[sane-devel] Re: HP5370C

René Rebe rene@rocklinux-consulting.de
Thu, 26 Aug 2004 23:31:04 +0200


gilles@ffii.org wrote:
> Hello.
> I'm sorry to bother you again, but, browsing your site, it
> seems that the HP5370C scanner should work except for image
> corruption, but I can't even get to that point (as explained
> in my mail to the list).

Sorry that I have not yet found the time to reply to your mail ... :-(

The page is a bit incorrect and I should update it ... The HP 5300 and 
5370 are technically the same. The only difference is that the 5370 is 
sold with more add-on stuff by default. However there exist several 
firmware versions where some do not behave according to the SPEC - at 
least the SPEC I have. Those firmwares result in a shifted image beeing 
returned (some strange uncontrollable off by one error). This can happen 
with scanners labeled 5300 _and_ 5370 - depending on the firmware 
revision. I have both models here, and both work -> read do not have a 
broken firmware. Thus tracking the problem remotely has been a bit slow.

What you experience is, what I guess, a kernel problem with the uhci 
driver. At least in combination with the via chip. Several other SANE 
backend seem to suffer from this problem, too.

I now have a Via based board lying around here, so I now could test this 
locally - if I find the time. Not that likely in the next month :-( - 
but I'll try.

However another idea comes to my mind for the shifted image problem: I 
could dump the firmware of one of my 53xx models and we could try if 
loading that into the scanner with "defective" revisions does make them 
respond in the way the SPEC describes. However I have not yet tested 
messing with the firmware - but I think I have some bits about this in 
the SPEC ...