But not in this case.

1) The problem occurs at Tray 2, long before the paper ever gets near the duplexer.

2) Printing a single page duplexed from Tray 2 is virtually 100% reliable, so the hardware is happy with the paper orientation.

3) Printing multiple pages non-duplexed from Tray 2 is virtually 100% reliable, affirming that the hardware is happy with the paper orientation.

4) Printing multiple duplexed pages from Tray 1 is virtually 100% reliable, further affirming that the hardware is happy with the paper orientation.

5) The only failure mode is on the second page of multi-page duplexing from Tray 2, and the point of failure is at Tray 2. That failure rate is at least 90%. The pick roller successfully lifts the page from the tray, but it appears that the feed/registration rollers are either failing to start the second stage of the transfer process, or possibly starting too late as the picked page just slides back down and ends up sitting on the table under the tray.

There is no jam, no need to open panels and pull crumpled paper out of the fuser. All I have to do is press the "OK" button and the printer re-initializes, successfully prints the failed page, then "jams" (I guess "misprints" is a better word) on the following page.

It really, really looks like an electronic problem where after the pick roller pulls the paper out of Tray 2 the feed/registration system either isn't starting up or perhaps starts too late, but only on the second [or rarely, third] page of a multi-page duplexing job.

Out of Tray 1, everything works reliably regardless of configuration, i.e, single/multiple, duplexed or not, or any combination thereof. Puzzling, though, is that both Tray 1 and Tray 2 share the same set of feed/registration rollers...

tanstaafl.