Maker Pro
Maker Pro

leaving pin floating

J

James Beck

Follow the thread back up, its about whether programming unused PIC
pins as outputs is a problem because of the time between power up and
when the pins actually get set to output.

I would never recommend leaving inputs floating.

Luhan
Well, it IS a floating input until the direction register gets set,
therfore it is a floating input. No matter what you see on hobby sites
or reference designs. If the MCU gets stuck in reset or keeps
resetting, those pins can spend a lot of time as inputs. Maybe I'm just
being anal, but that's the way I have always done it.

Jim
 
J

James Beck

Take for instance pins used as an address/data bus for an external
Flash/RAM memory. If the microcontroller starts with all pins defined
as generic inputs, and relies on firmware to define these pins as
outputs, there will be a finite time in which these pins can float.

For dedicated output-only pins, there is always a well defined level,
of course, but many microcontrollers have multiple functions per pin,
configurable by software, and they may default to input.
Yes, I guess we are talking about the same thing just using different
terms. When you say output you mean a programmable pin that is set to
an output. Naturally you will need to handle that time the pin is an
input and in fact I am discussing that with Luhan in another thread.

Jim
 
A

A. Lemon

A. Lemon said:
Hi!

Is it so bad to leave an input pin floating (from a PIC)?

Is the increased current consumption from input noise at threshold point
the only drawback?

Thanks!
Al

Holly Smoke, this subject is hot! :))

But I should've been more accurate in my request. Sorry about that.

What if the PIC (16F628) is kept in reset state for a long time, say or
hours or days and the floating pins are PORTA pins. These pins have Schmitt
trigger inputs. A middle point threshold should not promote oscillation or
current increase, is'nt it... or is it?

Al
 
J

James Beck

Holly Smoke, this subject is hot! :))

But I should've been more accurate in my request. Sorry about that.

What if the PIC (16F628) is kept in reset state for a long time, say or
hours or days and the floating pins are PORTA pins. These pins have Schmitt
trigger inputs. A middle point threshold should not promote oscillation or
current increase, is'nt it... or is it?

Al
I would still have a problem with keeping any high impedance input just
hanging there, just from an ESD suseptability stand point, it is a bad
idea. I have this thing about the electrons going where I want them to
when I want them to. I guess I'm just old school and those habits stuck
with me and have served me well.

Jim
 
A

A. Lemon

James Beck said:
I would still have a problem with keeping any high impedance input just
hanging there, just from an ESD suseptability stand point, it is a bad
idea. I have this thing about the electrons going where I want them to
when I want them to. I guess I'm just old school and those habits stuck
with me and have served me well.

Jim

Mmmmh!
Floating pins seems to make most people unconfortable. For sure, adding
resistors is cheap. I now consider adding few 4x100K SIP resistors soldered
directly to the PIC under the PCB, the common leg just falling nicely on a
Vss pin (or Vcc pin) and the other legs to I/O pins.

For ESD, these "floatting" pins are not used and no PCB trace go to them.
Al
 
J

James Beck

Mmmmh!
Floating pins seems to make most people unconfortable. For sure, adding
resistors is cheap. I now consider adding few 4x100K SIP resistors soldered
directly to the PIC under the PCB, the common leg just falling nicely on a
Vss pin (or Vcc pin) and the other legs to I/O pins.

For ESD, these "floatting" pins are not used and no PCB trace go to them.
Al
That's exactly what I do for a proto. The little 10 pin pull up packs
can be stuck under the IC.
Trust me about ESD. It doesn't matter that the pin has no copper
touching it. There is enouch there just counting the pin and the
bonding wire inside. Some of our stuff has to meet some pretty big
discharge tests and every little thing hanging out there can be a
problem. Some of those high V discharges don't care about insulation
and "isolation". I'll bet you never had much of a problem shocking all
kinds of "insulated" things using the 'ol rub your feet across the
carpet in the winter.

Jim
 
J

John Devereux

James Beck said:
That's exactly what I do for a proto. The little 10 pin pull up packs
can be stuck under the IC.
Trust me about ESD. It doesn't matter that the pin has no copper
touching it. There is enouch there just counting the pin and the
bonding wire inside. Some of our stuff has to meet some pretty big
discharge tests and every little thing hanging out there can be a
problem. Some of those high V discharges don't care about insulation
and "isolation". I'll bet you never had much of a problem shocking all
kinds of "insulated" things using the 'ol rub your feet across the
carpet in the winter.

But what does ESD have to do with a pin having a pullup, or leaving
floating?
 
J

James Beck

But what does ESD have to do with a pin having a pullup, or leaving
floating?
Gives the juice somwhere to go other than the guts of the IC.
If eveything is at 10KV it is the same as having nothing at 10KV.
Sure there are other things to consider when making industrial strength
ESD protection, but not having any loose pins is a great first step.

Jim
 
J

John Devereux

James Beck said:
Gives the juice somwhere to go other than the guts of the IC.

But I don't think this is true. It will go through the IC *before* it
goes through a 10k (or 1k) pullup. Through the input protection
diodes, for a start.

I don't see what benefit it has at all for ESD, in fact I would expect
it to make things worse. Before, you just had a tiny exposed pin. Now
you have the pin, and the tracking, and the resistor body to worry
about. On the other hand, connecting the unused pin direct to the
groundplane would work.

This is what I usually do, for production boards.
If eveything is at 10KV it is the same as having nothing at 10KV.
Sure there are other things to consider when making industrial strength
ESD protection, but not having any loose pins is a great first step.

In my experience you need a combination of parallel transient
suppression *and* series current limiting, for external signals.

________
I/P o-------[ R ]--------|CPU
| |
- TVS
^
|
======================== Groundplane
 
J

James Beck

James Beck said:
Gives the juice somwhere to go other than the guts of the IC.

But I don't think this is true. It will go through the IC *before* it
goes through a 10k (or 1k) pullup. Through the input protection
diodes, for a start.

I don't see what benefit it has at all for ESD, in fact I would expect
it to make things worse. Before, you just had a tiny exposed pin. Now
you have the pin, and the tracking, and the resistor body to worry
about. On the other hand, connecting the unused pin direct to the
groundplane would work.

This is what I usually do, for production boards.
If eveything is at 10KV it is the same as having nothing at 10KV.
Sure there are other things to consider when making industrial strength
ESD protection, but not having any loose pins is a great first step.

In my experience you need a combination of parallel transient
suppression *and* series current limiting, for external signals.

________
I/P o-------[ R ]--------|CPU
| |
- TVS
^
|
======================== Groundplane
So you think having a highZ input just floating there is better than
pulling it somewhere? Fine, do it your way. I agree grounding the line
is best, (depending on how "good" the ground is) but I don't like
programmable pins connected directly to power or ground. Like I said,
there is more to good ESD supression than what I went over, I just
didn't want to take the thread any further off topic.

Jim
 
J

John Devereux

James Beck said:
James Beck said:
But what does ESD have to do with a pin having a pullup, or leaving
floating?


Gives the juice somwhere to go other than the guts of the IC.

But I don't think this is true. It will go through the IC *before* it
goes through a 10k (or 1k) pullup. Through the input protection
diodes, for a start.

I don't see what benefit it has at all for ESD, in fact I would expect
it to make things worse. Before, you just had a tiny exposed pin. Now
you have the pin, and the tracking, and the resistor body to worry
about. On the other hand, connecting the unused pin direct to the
groundplane would work.

This is what I usually do, for production boards.
If eveything is at 10KV it is the same as having nothing at 10KV.
Sure there are other things to consider when making industrial strength
ESD protection, but not having any loose pins is a great first step.

In my experience you need a combination of parallel transient
suppression *and* series current limiting, for external signals.

________
I/P o-------[ R ]--------|CPU
| |
- TVS
^
|
======================== Groundplane
So you think having a highZ input just floating there is better than
pulling it somewhere? Fine, do it your way.

No, not at all, just that I could not see the difference *from the
point of view of ESD*. The main problem I have *actually* found, when
leaving floating, is that the pin may turn out to have some
configuration function that is read at power-on! Other hazards would
be an interrupt left enabled, connected to a floating input.

The point about increased power when floating is interesting, but I
think that switching the pin to output mode should eliminate this. The
brief window after reset would not cause a problem in my applications,
but perhaps in very low power work it might become important.
I agree grounding the line is best, (depending on how "good" the
ground is) but I don't like programmable pins connected directly to
power or ground.

Yes, good point, I usually only do this for inputs (e.g. unused ADC
inputs).
 
Top