Searching \ for 'isolating two pics...' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/microchip/devices.htm?key=pic
Search entire site for: 'isolating two pics...'.

Truncated match.
PICList Thread
'isolating two pics...'
1998\09\26@194446 by John Haggins

picon face
I hope somebody can help me with this...

I have a 16F84 programmed and working well on a breadboard which uses a
wire connected to one of the PIC pins as a touch sensor.

If I take an identically programmed pic and circuit, duplicating everything
except the shared power supply; I get a problem. Instead of each PIC
responding independently to touch on - they both respond to the same single
touch on either PIC.

The only way they can be "ensing each other is through the power line
and/or ground since there are no connections between them.

Question: Is there a simple way to totally separate/isolate one from the
other so that each will respond independently ? Is this possible sharing a
power supply?

Thanks In Advance,
John

1998\09\27@012159 by Dave VanHorn

flavicon
face
>If I take an identically programmed pic and circuit, duplicating
everything
>except the shared power supply; I get a problem. Instead of each PIC
>responding independently to touch on - they both respond to the same
single
>touch on either PIC.


Sounds like your touch circuit needs work.
Is V- actually earth grounded?
Without knowing which way you are detecting "touch", it's hard to say.
It looks like the touch on plate A is bouncing the ground, such that
it looks like another touch occurs on plate B.

1998\09\27@015147 by John Haggins

picon face
My "touch circuit" is a piece of grounded wire :)). The pic somehow detects
the touch (I programmed it in PicBasicPro with a simple rctime command that
measures a change) which works fine - except that I cant independently
reproduce it in another PIC at the same time. So I guess my question is -
is there a way to give each pic its own ground? I'm using a standard wall
wart 9 volt adapter with no real earth pin.

Alternatively, if anyone knows how to detect touch on a wire connected to a
specific pin in a more stable way... I'd be pleased to hear about it :))

thanks...

{Quote hidden}

1998\09\27@023838 by g.daniel.invent.design

flavicon
face
Check the SGSthomson application notes for a good demonstration of touch
sensor hardware.   Best to double check with an experienced electrician
and use resistors, capacitors with large over voltage rating if you do
this way though.

Your phrase "The pic somehow detects the touch" is a bit scary,
certainly knoledge at this level is somewhat distant from a succesful
commercial application.

Remember that the PICpins have limited static discharge protection, if
for example your finished circuit was installed in a a cold dry location
and someone walked down a long carpeted hallway.. zzzzAP goes your PIC.

You can "touch" sense in several ways:
1) resistance of finger etc pulls up/down PIC pin with very high
resistance to the other PIC supply voltage.
2) close proximity of finger etc loads capacitive sense circuit.
3) body coupled capacitivly/directly to "Earth" potential is detected
as  50/60 Hz signal referenced to active/phase wire through a VERY HIGH
RESISTANCE ie 15 mA of current can KILL!

Methods 1 & 2 are safe, method 3 should be implemented only with utmost
caution.

hope this helps,
Graham Daniel,
Electronic Product Enhancements.

John Haggins wrote:
{Quote hidden}

1998\09\27@092407 by Stig Brautaset

flavicon
face
>My "touch circuit" is a piece of grounded wire :)). The pic somehow detects
>the touch (I programmed it in PicBasicPro with a simple rctime command that
>measures a change) which works fine - except that I cant independently
>reproduce it in another PIC at the same time. So I guess my question is -
>is there a way to give each pic its own ground? I'm using a standard wall
>wart 9 volt adapter with no real earth pin.


I have humble answer to your problem, but I am afraid it contains two
transistors and one resistor for each detector.

I'm tryly sorry for the attachment, I know many of the PICsters don't aprove
of it, but sending the picture of the circuit is the only way I see to give
away this information.

Hope it is of any help, and PLEASE give me feedback if I have completely
screwed this one up, misunderstood something and so. :-) Please be
indulgent, I am only a student yet... :-))

Stig

===================================================
Stig Brautaset  -  spam_OUTsbrautasTakeThisOuTspamc2i.net  -  http://home.c2i.net/sbrautas
===================================================

Drive carefully; 90% of the people in the world are caused by accidents.

1998\09\27@094517 by Stig Brautaset

flavicon
face
part 0 996 bytes
>My "touch circuit" is a piece of grounded wire :)). The pic somehow detects
>the touch (I programmed it in PicBasicPro with a simple rctime command that
>measures a change) which works fine - except that I cant independently
>reproduce it in another PIC at the same time. So I guess my question is -
>is there a way to give each pic its own ground? I'm using a standard wall
>wart 9 volt adapter with no real earth pin.

I have humble answer to your problem, but I am afraid it contains two
transistors and one resistor for each detector.

I'm tryly sorry for the attachment, I know many of the PICsters don't aprove
of it, but sending the picture of the circuit is the only way I see to give
away this information.

Hope it is of any help, and PLEASE give me feedback if I have completely
screwed this one up, misunderstood something and so. :-) Please be
indulgent, I am only a student yet... :-))

Attachment converted: wonderland:touch.JPG (JPEG/JVWR) (00018230)

1998\09\27@122404 by John Haggins

picon face
Thanks - I'll give it a try. If you can recommend any particular
transistors - that will be helpful, since I'm without catalogs at the
moment. Also, you marked the drawing with 9v supply - but I assume you
realise it's 5 volts after going through a regulator.

Thanks again.

{Quote hidden}

1998\09\27@125953 by Dave VanHorn

flavicon
face
>Hmmm... all this excuse of the attached picture and I forgot to
attach it...
>Hope it works. I have not tested it.
>

This circuit needs a pullup resistor on it's input.  There's a TON of
gain here, and any leakage at all, static charge, you name it, and
it's going to turn on.  Maybe 4.7 meg or 10 meg, but it needs
something on that input to assure that it stays off.

1998\09\27@180121 by paulb

flavicon
face
part 0 1242 bytes
> Hmmm... all this excuse of the attached picture and I forgot to attach
> it...

 Fear not.  Although some have an objection to *any* attachment, yours
is a "list-friendly" one of reasonable size although I have taken the
liberty of re-encoding it in a format more suitable.  What *really* gets
up everyone's noses is the sending of uncompressed .BMP files.

 As to the circuit itself, I fear it has the limitations others have
mentioned; it is probably *too* sensitive to its own leakeages.  Since
the PIC has an extremely high impedance already, the only real advantage
of adding external components would be to isolate it from discharges
(for which a resistor and "catch" diodes to the supply rails should
really be sufficient) or increase its sensitivity to voltage.

 I am working on touch switches myself.  I feel the best approach is to
assess capacitance using a variant on the "Stamp" circuit for measuring
same: http://www.dontronics.com/psbpix/pot1.gif  Consider the capacitor
variable rather than the resistor.
Theory at http://www.dontronics.com/see.html#pot but I'd do it a little
different.
--
 Cheers,
       Paul B.

Attachment converted: wonderland:touch.gif (GIFf/JVWR) (0001831D)

More... (looser matching)
- Last day of these posts
- In 1998 , 1999 only
- Today
- New search...