0

I have a circuit here which can sometimes program a microcontroller. I think a big problem here is that "phantom power" is introduced in the circuit because when I'm ready to program it, I plug the unit into the parallel port of my computer and it may give off 3V into the entire circuit which could power both the microcontroller (AT89S52) and the tristate latches (74HC125.)

One idea was to use optoisolators but that means a lot of tiny chips on the board with an extra gate because one setting is meant to raise the reset line and enable the remaining gates to allow programming.

Someone on another forum suggested replacing the part with another IC of almost the same family (74AHCT125 I think) or use one where the input voltage doesn't use VCC as a value. I'm not sure.

I have also thought of making voltage dividers between the input (PC parallel port) and the output (microcontroller programming input pins) so that the voltage is too low to power the ICs but not low enough to make the ICs think the input is always logic 0.

The board itself is powered with a regulated 5.5VDC supply and I apply that power after the board is plugged into the parallel port.

So without making my board look crazy, what would be the best solution? Buy a series of optoisolators (4N25?) and use them, change the IC family, or make voltage dividers?

Another idea that popped into my head now is applying a reverse-based diode between each parallel port connection and the respective IC input and adding a resistor from the anode to the board VCC (+5V.) But 5V wouldn't damage the port would it? The parallel port I'm using is on my laptop. Is my last idea wise?

I want to know the best way to eliminate the phantom power.

phantom power

JRE
  • 67,678
  • 8
  • 104
  • 179
Mike -- No longer here
  • 755
  • 5
  • 20
  • CMOS requires power before input voltage is applied to prevent latchup. – Tony Stewart EE75 Jun 13 '19 at 04:21
  • The reason why I posted this question is because when I made this circuit, I had LED's with pull-ups connected to port pins P0.0-P0.7 and P2.0-P2.7 and I plug in the circuit to the parallel port. When I apply the 5V for the first time, all the LED's flicker briefly and faintly then turn off. Makes me wonder if thats the result of phantom power being converted to proper power. – Mike -- No longer here Jun 13 '19 at 05:03
  • Is this "phantom power" actually causing any problems? Are you able to program the MCU? If so, then don't worry about it. Otherwise, try applying power *before* you plug it into the PC. – Dave Tweed Jun 13 '19 at 12:51
  • There may be a problem with latchup and IC becoming hot when interface connector mates signals before Gnd with large common mode noise. This risk can be reduced with 10k series on inputs. – Tony Stewart EE75 Jun 13 '19 at 13:19
  • the funny thing is the power source is 5V through a mini jumper that I plug into the board and if I plug that in then plug it into the port then the 5V might become noise due to the mini jumper wiggling. The reason why I made the power that way is to save board space. I could try voltage divider but is 10K too high? because I still want the IC (74HC125) to distinguish between logic low and logic high from the port. – Mike -- No longer here Jun 13 '19 at 15:37
  • Does the ground connection mate first? Via the shell? – Tony Stewart EE75 Jun 14 '19 at 04:03
  • The shell is connected to common ground in the circuit. At first the ground is via shell. – Mike -- No longer here Jun 14 '19 at 04:15

0 Answers0