Shop OBEX P1 Docs P2 Docs Learn Events
Pin Protection — Parallax Forums

Pin Protection

ArchiverArchiver Posts: 46,084
edited 2004-06-11 04:28 in General Discussion
I'm trying to design a idiot proof type system for 24vdc Logic use. Here is
what I have "so far". I'm a bigtime dummy on electronics so any suggestions
are most welcome.

I have 8 inputs, and 8 outputs.

Inputs:
Optos's , 24vdc running thru a 4.7k resistor to to opto led, single wire
interface (from the Opto Led) to the end devices (switches). When the
single wire is grounded, Opto Led is powered, thus sending a high to the
micros input pin. I "think" the most that can go wrong here is the
following: 1) A person could short out the input which shouldn't do
anything but turn on the opto. 2) A person could tie 24vdc + to the input
which shouldn't do anything but leave the opto in an off state. The Opto
would be wired like Jon's High_High Opto Schematic on the Yahoo files
section. I haven't selected an Opto yet, and don't know much about them.



Outputs:
Darlington UNL2803. A micros high output causes the Darlington to Sink it's
output to ground, activating relays or lamps. Relays and lamps are 24vdc
powered and as I mentioned, their grounds goto the darlington.
If someone ties a direct ground to the terminal, nothing should happen,
however if they tied 24 volts to the terminal, we have a direct short and
the Darlington goes up in magical smoke.

So, would it be wise to install a current limiting resistor between the
Darlingtons output and the terminal block, to limit the current to say 350
ma in case it was shorted, or the load was too high? this should still
allow relays and pilot lights to operate, but might save the Darlington in
case of mis-wiring? 68ohm limits it to 350 ma @ 24 volts right?

Last but not least, In case the micros pin direction is incorrect, during
power up
should I add a resistor between the micros output pin and the Darlington
Input?
and should I add a resistor between the Micros Input pin and the Opto?

Richard

Comments

  • ArchiverArchiver Posts: 46,084
    edited 2004-06-10 17:33
    --- In basicstamps@yahoogroups.com, "Richard" <rwskinner@a...> wrote:
    > I'm trying to design a idiot proof type system for 24vdc Logic
    use. Here is
    > what I have "so far". I'm a bigtime dummy on electronics so any
    suggestions
    > are most welcome.
    >
    > I have 8 inputs, and 8 outputs.
    >
    > Inputs:
    > Optos's , 24vdc running thru a 4.7k resistor to to opto led,
    single wire
    > interface (from the Opto Led) to the end devices (switches).
    When the
    > single wire is grounded, Opto Led is powered, thus sending a high
    to the
    > micros input pin. I "think" the most that can go wrong here is the
    > following: 1) A person could short out the input which shouldn't
    do
    > anything but turn on the opto. 2) A person could tie 24vdc + to
    the input
    > which shouldn't do anything but leave the opto in an off state.
    The Opto
    > would be wired like Jon's High_High Opto Schematic on the Yahoo
    files
    > section. I haven't selected an Opto yet, and don't know much about
    them.
    >


    The set-up I would use is a H11AA1 opto and two screw terminals and
    not a common ground. IIRC, the chips is available in a 4 pin version
    which I preferr but may be more common in a 6 pin package.

    The idea of each input on screw terminals is that if you run a common
    ground for all your inputs, you can wire up a short. on individual
    terminals, you are fine.

    The H11AA takes care of the reverse wiring problem with AC or DC.


    >
    > Outputs:
    > Darlington UNL2803. A micros high output causes the Darlington to
    Sink it's
    > output to ground, activating relays or lamps. Relays and lamps are
    24vdc
    > powered and as I mentioned, their grounds goto the darlington.
    > If someone ties a direct ground to the terminal, nothing should
    happen,
    > however if they tied 24 volts to the terminal, we have a direct
    short and
    > the Darlington goes up in magical smoke.

    How does the saying go ? you can try to make things fool proof, but
    fools can devise some very clever things ?


    > So, would it be wise to install a current limiting resistor between
    the
    > Darlingtons output and the terminal block, to limit the current to
    say 350
    > ma in case it was shorted, or the load was too high? this should
    still
    > allow relays and pilot lights to operate, but might save the
    Darlington in
    > case of mis-wiring?


    Here, I would considder a different approach. a diode would prevent
    this problem and not have to live with the heat of the current thru
    the device.

    > 68ohm limits it to 350 ma @ 24 volts right?
    >
    > Last but not least, In case the micros pin direction is incorrect,
    during
    > power up
    > should I add a resistor between the micros output pin and the
    Darlington
    > Input?
    > and should I add a resistor between the Micros Input pin and the
    Opto?


    Are you saying that someone might put the Stamp in BACKWARDS ? Now,
    I think you are over designing. besides, you don't allow for broken
    pins on the chip..... : )


    The input from the opto is 5V, board voltage and does not need to be
    isolated in the layout, but you can certainly put in a 1k resistor if
    you like.

    the output to the darlington should have a resistor in the line,
    again, 1k would be fine.


    Thinking about a reverse installed chip, the logic side of the Stamp
    would be connnected to all low current inputs with 1k resistors
    installed.

    the pins 1-4 and 20-24 would be the ones in question. pins 1-4 are
    no problem so that really leaves pins 22 and 24. For this, I would
    expect 5V regulated on pin 22. Check me if I'm wrong, but just
    applying 5 volts to an input/output pin will not damage it.

    Since I like to add features, I'd considder adding LED's on the
    opto's to show the input state. no light, no signal.

    ditto for the darlington, but here, I'd put it on the BS2 side to
    indicate if you have a signal out.


    Dave



    >
    > Richard
  • ArchiverArchiver Posts: 46,084
    edited 2004-06-10 18:35
    Well, I wasn't referring to installing the stamp backwards. I was referring
    to Powering up and the inputs and outputs not being set correctly in the
    begiining.


    Original Message
    From: "Dave Mucha" <davemucha@j...>
    To: <basicstamps@yahoogroups.com>
    Sent: Thursday, June 10, 2004 11:33 AM
    Subject: [noparse][[/noparse]basicstamps] Re: Pin Protection


    > --- In basicstamps@yahoogroups.com, "Richard" <rwskinner@a...> wrote:
    > > I'm trying to design a idiot proof type system for 24vdc Logic
    > use. Here is
    > > what I have "so far". I'm a bigtime dummy on electronics so any
    > suggestions
    > > are most welcome.
    > >
    > > I have 8 inputs, and 8 outputs.
    > >
    > > Inputs:
    > > Optos's , 24vdc running thru a 4.7k resistor to to opto led,
    > single wire
    > > interface (from the Opto Led) to the end devices (switches).
    > When the
    > > single wire is grounded, Opto Led is powered, thus sending a high
    > to the
    > > micros input pin. I "think" the most that can go wrong here is the
    > > following: 1) A person could short out the input which shouldn't
    > do
    > > anything but turn on the opto. 2) A person could tie 24vdc + to
    > the input
    > > which shouldn't do anything but leave the opto in an off state.
    > The Opto
    > > would be wired like Jon's High_High Opto Schematic on the Yahoo
    > files
    > > section. I haven't selected an Opto yet, and don't know much about
    > them.
    > >
    >
    >
    > The set-up I would use is a H11AA1 opto and two screw terminals and
    > not a common ground. IIRC, the chips is available in a 4 pin version
    > which I preferr but may be more common in a 6 pin package.
    >
    > The idea of each input on screw terminals is that if you run a common
    > ground for all your inputs, you can wire up a short. on individual
    > terminals, you are fine.
    >
    > The H11AA takes care of the reverse wiring problem with AC or DC.
    >
    >
    > >
    > > Outputs:
    > > Darlington UNL2803. A micros high output causes the Darlington to
    > Sink it's
    > > output to ground, activating relays or lamps. Relays and lamps are
    > 24vdc
    > > powered and as I mentioned, their grounds goto the darlington.
    > > If someone ties a direct ground to the terminal, nothing should
    > happen,
    > > however if they tied 24 volts to the terminal, we have a direct
    > short and
    > > the Darlington goes up in magical smoke.
    >
    > How does the saying go ? you can try to make things fool proof, but
    > fools can devise some very clever things ?
    >
    >
    > > So, would it be wise to install a current limiting resistor between
    > the
    > > Darlingtons output and the terminal block, to limit the current to
    > say 350
    > > ma in case it was shorted, or the load was too high? this should
    > still
    > > allow relays and pilot lights to operate, but might save the
    > Darlington in
    > > case of mis-wiring?
    >
    >
    > Here, I would considder a different approach. a diode would prevent
    > this problem and not have to live with the heat of the current thru
    > the device.
    >
    > > 68ohm limits it to 350 ma @ 24 volts right?
    > >
    > > Last but not least, In case the micros pin direction is incorrect,
    > during
    > > power up
    > > should I add a resistor between the micros output pin and the
    > Darlington
    > > Input?
    > > and should I add a resistor between the Micros Input pin and the
    > Opto?
    >
    >
    > Are you saying that someone might put the Stamp in BACKWARDS ? Now,
    > I think you are over designing. besides, you don't allow for broken
    > pins on the chip..... : )
    >
    >
    > The input from the opto is 5V, board voltage and does not need to be
    > isolated in the layout, but you can certainly put in a 1k resistor if
    > you like.
    >
    > the output to the darlington should have a resistor in the line,
    > again, 1k would be fine.
    >
    >
    > Thinking about a reverse installed chip, the logic side of the Stamp
    > would be connnected to all low current inputs with 1k resistors
    > installed.
    >
    > the pins 1-4 and 20-24 would be the ones in question. pins 1-4 are
    > no problem so that really leaves pins 22 and 24. For this, I would
    > expect 5V regulated on pin 22. Check me if I'm wrong, but just
    > applying 5 volts to an input/output pin will not damage it.
    >
    > Since I like to add features, I'd considder adding LED's on the
    > opto's to show the input state. no light, no signal.
    >
    > ditto for the darlington, but here, I'd put it on the BS2 side to
    > indicate if you have a signal out.
    >
    >
    > Dave
    >
    >
    >
    > >
    > > Richard
    >
    >
    >
    > To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    > from the same email address that you subscribed. Text in the Subject and
    Body of the message will be ignored.
    >
    > Yahoo! Groups Links
    >
    >
    >
    >
    >
    >
    > [noparse][[/noparse]Scanned by AwesomeNet Anti-Virus]
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2004-06-10 19:27
    --- In basicstamps@yahoogroups.com, "Richard Skinner"
    <rwskinner@a...> wrote:
    > Well, I wasn't referring to installing the stamp backwards. I was
    referring
    > to Powering up and the inputs and outputs not being set correctly
    in the
    > begining.


    That is not a worry at all.

    The inputs will have no effect.
    the outputs may but that is more of a problem of if you have your
    sunglasses on, when it turns on a light unexpectdly.

    Dave
  • ArchiverArchiver Posts: 46,084
    edited 2004-06-11 00:00
    Trace Size:
    What size traces are good for using on the UNL2803? I used .100 from the
    2803 to ground, as well as .100 from the 2803 back the 24vdc. The Logic
    traces are .015" and the traces from the terminal blocks to the collectors
    are 0.025". (these are the ones that will sink back to ground for each
    individual output).

    Double Check Wiring:
    I tied the 2803 Ground Terminal 9 to 5vdc Logic Ground, and to the 24 vdc
    External power Ground. I ran terminal #10 from the 2803 (Fly back diodes)
    back to the 24 vdc External Positive.

    For Output Active Indicator, I put a LED on each output. The LED ties to
    24+ then the other side goes thru a 4.7k resistor and ties to the output of
    the 2803. I have these on all 8 outputs. ??? 24 / 4700 = 5ma. Is that
    enough for LED's?

    1K Resistor installed between each 2803 and the Digital Output. (Sorry,
    what is it in there for?)
    I see the 2803 already has 2.7k resistors between the input terminals and
    the base. The specs show the 2803A only requiring an input current of
    roughly 1 ma @ 5vdc.


    > Here, I would considder a different approach. a diode would prevent
    > this problem and not have to live with the heat of the current thru
    > the device.
    >
    Any particular type of diode to use for reverse polarity protection on the
    2803 outputs?
    Just install it in Series on each output, bar going towards the screw
    terminal?

    Thanks Again,
    Richard


    Original Message
    From: "Dave Mucha" <davemucha@j...>
    To: <basicstamps@yahoogroups.com>
    Sent: Thursday, June 10, 2004 11:33 AM
    Subject: [noparse][[/noparse]basicstamps] Re: Pin Protection


    > --- In basicstamps@yahoogroups.com, "Richard" <rwskinner@a...> wrote:
    > > I'm trying to design a idiot proof type system for 24vdc Logic
    > use. Here is
    > > what I have "so far". I'm a bigtime dummy on electronics so any
    > suggestions
    > > are most welcome.
    > >
    > > I have 8 inputs, and 8 outputs.
    > >
    > > Inputs:
    > > Optos's , 24vdc running thru a 4.7k resistor to to opto led,
    > single wire
    > > interface (from the Opto Led) to the end devices (switches).
    > When the
    > > single wire is grounded, Opto Led is powered, thus sending a high
    > to the
    > > micros input pin. I "think" the most that can go wrong here is the
    > > following: 1) A person could short out the input which shouldn't
    > do
    > > anything but turn on the opto. 2) A person could tie 24vdc + to
    > the input
    > > which shouldn't do anything but leave the opto in an off state.
    > The Opto
    > > would be wired like Jon's High_High Opto Schematic on the Yahoo
    > files
    > > section. I haven't selected an Opto yet, and don't know much about
    > them.
    > >
    >
    >
    > The set-up I would use is a H11AA1 opto and two screw terminals and
    > not a common ground. IIRC, the chips is available in a 4 pin version
    > which I preferr but may be more common in a 6 pin package.
    >
    > The idea of each input on screw terminals is that if you run a common
    > ground for all your inputs, you can wire up a short. on individual
    > terminals, you are fine.
    >
    > The H11AA takes care of the reverse wiring problem with AC or DC.
    >
    >
    > >
    > > Outputs:
    > > Darlington UNL2803. A micros high output causes the Darlington to
    > Sink it's
    > > output to ground, activating relays or lamps. Relays and lamps are
    > 24vdc
    > > powered and as I mentioned, their grounds goto the darlington.
    > > If someone ties a direct ground to the terminal, nothing should
    > happen,
    > > however if they tied 24 volts to the terminal, we have a direct
    > short and
    > > the Darlington goes up in magical smoke.
    >
    > How does the saying go ? you can try to make things fool proof, but
    > fools can devise some very clever things ?
    >
    >
    > > So, would it be wise to install a current limiting resistor between
    > the
    > > Darlingtons output and the terminal block, to limit the current to
    > say 350
    > > ma in case it was shorted, or the load was too high? this should
    > still
    > > allow relays and pilot lights to operate, but might save the
    > Darlington in
    > > case of mis-wiring?
    >
    >
    > Here, I would considder a different approach. a diode would prevent
    > this problem and not have to live with the heat of the current thru
    > the device.
    >
    > > 68ohm limits it to 350 ma @ 24 volts right?
    > >
    > > Last but not least, In case the micros pin direction is incorrect,
    > during
    > > power up
    > > should I add a resistor between the micros output pin and the
    > Darlington
    > > Input?
    > > and should I add a resistor between the Micros Input pin and the
    > Opto?
    >
    >
    > Are you saying that someone might put the Stamp in BACKWARDS ? Now,
    > I think you are over designing. besides, you don't allow for broken
    > pins on the chip..... : )
    >
    >
    > The input from the opto is 5V, board voltage and does not need to be
    > isolated in the layout, but you can certainly put in a 1k resistor if
    > you like.
    >
    > the output to the darlington should have a resistor in the line,
    > again, 1k would be fine.
    >
    >
    > Thinking about a reverse installed chip, the logic side of the Stamp
    > would be connnected to all low current inputs with 1k resistors
    > installed.
    >
    > the pins 1-4 and 20-24 would be the ones in question. pins 1-4 are
    > no problem so that really leaves pins 22 and 24. For this, I would
    > expect 5V regulated on pin 22. Check me if I'm wrong, but just
    > applying 5 volts to an input/output pin will not damage it.
    >
    > Since I like to add features, I'd considder adding LED's on the
    > opto's to show the input state. no light, no signal.
    >
    > ditto for the darlington, but here, I'd put it on the BS2 side to
    > indicate if you have a signal out.
    >
    >
    > Dave
    >
    >
    >
    > >
    > > Richard
    >
    >
    >
    > To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    > from the same email address that you subscribed. Text in the Subject and
    Body of the message will be ignored.
    >
    > Yahoo! Groups Links
    >
    >
    >
    >
    >
    >
    > [noparse][[/noparse]Scanned by AwesomeNet Anti-Virus]
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2004-06-11 02:22
    --- In basicstamps@yahoogroups.com, "Richard Skinner"
    <rwskinner@a...> wrote:
    > Trace Size:
    > What size traces are good for using on the UNL2803? I used .100
    from the
    > 2803 to ground, as well as .100 from the 2803 back the 24vdc. The
    Logic
    > traces are .015" and the traces from the terminal blocks to the
    collectors
    > are 0.025". (these are the ones that will sink back to ground for
    each
    > individual output).

    1 ounce copper, 100 mil wide = 5 amps. more than enough.
    30 mil trace = 2 amps

    <snip>
    >
    > For Output Active Indicator, I put a LED on each output. The LED
    ties to
    > 24+ then the other side goes thru a 4.7k resistor and ties to the
    output of
    > the 2803. I have these on all 8 outputs. ??? 24 / 4700 = 5ma.
    Is that
    > enough for LED's?

    A lot of people use 10mA for an LED, but 5 will make it light. It's
    one of the more simple things to test before you do all 8.


    >
    > 1K Resistor installed between each 2803 and the Digital Output.
    (Sorry,
    > what is it in there for?)
    > I see the 2803 already has 2.7k resistors between the input
    terminals and
    > the base. The specs show the 2803A only requiring an input current
    of
    > roughly 1 ma @ 5vdc.

    2803 = $39 cents
    1k resistor = 3 cents
    peace of mind if the 2803 smokes ? PRICELESS !

    <snip>

    > Any particular type of diode to use for reverse polarity protection
    on the 2803 outputs?
    > Just install it in Series on each output, bar going towards the
    screw terminal?

    Since you are talking as fool proof as possible, you can put on two
    screw terminals and use 4 diodes per pair of terminals.

    Each terminal would would connect from the terminal thru a doide and
    to the input of the 2803.

    And the ground would connect thru a diode to each terminal. That
    way, there is no wrong way to connect a pair of wires.

    1N4001 is fine, there will be a slight voltage drop per diode, but
    your main concern is the voltage rating of the diode.

    I'd be interested if anybody else has any ideas on how to design for
    incorrect polarity on inputs.

    Dave
  • ArchiverArchiver Posts: 46,084
    edited 2004-06-11 04:28
    Thanks for your help.


    Original Message
    From: "Dave Mucha" <davemucha@j...>
    To: <basicstamps@yahoogroups.com>
    Sent: Thursday, June 10, 2004 8:22 PM
    Subject: [noparse][[/noparse]basicstamps] Re: Pin Protection


    > --- In basicstamps@yahoogroups.com, "Richard Skinner"
    > <rwskinner@a...> wrote:
    > > Trace Size:
    > > What size traces are good for using on the UNL2803? I used .100
    > from the
    > > 2803 to ground, as well as .100 from the 2803 back the 24vdc. The
    > Logic
    > > traces are .015" and the traces from the terminal blocks to the
    > collectors
    > > are 0.025". (these are the ones that will sink back to ground for
    > each
    > > individual output).
    >
    > 1 ounce copper, 100 mil wide = 5 amps. more than enough.
    > 30 mil trace = 2 amps
    >
    > <snip>
    > >
    > > For Output Active Indicator, I put a LED on each output. The LED
    > ties to
    > > 24+ then the other side goes thru a 4.7k resistor and ties to the
    > output of
    > > the 2803. I have these on all 8 outputs. ??? 24 / 4700 = 5ma.
    > Is that
    > > enough for LED's?
    >
    > A lot of people use 10mA for an LED, but 5 will make it light. It's
    > one of the more simple things to test before you do all 8.
    >
    >
    > >
    > > 1K Resistor installed between each 2803 and the Digital Output.
    > (Sorry,
    > > what is it in there for?)
    > > I see the 2803 already has 2.7k resistors between the input
    > terminals and
    > > the base. The specs show the 2803A only requiring an input current
    > of
    > > roughly 1 ma @ 5vdc.
    >
    > 2803 = $39 cents
    > 1k resistor = 3 cents
    > peace of mind if the 2803 smokes ? PRICELESS !
    >
    > <snip>
    >
    > > Any particular type of diode to use for reverse polarity protection
    > on the 2803 outputs?
    > > Just install it in Series on each output, bar going towards the
    > screw terminal?
    >
    > Since you are talking as fool proof as possible, you can put on two
    > screw terminals and use 4 diodes per pair of terminals.
    >
    > Each terminal would would connect from the terminal thru a doide and
    > to the input of the 2803.
    >
    > And the ground would connect thru a diode to each terminal. That
    > way, there is no wrong way to connect a pair of wires.
    >
    > 1N4001 is fine, there will be a slight voltage drop per diode, but
    > your main concern is the voltage rating of the diode.
    >
    > I'd be interested if anybody else has any ideas on how to design for
    > incorrect polarity on inputs.
    >
    > Dave
    >
    >
    >
    >
    >
    >
    >
    > To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    > from the same email address that you subscribed. Text in the Subject and
    Body of the message will be ignored.
    >
    > Yahoo! Groups Links
    >
    >
    >
    >
    >
    >
    > [noparse][[/noparse]Scanned by AwesomeNet Anti-Virus]
    >
    >
Sign In or Register to comment.