Scanning based 3D printer

Talk on the diy_3d_printing_and_fabrication Yahoo group has turned recently towards the possibility of using the polygon scanning mirror and lenses from a laser printer to create a resin based 3D printer for large areas. The scanner scans a line with a laser and then you move the scanner to "develop" a layer, then repeat. Basically the same idea as a SLA or DLP based resin projector just a different method of imaging. A few guys have done this sort of thing to try and image PCBs directly in photoresist. The visible light curing resins that can be used in 3D printers may however work a lot better than the etch resist.

I have a Samsung laser printer with a perpetual paper pick up problem so decided to take it to bits to investigate. It is a pleasingly simple module, pictures now follow.

The scanning module as it is found in the printer, very possibly this is a part obtained from a 3rd party. It is a really neat sealed unit just waiting to be mounted into a machine.

The other side

At the bottom left we have the spinning mirror, at the top is the weird looking lens that corrects for the variation in distance from the mirror to the drum in the printer, probably does some focussing too, not sure.

This is a brushless motor controller by Rohm especially designed for polygon scanners. The control is very simple from what I have managed to work out, there is a clock to set the speed, a start/stop input and an output to say then the PLL has a lock (i.e. the speed is constant). Then power which is 24v (up to 36 but I suspect 24).

The optics consist of a collimator, a metal aperture and then a focussing lens with a long focal length. The mirror is very thin so perhaps the aperture stops stray reflections from other parts of the motor.

The collimator

The laser diode, a simple uncollimated source. At the very top you can also see a photodiode which is used to sync the control signals to the diode with the rotation of the mirror. The beam hits a small mirror at the other end of the correction lens and then shoots over to this sensor.

Should be quite easy to get a basic system up and running, I'll probably use the propeller microcontroller which should eat the task up but I will need a UV laser to do some real tests, not to mention some resin. And I have the Zcorp to finish! Argh, maybe tomorrow!