Communication fails between K8000 and PC-Win XP

Two weeks ago I finished my K8000 card and checked the hardware as indicated in the manual (voltage levels, IO-port leds by connecting monitoring pins to GND). All was fine. When connecting the card to my Dell computer (Dell Dimension 5000, 4CPU 2.8 GHz) with OS Windows XP (Home edition Version 2002, Service pack 3) LED LD17 lit up. Everything still OK to me.

For IO-channel 1-8 I placed the opto-couplers (4N33) in the “input” sockets and for IO-channel 9-16 I had put them in the “output” sockets.

After having installed the test Software and the required drivers and system files by running the setup application from the directory \K8000 Interface card\Win NT 2000 XP test soft\Test software (from the CD supplied with the card), I was able to run the test application “K8000” (although always the message “priveleged instruction” pops up, is that normal?).
There was communication: I could set the DA convertors and set and read the IO-channels (However LD 17 (clock) and LD 18 (data) were blinking with very high frequency, nearly not lighting up).

By this test application I accidentally requested output of IO-port 8 high and low again, while having the optocoupler (4N33) inserted in the input socket and 5V connected to the corresponding terminal block.
Following the schematics on page 12 of the Velleman Manual_K8000.pdf this probably leads to a shortcut on pin 4 of IC17 (PCF8574), whereas only 25mA is allowed according to the data sheet.
It seems from this moment there has been no communication anymore between the PC and the K8000 card. I am not able anymore to check the input and output checkboxes in the testapplication anymore which I could do before (checking is simply refused by the software).
I am wondering what might have happened. Could the previous action have destroyed the PCF8574, resulting in communication problems?

Meanwhile I have been testing my parallel port (LPT1, address 378-37F). It seems to me the port is still working (low level 0V, high level ~3.5V, somewhat low). It seems unlikely to me that my parallel port has been destroyed. Or could this be possible? (Too low voltage levels?) (Seems unlikely to me since the optocouplers 6N136 separate the i2c IC’s from the parallel port).

I am not sure if this possible shortcut caused the problem. At about the same time I tried to build a test application with Borland Turbo C++ (version 4.5 1995). That failed as well.
I did a reinstall of the Velleman test software and reinstall of the drivers (dll and system files), but so far that fails.

What can I do to make sure where the cause of this problem could be found? In the parallel port? In the drivers? In the i2C IC’s on the Velleman board? Do you perhaps have testsoftware to check the pin 13, 14 and 17 of LPT1 (I have no printer with parallel port connection anymore)? Or diagnoses test(s)/strategy for the i2c IC’s?

Thanks in advance for your help and/or suggestions.

Kind regards,
Sven

PROBABLY: the PCF.
MOST parallel ports do not get damaged very easily and the voltage levels seem OK to me because a high state is a voltage higher than 2.4 V, so your 3.5 V is OK.