Expensive but worth it?

[DanBarlow]

Dan, 4 diodes = 4 * 172 = 688 lines. The line scanner is the polygon and the page scanner is a simple galvo.

Where you get the frame buffer is the fun one... The guts of the copier perhaps? Some VGA cards were quite programmable.

RAM is cheap these days, if you just wanted a still picture...
Steve

I've been working on a scan mangler based on surplus x86 cache chips. NTSC in, whatever you want out. Looks like it would come in around $200. Since I now have AOD driver to test it with, might move up in the project queue. I also have an evil hack for the polygon to multiply the lines per facet, but it is tweaky and so I'm not sure it will be worthwhile for anyone else. Sometimes an image defect looks like a special effect... and the jet engine sounds of the mirror spinning up are cool.

[300EVIL]

Dan, 4 diodes = 4 * 172 = 688 lines.

That's exactly how they do it in the copier... The wavelength is only good for writing on the photoreceptor though...



Steve,
PM me your address again and I'll get you one of these scanners.
Adam

[catalanjo]

Shit ... loads of things have been going on in this thread while I was mooning about at Plaza here in London.

As I understand it, the evil one has a bunch of polygon motors (with drives?) and all of a sudden, after a slight plumbing detour, everyone here is discussing the resolution of a hypothetical TV set.
Low resolution is GOOD!
Flicker is GOOD
I'll have to blurt this given my hinibriyated condition.

IMHO screw the resolution just make it look nice and funky !

I'll drink to that !
Cheers

[DMills]

Works, but the alignment pain.....

Try this variant:

Take a single source and pass it through an AOM fed with FM so that the first order beam deflection angle is time varying over a small amount, then bounce this beam off the polygon scanner.

If the drive to the FM modulator is sinusoidal then each 'line' scanned out by the polygon will consist of a sine trace where the peak to peak amplitude should match the width of the mechanically scanned line.

The Video used to modulate the thing would be read out of the buffer in a suitable pattern to match the spatial beam position.

Incidentally have a look at some of the software defined radio output boards for scary quick D/A, should be well up to this, or just use some dram, flash converters and a couple of PLLs (The refresh will be taken care of by the read cycles so no worries there), page flip in the output side VBI.

Regards, Dan.

[300EVIL]

Here are those scanners I was talking about...











The pinout is labeled but it's under a piece of the frame. I can clearly make out 24V and ground pins but the others control the speed. I'll probably have to cut off the arm to look and see what they say without completely destroying the case.

[DanBarlow]

Here are those scanners I was talking about...

The pinout is labeled but it's under a piece of the frame. I can clearly make out 24V and ground pins but the others control the speed. I'll probably have to cut off the arm to look and see what they say without completely destroying the case.

Link to datasheet

I think it's related to the ones I have from color laser printers. They accept a frequency to lock to, up to 10KHz, and a "go" signal. Output is a "PLL Lock" signal.



This one is much better heat-sinked. Can I get one?

[300EVIL]

Link to datasheet

I think it's related to the ones I have from color laser printers. They accept a frequency to lock to, up to 10KHz, and a "go" signal. Output is a "PLL Lock" signal.



This one is much better heat-sinked. Can I get one?

UH DUH!! Yeah, I guess looking at the driver datasheet would make sense! Thanks man!

Not only heatsinked but ceramic air bearinged as well!!! These come out of $50K copiers so I'd imagine the quality to be high. Tell ya what... If you promise to do something with it and post your results, I'll send you one for the cost of shipping.
Deal?
Adam