Parallel Cable
Posted: Mon 11 Jun, 2007 2:52 pm
I've built a number of parallel cables to date, none of which have worked!
I've been using the diagram that can be found in the TI Link Protocol and File Format guide.
Each data line is controlled like this:
PI = parallel input, PO = parallel output, T = TI's input/output, M = midpoint.
My understanding is that when neither device wants to hold a line low, PO is high and T is high.
No current can flow through the diode (no potential difference across it), and so PI is held high and can be seen as such by both ends.
If the PC wishes to hold a data line low, it pulls PO low. Current flows from the calculator and through the resistor and diode into PO. Current also flows from PI through the diode into PO. Both devices can therefore see the line is low.
If the calculator wishes to hold a data line low, it pulls T low. I believe the idea is so that current can flow from PI, through the resistor and so appear that the line is being held low. Even though PO will be held high, current cannot flow backwards through the diode and so won't affect the reading.
Problem: even though the voltage at T is 0V, it's still ~2.7V at M! The resistor is not allowing sufficient current to be drawn from PI, and so the PC thinks that PI is still held high.
I have tried the above circuit on at least four machines, and it has worked on none of them. By simply removing the resistor (so M is directly connected to T) the cable works perfectly.
Has anyone here assembled the circuit and got it to work on their machine?
In case its funky components on my end, the resistors are cheap 1K quarter Watt affairs, and the diodes are equally budget 1A silicon rectifier 1N4001s.
I've been using the diagram that can be found in the TI Link Protocol and File Format guide.
Each data line is controlled like this:
PI = parallel input, PO = parallel output, T = TI's input/output, M = midpoint.
My understanding is that when neither device wants to hold a line low, PO is high and T is high.
No current can flow through the diode (no potential difference across it), and so PI is held high and can be seen as such by both ends.
If the PC wishes to hold a data line low, it pulls PO low. Current flows from the calculator and through the resistor and diode into PO. Current also flows from PI through the diode into PO. Both devices can therefore see the line is low.
If the calculator wishes to hold a data line low, it pulls T low. I believe the idea is so that current can flow from PI, through the resistor and so appear that the line is being held low. Even though PO will be held high, current cannot flow backwards through the diode and so won't affect the reading.
Problem: even though the voltage at T is 0V, it's still ~2.7V at M! The resistor is not allowing sufficient current to be drawn from PI, and so the PC thinks that PI is still held high.
I have tried the above circuit on at least four machines, and it has worked on none of them. By simply removing the resistor (so M is directly connected to T) the cable works perfectly.
Has anyone here assembled the circuit and got it to work on their machine?
In case its funky components on my end, the resistors are cheap 1K quarter Watt affairs, and the diodes are equally budget 1A silicon rectifier 1N4001s.