Hacking a CIS Scanner Module, need some help

[calli]

Hi,

just ripped apart one of these cheep ink jets which just get financed by selling ink... It is (was) a hp photosmart 6510 scanner/printer combination. I salvaged quite some nice parts including a pump etc.

So the scanner part came off nicely and just for the fun I hacked a bit around. The motor part is easy to do, just a DC motor and an optical encoder. Just one thing wonders me: Why is there a quadrature encoder, I mean the printer (or myself) saying which direction the motor runs, why do I need this? Getting the speed maybe?

So now for the CIS module/line: I know there will be a VCC, GND and (guessed) RGB LEDs. And a Clock and maybe Line Clock and three Analogue Outputs which gives the RGB from the pixels which are read advanced with the "pixelclock" .

I will attach some photos maybe you can help me identify the module, the part numbers are all leading to HP replacement parts not very helpfull...

I can maybe check the RGB LEDs with a multimeter/diode test, but other than that I am lost. Oszi is avaible. No I cant check on the workin scanner, already tossed the remaining parts into the dumpster...

Cheers,
Carsten

[amyk]

The encoder is to precisely measure the position of the scanner on the page.

You may be able to find a service manual for it which has the relevant pinouts/schematic.

[calli]

The encoder is to precisely measure the position of the scanner on the page.

Yea, but I know if the motor turns forward OR backward, so I would say I only need just a pulse and not a quadrature with two phased pulses?

For the service manual, I guess these $60 printers are not supposed to be repaired :-/ so maybe thats not findable.

Cheers,
Carsten

[calli]

Edited my first post here to actually show BOTH images and not one two times   

Carsten

[tatus1969]

The encoder is to precisely measure the position of the scanner on the page.

Yea, but I know if the motor turns forward OR backward, so I would say I only need just a pulse and not a quadrature with two phased pulses?

For the service manual, I guess these $60 printers are not supposed to be repaired :-/ so maybe thats not findable.

Cheers,
Carsten

they use brushed dc motors to replace steppers. the right way to do this  is to run them as a closed-loop servo system with positional control. this requires the quadrature encoder. doing this gives superior dynamic behavior over the traditional stepper solution, and it is more cost effective because the motor can be much smaller.

[bktemp]

CIS sensor modules are nice. Somewhere I have a module connected to an AVR streaming the image data over uart.
If you still have any other parts of the scanner the module was connected to, try to power it up again to identify most of the signals.
The one I have used 3 signals: Start, clock, output. It is only a monochrome sensor, because the RGB LED is used to scan R, G and B sequentially.
Reading the signal is simple:
Set start high, apply a single clock cycle, set start low again. Now every clock pulse outputs the next pixel.
The integrating of the next image data starts after each start pulse. Maybe there is an additional clear or integration start signal, but most sensors handle this internally.

[calli]

they use brushed dc motors to replace steppers. the right way to do this  is to run them as a closed-loop servo system with positional control. this requires the quadrature encoder. doing this gives superior dynamic behavior over the traditional stepper solution, and it is more cost effective because the motor can be much smaller.

Ah, ok that triggers something in my brain. I will remember this when and if I one time get that thing to run (and scan something).

Thanks!
Carsten

[calli]

The one I have used 3 signals: Start, clock, output. It is only a monochrome sensor, because the RGB LED is used to scan R, G and B sequentially.

Ah yes, that makes sense! I did let me fool by an article which got this RGB thing wrong or talked about a different kind of sensor.

Thanks!
Carsten

[calli]

So I think the RGB-LEDs are not controlled directly but by a logic signal from the "motherboard" through the flat cable and the there seems to be a chip which turns on the LED. There is no direct connection from the motherboard through the flat cable to the LEDs.

At least I could light up all LEDs (and did burn the RED one nearly because of having too much voltage and bypassing the resistor).

Carsten