Self-Adjusting Spine Support

[dpc]

Thanks for the posting the picture of your scanner - that helps. (Good work, BTW!)

If you open that book that is pictured in your scanner to the exact middle of the book, then ensure that the spine is supported against the flat cradle bottom, can you then get the platen to go down into the gutter, flattening the pages, and get a quality shot? I'm wondering how much of the problem you're seeing is because the spine isn't supported at the correct angle to ensure that the platen apex can press into the gutter evenly between the two pages?

From just looking at that book in the cradle, it would seem like it would press against the platen sides as you forced the book into the platen because the pages are a bit wrinkly, making the stack of pages on the left and right of the platen rather "spongey".

You have different design requirements compared to most folks who are building scanners. The thick old warped, wrinkly books with notes near the gutter may require something special in the cradle/platen design to get you the images that you need. What ends up working for these type of books may very well be overkill, or even non-optimal for scanning books that aren't in this same condition. For example, those rollers would work OK for hard-backed books, but something with a cover that is more flexible would bend outward over the roller and likely not contact the platen near the outer edge of the page (most of my software reference books are this style and I wouldn't want to be forced to resort to scanning in paperback mode for these). This discussion reminds me that I need to make a note to swing by the local public library and look for some thick (3"-4") books to try a few ideas I've had for a self-leveling cradle.

If you absolutely want to get in tight to the book's gutter, a single page scanner will do it. You can make the platen angle less than 90 deg (how much depends on your camera lens viewing angle) and that will make the platen "knife" into the gutter a bit tighter. 1600 pages would be a bear to scan one at a time though. Just thought I'd throw the idea out there if you can't find any other way to get the results you require. Or maybe that's an option for a few of the more troublesome pages?

BTW, how do you like the T2i cameras for scanning? I'm considering moving to these eventually because of the lens quality, remote software, and absolute control that they offer over the digital compact cameras. Any complaints other than having to do intermediate file transfers due to the large size of the RAW images?

[jck57]

Working prototype of a spine tilting cradle. Because the ropes run underneath the rolling tables, the distance between the endboards of the book is less when the middle of the book is scanned than when all the pages are on one side. This causes the spine to arch up when scanning the middle portion of the book. It's all about the plywood thickness of the tables.

[rob]

Wow, that is really impressive, and it seems simple enough. I did notice that the location of the gutter moves, which means that we'd keep the cradle able to move horizontally (since the platen is fixed), and sort of add this mechanism on top. Some easy way, maybe adjustment knobs, could be used to adjust the separation for the overall thickness of the book, along with adjusting the tension of the string(s). The whole thing would also have to be as light as possible, so that it is relatively easy to remove the cradle, place the book on it, adjust the thing, and then place the cradle back in the scanner...

[jck57]

Wow, that is really impressive, and it seems simple enough. I did notice that the location of the gutter moves, which means that we'd keep the cradle able to move horizontally (since the platen is fixed), and sort of add this mechanism on top.

Yes. By simply adding four skate bearings to this cradle it should replace the existing cradle.

Some easy way, maybe adjustment knobs, could be used to adjust the separation for the overall thickness of the book, along with adjusting the tension of the string(s).

I'll post pics of the adjustment mechanism for the springs. I think a single strap near the center with a simple spring clamp should efficiently replace the ropes.

The whole thing would also have to be as light as possible, so that it is relatively easy to remove the cradle, place the book on it, adjust the thing, and then place the cradle back in the scanner...

Good point. I think many of the pieces including the rolling tables could be made from thin plywood. Also, there is a need for easily adjustable clamps for the end boards.

[daniel_reetz]

Whoa. I actually gasped out loud when you flipped the book open and it self-centered.

[jck57]

I'd like to try to design one that you guys could CNC out of 3/4 ply. Like the rest of the Hackerspace scanner it should be easy to assemble and use common hardware. I have some ideas but I need dimensions for the original cradle and carriage. Can these numbers be taken off the CNC files or will somebody have to help me out here?

I want to follow Rob's suggestion and have one knob for adjusting gap and one knob for adjusting both springs. I'm anxious to apply some new ideas and make a prototype for CNC cutting.

For starters, what are the dimensions for the cradle's plywood base to which the skate bearings are attached?

[daniel_reetz]

Do you have/like any CAD software? I did the original drawing in CorelDraw, and Rob has converted it to Adobe Illustrator. If you bring the drawings into those packages, the dimensions will be apparent. Otherwise, I can annotate the complete cradle for you.

[daniel_reetz]

Here's a quick shot at an annotated drawing showing dimensions.

[jck57]

Do you have/like any CAD software? I did the original drawing in CorelDraw, and Rob has converted it to Adobe Illustrator. If you bring the drawings into those packages, the dimensions will be apparent. Otherwise, I can annotate the complete cradle for you.

Used to have AutoCad but nothing on this new computer. There is a free program that will scale any screenshot if you give it a single known distance. All I really need for now are the outside dimensions of the cradle base, por favor.

EDIT: Ah, there it is. Thanks, Dan.

[daniel_reetz]

No problem and please let me know if I can give you anything else. The dots in that image are 1/4" spacing.