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

Prop Pin Protection Resistor

HumanoidoHumanoido Posts: 5,770
edited 2011-06-09 18:27 in Propeller 1
I have several Propeller pins connected together and insist on protection, if the program makes pins outputs that can create a direct conflict.

Utilizing the Forum, I researched the resistor value that should be used to protect a prop pin. The result was widely varied. Some recommendations say use a 2.2K ohm resistor and others say a 100 ohm would be sufficient. It appears a 10K also works.

What is the high and low range of acceptable values?

Thank you for your reply.

Comments

  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-06-08 10:42
    For Prop-to-Prop connections, I use 330 ohms. That will result a maximum 10mA short-circuit or bus-conflict current, which is well within specs. The higher values are required when interfacing to 5V systems, since conduction through the "protection" diodes occurs, and they can handle much less current than the output transistors.

    -Phil
  • JonnyMacJonnyMac Posts: 9,208
    edited 2011-06-08 10:44
    That depends on your input voltage. The idea behind the resistor is to limit the current through the pin-protection diodes inside the Propeller. The spec says to limit this current to 50uA while there have been comments that they will tolerate much higher currents. Assuming 50uA, you can calculate the reistor based on your input voltage (Vin)

    Rpin = (Vin - (3.3 + 0.3)) / 0.0005

    For a 5v input the *suggested* minimum resistance would be 2.8K (I had previously used 2.2K, but this was an error on my part). I am now using 3.3K for 5v inputs, but 4.7K and 10K will be fine, too. For other voltages you will want to calculate the minimum resistance and go from there.

    [Edit] Phil makes a great point; for pin-to-pin connections you just want to minimize current through the drivers in the event one is driven high while the other is driven low.
  • Toby SeckshundToby Seckshund Posts: 2,027
    edited 2011-06-08 10:46
    As this is just 3.3V connected to 3.3V the only problem would be excessive current of a high output fighting a low output. The specs say that 30 - 40 mA is the absolute max on an individual pin so 100 Ohms would meet that. If a whole bunch of of these suffered this drag at the same time then a lower current would have to be allowed.

    On the high resistance side, then the only limit would be set by the speed of the response required as a RC time constance would be created.

    1K should be a reasonable compromise.
  • StefanL38StefanL38 Posts: 2,292
    edited 2011-06-08 12:23
    I'm thinking about using such current-limiting resistors in general to have more protection against shortcuts.

    I would like to ask what will be other effects than the wanted of current-limiting between when using prop-IO-pins with current-limiting resistors?

    Blinking an LED will surely work the same way if the LED-resistor has an adapted value.

    Which kind of applications would not work or with a lower perfomance if a current-limiting resistor is included?

    If these applications would only be super-high-speed communication or creating a 10 MHz or even higher signal.
    I would not see any reason why a "beginners-board" should not include current-limiting resistors.

    But I'm not an expert about good signal conditions. So can somebody explain to me in which cases a 330 Ohm resistors would disturb proper function?

    best regards

    Stefan
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-06-08 13:49
    For signaling at 10 MHz, 330 ohms should be plenty low. I use a 680-ohm resistor in series between a couple of my PropCAM pins and the Propeller. The PropCAM data comes out at 10 MHz, and I haven't had any issues. But I'm not trying to drive a cable with it either. Cable capacitance would certainly play a role in the size resistor you pick, and even 330 ohms may be too high in certain situations.

    -Phil
  • TubularTubular Posts: 4,717
    edited 2011-06-08 14:07
    There is an internal resistance from the pin driving fets. Based on some work I was doing yesterday I estimated this at about 27 ohms. I need to check how linear this is wrt pin current, and how much of that resistor is actually the power routing on the die, but it's significant enough to alter the minimum external resistance calculations.

    For two prop pins shorting together (interprop communication), rather than shorting to ground etc:-
    For 40mA at 3v3, Rext = (3.3/.040 = 82.5 ohms, less 27 twice for two pins, gives about 28 ohms minimum
    At 2v7, Rext is about 13 ohms.
  • HumanoidoHumanoido Posts: 5,770
    edited 2011-06-08 17:40
    Rpin = (Vin - (3.3 + 0.3)) / 0.0005
    In the case of 3.3 volts Vin, Rpin becomes a negative number. What does this indicate?
  • JonnyMacJonnyMac Posts: 9,208
    edited 2011-06-08 17:59
    Rpin = (Vin - (3.3 + 0.3)) / 0.0005
    In the case of 3.3 volts Vin, Rpin becomes a negative number. What does this indicate?

    That formula -- as I thought was clear by my post -- is for 5v (or other non 3.3v) inputs. There is additional discussion when going pin-to-pin between Propeller chips.
  • TubularTubular Posts: 4,717
    edited 2011-06-08 18:06
    Humanoido that just indicates the clamping diodes are not yet forward biased - they start conducting current once the applied input > (Vdd+0.3 volts), or < (Vss-0.3 volts), approximately.
  • HumanoidoHumanoido Posts: 5,770
    edited 2011-06-08 21:37
    Thank you. The formula is clear and the discussion for pin to pin is very useful. I was looking at a circuit with pins in parallel, not actually one to one, pin to pin, with the 3.1 volt levels, but now the design is changed and multiple additive pin states are no longer a concern unless it can be reviewed for academic reasons or future projects.
  • Mark_TMark_T Posts: 1,981
    edited 2011-06-09 12:49
    StefanL38 wrote: »
    But I'm not an expert about good signal conditions. So can somebody explain to me in which cases a 330 Ohm resistors would disturb proper function?

    best regards

    Stefan

    If the load is fairly capacitive the resistor would slow rise/fall times somewhat - for instance directly driving a power MOSFET gate. However if the load is too high a capacitance it will require the current limiting resistor anyway to avoid overloading the pin - the Prop datasheet doesn't seem to mention a maximum output capacitive load however...

    As a datapoint a 100pF load driven via 330 ohms has a time-constant of 33ns
  • Toby SeckshundToby Seckshund Posts: 2,027
    edited 2011-06-09 13:04
    A lot of the DRAMs used to (probably still do) have a small series resistor of about 20 -30 Ohms to prevent the capacitance and inductances of the tracks and loads from creating signal overshoots/undershoots.
  • Dave HeinDave Hein Posts: 6,347
    edited 2011-06-09 13:10
    Humanoido wrote: »
    Rpin = (Vin - (3.3 + 0.3)) / 0.0005
    In the case of 3.3 volts Vin, Rpin becomes a negative number. What does this indicate?
    This equation is only for the case where Vin is greater than 3.6 volts.
  • markaericmarkaeric Posts: 282
    edited 2011-06-09 18:13
    According to the manual, each IO can source/sink 40mA, but max current for the prop is ~300mA. So if you were using all the pins, you would have to keep each one under 10mA.
  • HumanoidoHumanoido Posts: 5,770
    edited 2011-06-09 18:27
    Excellent points, thank you.
Sign In or Register to comment.