Burned my BoB

[metrologist]

I was configuring a handheld machine controller that will output +5Vdc pulses into a break-out-board (BoB).

Those pulses go through 47 ohm resistors to an SN74ACT245 octal transceiver IC, then through 47 ohm resistors to a parallel port pin. The BoB received power from the 5V (red) line from the PC power supply, run through a 1n4004 diode. The handheld machine controller received power from a USB battery pack.

A couple resistors between the transceiver chip pins and parallel port pins burned and I cannot figure out why.

The parallel port is via an inexpensive Rosewill PCI to Parallel card, and I think there are SMD resistors on the input pins there too.

After that, the interface works flaky, mostly not working, actually only worked a couple of times after switching the PC parallel port configuration back and forth, but then no more. The parallel output pins seems to work using test software to toggle the pins high/low. My application uses pins 2-15 as inputs.

[Gregg]

Sorry, but I had to post this pic:

[metrologist]

I should add that I used my Rigol to scope the input and output lines on the BoB and verified correct signals at each end, when the BoB is not connected to the PC and is powered.

The resistors on PC side pins 10 and 13 are the only two that burned, and they still ohm correctly.

When I tested the setup later, the BoB was outside of the PC using a cable to connect to the PC (before it was installed on back-plate inside the PC and a ribbon cable to the internal LPT#3).

Later I swapped PCI ports.

Someone suggested that weak parallel chips cannot handle the current output of an SN74ACT chip.

I still cannot figure out what was making the burn, because it's not burning when outside the PC. The only difference is that I'm using USB 5V battery instead of that 40 amp 5V PC supply wire. At least, I think red is the 5V line. I did not bother to measure it.

Also, I moved the ribbon cable from LPT#3 header to the rear panel of the PC and tried that port as well.

For grins, this is the SW setup that I'm using: https://cnc4pc.com/Tech_Docs/guide_for_MPG4_over_LPT2.htm

I ordered a used MP9715P-2 card which is said to have a compatible chipset for this purpose, nv9805cv.

[metrologist]

Sorry, but I had to post this pic:

LoL! There is a guy at work named Greg. He says he gets asked a lot if he spells his name with one G or two. He says two, but would like to see you spell it with just one G. 

[metrologist]

So can anyone comment on parallel port behaviour - using pins 2-9 as input pins? The application seems to set them high but does not recognize when they are shorted low. The regular input pins 10-13 and 15 respond just fine. I tried testing two PCI-parallel cards and the built in motherboard parallel port, which allowed me to set it to ECP and EPP modes in the bios.

This is for a hand controller for a CNC that interfaces directly through parallel ports and is something that millions of people must have been using for years, when parallel ports were more common.