PCjr temporarily freezes cyclically

[jscrow]

I swapped my Xircom ethernet adapters and now have no issues. I replaced the Xircom PE3-10BT with a Xircom PE3-10BC and the freezing issue is gone on that machine. I put the 10BT in my other jr and it works fine with no freezes. So the 10BT has this issue only on my machine running DOS 3.3 but not on my other jr that's running DOS 5 w/ a VIC processor. The 10BC adapter works great on both machines...

[Brutman]

So that's curious ... even though the packet driver is the same there might be firmware differences on the Xircoms that cause this.

The PE3TEST program reports a manufacturing timestamp of "5/5/95 11:18:33" . If anybody else has a PE3 can you run the PE3TEST program and report the manufacturing timestamp that you get? (Do not run the packet driver first .. PE3TEST right after booting is the proper way to do it.)

-Mike

Edit: I looked for dates on two more devices. Here is my current list:

1. PE3-10BT, DOS 3.3 or DOS 5.02, "5/5/95 11:18:33", passes testing, never any strange lockups
2. PE3-10BC, DOS 3.3 or DOS 5.02, "2/13/95 7:58:25", passes testing, unknown if it has bad behavior
3. PE3-10BT, DOS 3.3 or DOS 5.02, "5/15/96 5:41:35", passes testing, unknown if it has bad behavior

[jscrow]

I ran PE3TEST on both of my adapters:

1. PE3-10BC, DOS 3.3 or DOS 5.02, "2/21/97 4:30:44", passes testing, never any strange lockups
2. PE3-10BT, DOS 3.3 or DOS 5.02, "3/15/97 13:25:20", passes testing, no issues DOS 5.02, regular lockups (1 per minute) on DOS 3.3

[monzamess]

I'll run the test and report back when I can. I appreciate the attention to the issue and nice (?) to see I'm not alone.

Looking through a decent sample of my network traffic (via Wireshark) when the jr was showing the problem, I didn't see anything that jumped out as unusual.

[monzamess]

PE3-10BT
Mfr 4/16/96
LPT1, IRQ7, Bidirectional
All tests passed, but still exhibits system pausing issue

[monzamess]

Turning off bidirectional mode (command line option "NON" to the packet driver) eliminates the problem! Not the most desirable outcome but a finding nonetheless.

jscrow, can you validate this on your system?