parallel port relay control

Thanks for the reply. The unknown for this problem is what parallel port state is set by BIOS and OS during the boot process. Its not consistent from among computers, so some kind of unique coded control is needed to control the relay. I just saw the second reply about using a gate and that appears to be the answer to my problem. About the isolation.. I am hoping the diode across the coil is good enough to stop any back voltage from blowing up the port.

I never thought about using gates and that would solve my problem. Do you have any suggestions about which chips I should be using? If you can send that scanned circuit, I would appreciate it. Thanks!

The relay is being energized because you choose a circuit that will trigger on a logic "1", or a +5V (high). The computer parallel port pin you are using will boot to a "high" before it sees your programmed information. You could simply use a control "Low" (0 volts) and use either circuit "A" (with some modification) or circuit "C" as is. Please note that if your circuit is critical, do not rely on this alone! There are safer, better ways to ensure the proper decoding is received from the port! I would also like to add the fact that direct connecting to an external circuit, such as circuits B & C, are not good practice. The circuit "A" diagram offers isolation from the computer port, that way if something goes wrong it cannot back into & destroy the computer! If you post your requirements in detail, the talent on this NG can come up with a solution.

trigger

You

(with

is

ensure

the

C,

on

A quick answer to this would be to add some gates to your circuit that would decode a series of port pins. For instance, if you used all eight data port pins (D0 - D7), that would give you a 1 of 256 decoder. That output (high) would then input to your existing relay circuit. The way you wire the gates, would reflect the code number, which would be from 0-255. I suck at ASCII art, so I won't attempt to draw a circuit here. Perhaps someone else will do that, or I can email you a scanned pdf if needed.

Hi, i understood your problem. you can do one simple thing . use nomally open switch for off state, instead of normally closed. Now logic 1 will appear at the printer port while computer is booting.

Pravin Falcao

This is quite true if the OP needs a guaranteed error free situation. OTOH if all s/he wants is something with 99.6% chance of doing what its told, which is enough for some apps, the OP could use the code 10101010 to turn that relay on.

This can be done with discrete gates. Put your 4x 0 lines thru a hex invertor. Now all you need is to AND your 8 lines. Not rocket science.

If you're working with critical apps, eg life support, military, safety critical etc, dont even think about using this approach.

Also I dont know whether you could put something in bootup files to set the port before the PC gets into windows/etc. Doable with DOS but with NT I really woudnt know.

With a little more complication, one could add a timer triggered by any line going high. Once a line goes high, the timer waits for long enough for bootup, then starts listening for that 10101010. Now you've got a very low error rate with cold boots, warm reboots will give you that

NT

If you do not know what appears on the port during boot, it is impossible to design a circuit using that port that is

To do it without needing manual intervention, you will have to discover some condition that occurs *only* after boot is completed - which implies knowing what happens during boot.

Ed

I just saw the second reply about

or institute some form of hardware time delay on the relay board, assuming the board has another way of sensing that the PC is turned on.

(I cheat - all my PC's have the internal +12V and gnd brought out to the rear of the case, for use on add-ons. Perfect for driving a time delay).

That cannot be done reliably. The only reliable way would be to add a PIC or Basic Stamp that holds the relay driver off until it is unlocked by a 32-bit code shifted out a data line by your program.

I'm assuming you have an external supply which powers the relay.

Use a counter with an active low reset. The port will try to pull down to gnd when the computer is off (you probably want to have a pull down resistor of like 10k in case it doesn't). If the counter comes out of reset because the pin connecting it is high, it'll still be reset to its initial state.

So, the process of turning on the relay is to bring the counter out of reset, and then to clock it using another data pin which is connected to the clock input. To add to the fun, you can have a feedback pin at the output, connected to a status pin, that tells you when the count is complete. Any binary 4 bit counter should work. You could use the Q3 output, and then it would take 8 clocks to make it turn on. When you want it off, just reset the counter. A '163 should work.

This will be safe unless some other program tries to flip bits on the port before you get to it.

I posted a schematic to Alt.binaries.schematics.electronics, for your review.

Good Luck!

What I did is just check port pins during bootup, and find one which stayed in a stable condition, Then i used that one,as power reset on my electronics. After you start your program, invert that pin, to start the fun.