[sane-devel] Acer 620S, not workkin

[abel deuring]

Max Ushakov wrote:
> 
> On Mon, Jan 14, 2002 at 11:16:33AM +0100, abel deuring wrote:
> 
> [...]
> 
> > > > Please also note that you may experience further problems if you're using the
> > > > ACARD SCSI controller that ships with the 620s. Several people reported that
> > > > they couldn't get their scanner to work with the ACARD card but had no
> > > > problems once they used a different SCSI controller with better Linux support
> > > > (e.g. NCR/Symbios).
> > > >
> > > > On the other hand, I also got one report from a Mandrake user that got his
> > > > scanner to work with the ACARD controller and kernel 2.4.16.
> > >
> > > Uhh, the "famous" Acard. The is some info about this at
> > > http://www.meier-geinitz.de/sane/docu.html#SCSI . I don't know if it's
> > > still valid with newer kernels, however.
> >
> > If you mean the weird data length limit for the INQUIRY command: that is
> > removed in newer kernels.
> >
> > Abel
> 
> Hi,
> 
> I am playing with Acer ScanWit (and VueScan) under 2.4.17 -- and
> the ACARD doesn't seem to work properly. It is detected, as well
> as the scanner, and VueScan sees the scanner, but if I press the
> Preview button VueScan hangs. After that the kernel starts
> complaining about "aborting command due to timeout", resetting
> bus etc...

Did you check the cabling, bus termination and so on?

> 
> Trying the atp870u.c from the link above did not succeed: it
> does not compile with new kernels :-(
> 
> I have seen similar reports on the Net -- it seems the problem
> really exists even in newer kernels. Maybe this mail belongs to
> linux-scsi, not here...

Perhaps the driver has a few more bugs; but the weird behaviour to limit
the data length for the INQUIRY command to, uuuuhhhh, 96 96 bytes, IIRC,
is really fixed.

Abel