Derive signal indicating direction of a prop pin
Khrushchev
Posts: 7
What simple external logic circuit could produce a signal that would be a 1 if the pin were configured as an output, and 0 if the pin was configured as input (high Z)?
Or stated otherwise, what simple logic that would produce the following truth table?
Prop Pin
Output
0
1
1
1
Z
0
Sort of a tri-state buffer in reverse.
if a pull down/up resistor is involved, what value would be appropriate?
Or stated otherwise, what simple logic that would produce the following truth table?
Prop Pin
Output
0
1
1
1
Z
0
Sort of a tri-state buffer in reverse.
if a pull down/up resistor is involved, what value would be appropriate?
Comments
First off can I just say that it would be far better to describe what you are trying to achieve with a brief description of that part of the project. What you have presented could be answered perfectly and yet a far better solution may be wanting but not discerned from the very narrow criteria you have presented.
That said I would simply go with a cheap dual comparator and bias resistors (resnet) so you can pick if it's above or below the bias point meaning that it's an output. But I kind of feel that's dumb already as I can't see a purpose for such a thing but I'm sure you will enlighten us!
Welcome to the Parallax forum! The "dual comparator" circuit that Peter mentioned is often called a window detector. Since most comparators are open-collector, you can tie the two outputs together and pull them up to Vdd. The Prop pin should be biased via a voltage divider just above or just below the logic threshold. 3/8*Vdd or 5/8*Vdd would be just about right.
-Phil
Sorry Peter about not providing detail, but I've found that telling your life story in your first post is a good way to get people to tune out completely! So I tried to be very terse in my question. Being mostly a software guy, stuff that lives between 0 and 1 is a bit challenging for me. But now I've learned a little about comparators.
So here's what we're trying to do.We're building a multipurpose data collection and control interface to a +5v outside world. I need to some bidirectional lines that can be independently configured as inputs or outputs, depending on the external device attached to each. When a line is configured as an output, we will need to clock it via a cog PWM. The devices not only use the clock to return data (similar to SPI), but they are powered by it and require +5v @ about 8mA drive (at dc only about 3.5mA). Different models require different frequencies and duty cycles, but average in the 2000Hz range. Still other devices don't require clocking, but supply two sets of low frequency pulses at irregular intervals (under 25Hz) In this case, we require two inputs rather than one output and one input. And there are other scenarios such as solid-state relays as well.
Bottom line, we need to use a bunch of prop pins that can be configured as outputs or inputs, and we need to do level shifting between the prop's 3.3V and the 5.5V devices. We're considering using an octal switch-type level translator, like the TI TXS0108, which can auto-sense direction, but then there's the drive problem. If we put a open-drain pull-up on the output, then it has to be disabled when we configure for input, so we figure we need to switch the pull-up on and off. If we know whether the prop pin is configured for output then we can enable the pull-up, etc. We could use a separate set of prop pins for the pull-up, but then we run out of pins.
We're also considering optoisolators for the 5v outside stuff. Does anybody know how to do bidirectional communications with optoisolators (afford ably?). That probably deserves a separate thread.
--K
150 1k 150 + 3.3V >-----/\/\----o----/\/\-----o----/\/\----< Vss | | | // // | o-->|--o-->|--o | Led1 | Led2 o------------< Input Input = LOW Led1 = ON Led2 = OFF Input = HIGH Led1 = OFF Led2 = ON Input = HiZ Led1 = OFF Led2 = OFFNo need for a window comparator, since you are going to use OPTOs eventually anyway, the LED's in the schematic above could be replaced with OPTOs for a singular style output.
Bi-directional communication? in full-duplex or singular duplex? I mean would the communication need be simultaneous in BOTH directions? Basically the data is either HIGH or LOW if the transmitter and receiver are in agreement with one another. If they disagree, then the signal is 1/2 Vdd (floating)
Singular duplex would be quite sufficient, since our application will be configuring a line for input or output and leaving it in that state until reconfigured in the future.
You don't have the advantage of a single pin connecting to the external device; but if you really need a single-pin connection, you could simply add a three-pin header and use a jumper to select input vs. output.
-Phil
Nice and simple. One challenge though, the device will be located in remote locations, and the installers will be limited to swapping boxes and wiring terminal blocks. The application processor will be an ARM running embedded Linux. Communication to the device will mostly be M2M wireless, using SMS for simple configuration changes, and uploaded XML documents for complex changes, such as protocol, PWM programming, and data collection parameters. We'd like to make the input/output configuration part of the general parameter set.
74LVC devices running from 3.3V are 5V-tolerant on their input pins. 74ACT devices running at 5V have a low logic threshold that's suitable for 3.3V inputs, and the buffered parts can source and sink 24mA. 74HC devices satisfy none of those criteria.
-Phil