remember that we were talking about unused pins. The discussion is in part
about what, if anything, they should be connected to since the circuit does
not have them connected to anything. If they are left unconnected then
current can't flow to them. Can't you see that limiting current in this
context makes no sense? There already is no current to limit. And that is
assuming you meant current should be limited and not some other value. Note
that your statement never specified what these resistors are supposed to
limit. Normally one would assume current from context, but since that makes
no sense in this situation it makes your statement ambiguous.
Then I said:
> And it's not clear what is supposed to be on the other end of those
> resistors from the PIC pin.
This was more to point out how poorly written your statement was. The best
result would have been that you would have been embarassed that you made
such a ambiguous statement and that someone pointed it out in front of 2000
people, and you'd hopefully be more careful in composing future statements.
Trying to defend it only makes you look worse.
Explanation: Look at a resistor carefully and you will see it has two leads,
not just one. That means when you describe how a resistor is connected, you
have to specify what both ends are connected to. Your statement only
specified one end be connected to the PIC pin. Not describing what the
other end is supposed to be connected to is poor communication and bad
engineering.
I and others here keep beating on this, but unfortunately it still refuses
to sink in too often. Engineering is about intelligent tradeoffs and
requires attention to detail. This includes not only the design itself, but
the documentation and any other descriptions of that design, such as
discussions about it here. A engineer that designs something but doesn't
document or communicate that design is no engineer at all. As Dave Tweed
pointed out recently, you can view the end result of engineering as a pile
of documentation. That's really all the schematic, bill of materials,
assembly drawings, parts locator, special build instructions, theory of
operation, test procedure, etc, are. These need to be taken just as
seriously as producing the design in the first place, since without them the
design has no value.
And all this documentation needs to be clear and precise. There is no place
for sloppiness. This is especially true the wider and more diverse the
audience is, since there will be less common knowledge and communication
conventions you can assume. The PIClist is such a place. Inexact and
sloppy descriptions are especially bad engineering here. There are also a
significant numbers of newbies here that might not be able to tell bad
engineering from good. So yes, I and some others are going to continue to
hammer on sloppiness presented here. Get used to it, or better yet, be more
careful so that there is no need for such comments.
Since my first "unfortunate" response in this topic 18 (engineering
speaking: non productive) more responses happened, none of them from the OP
(Roger Weichert). And it seems it's easy to continue that way. Not my way.
In the official web page of this list there is no a recomanation of the
grade of concretion/assumtion we have to write in the posts.
Roger Weichert, if you still have any interest on what to do with unused
pins, here (in our piclist page) you have a very explanatory doc:
www.piclist.com/techref/logic/xtrapins.htm
And, Roger, sorry again, to make you loose your time with silly discussions.
> Since my first "unfortunate" response in this topic 18 (engineering
> speaking: non productive) more responses happened, none of them from the
> OP
> (Roger Weichert). And it seems it's easy to continue that way. Not my way.
> In the official web page of this list there is no a recomanation of the
> grade of concretion/assumtion we have to write in the posts.
>
> Roger Weichert, if you still have any interest on what to do with unused
> pins, here (in our piclist page) you have a very explanatory doc:
> www.piclist.com/techref/logic/xtrapins.htm
> And, Roger, sorry again, to make you loose your time with silly
> discussions.
clear.net.nz>
