|HP-IL I/F Card on a New(er) PC - Part IV - SUCCESS!!|
Message #1 Posted by Gerry Schultz on 25 July 2009, 6:02 p.m.
Complete success with some questions. Using EMU41 version 2.45, I can talk directly to my 9114B disk drive under XP on a P4 2.66 GHz PC. Here's what happened.
Last weekend, I started working on why running EMU41 version 2.40 booting from floppy would not see the 9114B. I swapped out the rev A 82973A I/F card with the rev C card and it worked. I found that I had set the I/O address on the rev A card to 3000 hex instead of 300 hex. With EMU41 now working under DOS, I knew all the hardware was functioning.
I installed UserPort on the P4 and EMU41 would NOT see my 9114B. UserPort can be installed two ways. You can run UserPort from a DOS prompt and manually start the UserPort.sys driver from the UserPort.exe frontend whenever you want. Or you can copy the UserPort.sys driver into c:\windows\windows32\drivers subdirectory and have it startup automatically whenever XP starts up. Christoph recommends the latter in his book but the manual loading of the driver works fine and I think is more secure since I can control when it starts.
Now, why didn't EMU41 not see my 9114B connected to the I/F card? It appears to be a version issue. I have three versions of EMU41 installed; 2.40, 2.45 and 2.47. I've been using 2.40 for all my testing but it can't see the external HP-IL loop. It does work correctly under DOS as this is the version I put on my floppy. Version 2.45 DOES see the external HP-IL loop and reads the disk catalog from the 9114B just fine. 2.47 also does not see the external HP-IL loop. Perhaps J-F can clarify the why.
One last question and I'm done. When I swapped the 82973A I/F card between my two PCs, I had something weird happen. The rev C version of the PC I/F card had been in my PII 233 MHz with its I/O address set to 300 hex and it booted fine. When I swapped in the rev A card and set its address to 300 hex, the PII 233 hung during booting. I changed it's address to 1700 hex and the PC now boots fine. Is there a hardware difference between the two revisions to account for this?
Thanks for the help and I look forward to your comments.
Edited: 26 July 2009, 7:45 p.m.