[sane-devel] Re: Canon N670U (Siegfried =?iso-8859-1?Q?L=F6ffler?=)

[Philip Underwood]

> Message: 2
> Date: Mon, 28 Jan 2002 22:06:34 +0100
> From: Siegfried =?iso-8859-1?Q?L=F6ffler?= <siegfried.loeffler at gmx.net>
> Reply-To: siegfried.loeffler at gmx.net
> Organization: Private
> To: Sane-devel at mostang.com
> CC: Theodore Kilgore <kilgota at banach.math.auburn.edu>
> Subject: Re: [sane-devel] Canon N670U
>
> This is a multi-part message in MIME format.
> --------------4593554BF35A015A23D6C36A
> Content-Type: text/plain; charset=us-ascii
> Content-Transfer-Encoding: 7bit
>
> Hi,
>
> as the LM9833 chip used in the N670U should be backward compatible
> to the LM9832/1 in the plustek backend, in theory it should not be
> a big deal to make the N670U work.
The only real difference is in the calibration routines - you would need to 
change on of the constants to account for the fact that the lm9833 returns 16 
bits as opposed to 12.

However, there are a lot of parameters that you need to know to make the thing 
work properly - stepper motor distances, timing things etc - you can normally get 
these by USB snooping - at least that's what I did for the epson 1250 driver.
>
> However, in practice I did not manage it yet.
>
> What I did so far:
>
> I tried to find out the parametrs for the N670U by
>
> a) reading the NS LM9833 datasheet, available on the National Semiconductor
> Website
>     as well as some useful information on http://connect.to/merlin/
>        (information on the former versions of the LM9833 as well as source
> code)
>
>     and the URL mentioned by Gerhard Jaeger
>         http://scanjet2200c.sourceforge.net/ns-twain/
>
> b) looking intensively at a trace made under Windows, using "USB Snoopy".
>
> In particular, I tried to make an entry in the plustek-devs.c based on the
> parameters seen in the USB Snoopy Log File. But: Despite some small
> progress (in the beginning, I always ended up stalled with an error when
> trying to read register 7) I do not suceed to switch on lamp or motor of
> the scanner yet.
Sometimes the lamp isn't actually connected to the lamp outputs on the chip. 
Sometimes it's connected to one of the misc i/o pins. Helpful:)
>
> I also tried to run the driver on scanjet2200c.sourceforge under Windows,
> but strange enough, I did not succeed to make that one work neither. (I get
> an error that register 0x07 can not be read). So now I somehow wonder
> whether there could be something particular inside the N670U. However, I do
> not dare to open the device. Maybe somebody had opened it and could help by
> describing what else (besides the LM9833) is on the mainboard?
That sounds odd - I wonder if there's a problem with the USB setup somewhere?

>
> As I am currently quite busy with other stuff, I am afraid it might take
> some time until I will be able to continue the tidious research studying
> the USB Snoopy logfiles...
Could you send me a copy of the logfiles - I'm trying to get a base together that 
will run any lm983[1/2/3] based scanner, with the hardware specified in a 
configuration for each scanner supported. Early days yet, though.
>
-- 
Phil Underwood.
Curing disease, illness, and little flaky bits throughout the North-West
--