Self-Adjusting Spine Support

[jck57]

Another idea for a self-adjusting spine. A slippery fabric like satin is placed under the book. Each side of the satin is clamped to the carriage that holds the cradle. a rubbery sheet is placed between the book and the satin. As the pages are turned, the misalignment of the platen to the book forces the cradle over. As the cradle moves sideways, the satin fabric slips in relation to the cradle, but because of the rubbery sheet, the book stays with the satin. This means the book spine moves in relation to the cradle, thus keeping the open book always in alignment with the vee of the platen.

spinesupport2.jpg (23.22 KiB) Viewed 8917 times

EDIT:

After mocking up a model of this design I discovered it fails because the fabric slides the book an inch for every inch of lateral movement of the cradle. To work, the fabric would need to slide the book 1.41 inches for every inch of lateral movement of the cradle.

[freemab]

Use wooden dowels. For sturdy, modern volumes only, especially bound periodicals. Such volumes have a space between the sewn, bound pages and the spine of the book cover. It's as if the book spine were hollow. Obviously, this does not apply to all, or even to most, books.

But when you have such volumes to scan, you can slide a dowel into the hollow spine to open the pages almost flat by flexing the spine into a gentle arc. You will need a range of dowel diameters, and you choose the dowel size to best open the pages evenly. This is so effective that such volumes can be scanned easily and accurately on a flat-bed scanner, such as a Xerox machine. This is a standard technique in libraries with many such bound periodicals.

For some books without hollow spines, such as some hardcover books and some thick perfect-bound softcovers, you can place a dowel below the spine on a scanner that scans from above. This has a similar effect as placing the dowel in the hollow spine, flexing the book gently into an arc that can better be photographed.

[daniel_reetz]

That's true, but what I learned is that the dowel has to be able to shift left and right as you traverse the book. I can make a more detailed description/video of the full problem in a week.

[jck57]

selfspine3.1.jpg (71.6 KiB) Viewed 8730 times

Yet another self adjusting spine idea. There is no direct mechanical link between the slides. The red line represents a rubbery sheet that prevents undue stress between book endboards and the spine. The rubbery sheet also helps hold the book in place in relation to the slides. Springs pull up on each slide to help balance the book at every position it is open. Otherwise, on very thick books, it would be difficult for the platen to recenter the book.

EDIT:

I fixed the picture to show the springs and a way to adjust tension. I have a better scheme for this but hey I'm using Paint. I really like this concept now. You adjust spring tension so that when all the pages are one side the book "balances." Then you set the other side for the same tension. As the extension spring is loaded by more pages, the tension increases so that for every book opening the book is "balanced" side to side.

[jck57]

Previous post edited.

[sethg]

I've been thinking about this, but it's taking me so much longer than I originally planned to do a regular build that I probably won't be actually trying anything for a while. Still, I like knocking ideas around.

This is what I'm thinking: I like the simplicity of some of these designs. My original plan was way too complicated and probably wouldn't have worked that well. However, I wonder if this whole issue stems more from the actuall support of the spine than from the need to adjust the cover. Here's a crude drawing:

BookSupport.png (24.66 KiB) Viewed 8634 times

I think any method that does not push the spine up into the v of the platen will not get it quite right. Now, how to do that? I'm still working on that part.

I wonder if a wire could be attached to the frame in which the cradle slides, so that it is always centered to the v of the platen. It could have an adjustable height to accommodate books of different sizes. Then you could use the sliding rubber thing from jck57's post.

[dpc]

I've also been thinking about this a bit more and have come up with a far simpler design than the one that posted earlier in this thread. I'm still hashing out the details and would like to make a prototype of part of it before posting here to ensure that it will work. It's difficult to test though without a platen "wedge" to press down to force the book to level and center itself in the cradle. I need to think about a test setup too.

I've been wondering about some of the design constraints. When does the thickness of the book necessitate some sort of leveling mechanism? I assume that thin books (< 1.00" or so) will center themselves enough to scan properly. Also, what is the maximum book thickness that the self-leveling mechanism should support?

[sethg]

I would think that you can make it for any size of thickness that you want. If you look higher on this thread you can see that books have been made that are pretty stupidly thick. For my purposes, I'm using an oversized build that can accommodate books up to 8 inches thick, but even with the old books I'll be scanning the vast majority will fit under 6, and most are probably closer to 3 or 4. It's really a matter of where it will be used and what are the acceptable limitations. A little bit of research tells me that most commercial scanners won't do more than a couple inches.

[daniel_reetz]

my experience has been that most books are smaller than 3" across. I can't remember exactly, but the new CNC scanner supports up to 4 inches, I think.

[dpc]

The amount of horizontal travel required by the cradle carriage is greater than the thickness of the largest book you plan to scan, and can be determined using the following equation:

Cradle shift = 2.0 * book_thickness * sin(platen_angle / 2.0)

So if you're using a 100 deg. platen angle and you plan to be able to scan books that are 3" thick, the cradle must be able to shift horizontally by at least 4.6" (2.3" either side of the center of the platen) during the course of the scan.