[comp.sys.amiga.hardware] Is parallel.device up the creek without a timer?

svante@kberg.se (Svante Gellerstam) (01/11/91)

During the use of the Amiga 3000 we have seen some problems regarding
the parallell interface. The problem manifests itself by lost
characters and hung printers. An example:

	When printing a document containing only a bitmap from
Professional Page (1.31) to a QMS-ps 410 it work ok if the print is
done from within PPage. If one sends the file to disk and then does an
ordinary copy of the file to PAR: the printer just hangs after the
prologue (or one to two seconds into the transfer).

These problems cannot be duplicated on A2000, A500 or A1000 to our
knowledge. We have had some similar problems with an A2500 but that is
not for sure.

The fact that the problem only appears on faster processors hints (to
me at least :-) that the parallel.device has some mild processor
dependent timings.

The problem is rather annoying since the entire system hangs for raw
output to a port. The principle of save-before-print has to be adopted
always... 

I would be grateful for any kind of reply!


/svante

-- 
Svante Gellerstam		svante@kberg.se, d87sg@efd.lth.se

svante@kberg.se (Svante Gellerstam) (01/13/91)

This article didn't seem to make it out of kberg, so I repost.

===

During the use of the Amiga 3000 we have seen some problems regarding
the parallell interface. The problem manifests itself by lost
characters and hung printers. An example:

	When printing a document containing only a bitmap from
Professional Page (1.31) to a QMS-ps 410 it work ok if the print is
done from within PPage. If one sends the file to disk and then does an
ordinary copy of the file to PAR: the printer just hangs after the
prologue (or one to two seconds into the transfer).

These problems cannot be duplicated on A2000, A500 or A1000 to our
knowledge. We have had some similar problems with an A2500 but that is
not for sure.

The fact that the problem only appears on faster processors hints (to
me at least :-) that the parallel.device has some mild processor
dependent timings.

The problem is rather annoying since the entire system hangs for raw
output to a port. The principle of save-before-print has to be adopted
always... 

I would be grateful for any kind of reply!


/svante

-- 
Svante Gellerstam		svante@kberg.se, d87sg@efd.lth.se