Daniel Reetz, the founder of the DIY Book Scanner community, has recently started making videos of prototyping and shop tips. If you are tinkering with a book scanner (or any other project) in your home shop, these tips will come in handy. https://www.youtube.com/channel/UCn0gq8 ... g_8K1nfInQ

Archivist Quill Kit Plans vs. Archivist Book Scanner Plans?

Built a scanner? Started to build a scanner? Record your progress here. Doesn't need to be a whole scanner - triggers and other parts are fine. Commercial scanners are fine too.
Doranwen
Posts: 45
Joined: 13 Aug 2014, 04:19
E-book readers owned: Onyx Boox (dead), Astak EZ Reader (dead)
Number of books owned: 2662
Country: USA

Re: Archivist Quill Kit Plans vs. Archivist Book Scanner Plans?

Post by Doranwen » 20 Apr 2021, 21:04

I pointed out the issues with dust and scratches on acrylic platens to my brother, and he agreed that glass was probably the better choice. His main reluctance to use glass was that he has no idea how one can affix them to the scanner when following the plans that we're looking at (the PVC pipe style). It involved drilling the acrylic to use a T-bracket to hold it, and drilling glass is likely to end up in cracked and shattered glass, he thinks. So he's still stumped on that one.

Home Depot actually has panes of 11"x14" glass for very cheap - less than $4 per pane (compared to the $12 per pane that Michael's quoted me). The difference between theirs and Michael's is in the thickness, though. Michael's had 1/8" thick, which came to about 3.175 mm - slightly thicker than the 3 mm recommended. Home Depot, on the other hand, has 3/32" thick, or approximately 2.5 mm. Will it likely be a problem using glass that thin? I can see the lower weight being beneficial as the counterbalance was designed for the acrylic (though with the older cameras, so I really have NO idea how the weight is going to balance out in the long run).

I picked up three of those because it was only about $10, and that's pretty cheap to play around with and see if we can figure out how to fasten them (my brother figured I ought to get one just to experiment on, lol). But that really is the question: How do you fasten a pane of glass from the *top* only? Clamps? Is there a way to drill through glass and have it actually be supported well from that thin bit? Some super strong adhesive? My brother said he'll ask some construction guys he knows if they have any ideas, but that's the real challenge that we can think of.

cday
Posts: 300
Joined: 19 Mar 2013, 14:55
Number of books owned: 0
Country: UK

Re: Archivist Quill Kit Plans vs. Archivist Book Scanner Plans?

Post by cday » 21 Apr 2021, 09:12

@dpc

100º Platen_3.png
100º Platen_3.png (7.83 KiB) Viewed 722 times

I may have a particular ability to identify possible ambiguity, could you please confirm whether you have used epoxy to fill the space where the two sheets in a platen like the one shown above join?

dpc wrote:
20 Apr 2021, 17:53
I haven't cut acrylic with a circular saw, but if I did, I'd look for a blade made for cutting laminate/plastic which is going to have a lot of teeth and a narrow kerf.

While acrylic is soft compared to laminate, a laminate blade might be best if it is necessary to buy a finer blade, but I think the blade on my circular saw which has fairly high number of teeth would likely also work.

You'll definitely want to clamp a straight edge to the piece for the saw to cut a straight line. You might have to sandwich the acrylic with some plywood/mdf to prevent tear out chipping. You need to be careful that you don't scratch the acrylic when pushing a saw across the surface. Keep the protective paper on the acrylic until you are finished cutting/drilling on it. Double-check that the bottom of the saw doesn't have any metal burrs that could scratch through that protective paper layer.
All that was implied in my suggestion, if possibly not made sufficiently clear. I envisaged placing the acrylic sheet on a clean workbench or other flat surface, with just the minimum acrylic sheet extending over the edge so as to minimise possible flexing. Then clamping a thick board above the sheet, on which the circular saw base would slide, guided by a straight edge clamped or screwed to the thick board. So the base of the saw wouldn't run across the surface of the sheet.

If you're going to cut acrylic with a saw that allows you to tilt the blade and make mitered cuts, you should just cut both panels with 40 deg edges and glue the mitered joint with acrylic cement. That's going to make a stronger joint than leaving the panes with a 90 deg edge and filling the gap with epoxy (and/or an acrylic rod). I don't know how well epoxy adheres to acrylic. Some experimentation may be necessary. The epoxy that I've used on glass platens was the type used for auto windshield chip repair and is specifically designed for adhering to glass, filling a void, and is clear.
Unfortunately I have spotted a practical issue: I had been envisaging the circular saw blade being inclined only slightly, say 10º, from the vertical, but thinking about it now I have realised that with the above setup the blade would actually need to be inclined at 50º from the vertical... That would be at the limit or possibly beyond the angle at which the saw could be set, and probably as important would probably not be an angle at which a satisfactory cut could be made in practice with the above setup.

That issue could be overcome by constructing a structure that held the acrylic sheet at an angle, such as 50º, allowing the blade to be set at 0º, or at any rate a small angle from the vertical, while the saw was run across a horizontal platform above it, but constructing such a support although probably sound in principle would be a significant extra complication. Maybe you would like to go into production producing beveled acrylic sheets?

dpc
Posts: 347
Joined: 01 Apr 2011, 18:05
Number of books owned: 0
Location: Issaquah, WA

Re: Archivist Quill Kit Plans vs. Archivist Book Scanner Plans?

Post by dpc » 21 Apr 2021, 12:15

cday,
Yes, that's the gap that I filled in a platen comprised of two glass panes. I wouldn't do that with acrylic though.

You need to set the saw cut to 40 degrees, not 50. It's the opposite of what you think it is. The smaller the angle, the more "open" the platen V will be when the two cut pieces are mated together.

Think about it this way, if you set the saw angle to 0 degrees and mated the two pieces together you'd have a 180 degree platen. If you set it to 45 degrees (the max angle most saws will cut) you'll end up with a 90 degree platen when the two pieces are mated. So you want something between the two. That "something" is a 40 degree angled cut which will result in a 100 degree mitered joint.

Doranwen,
If you decide to go with glass panes for your platen, you'll need to change the design and from what you've hinted of the tools, experience, and time that you have available it might be better if you stuck with acrylic for this first scanner. That being said, here's some thoughts on using glass in a Landin-type scanner. The glass will need to be supported with a V-shaped framework (typically made of wood) that positions the planes to 100 degrees and has a mount on either side for your lifting arms. Attempting to attach those lifting arms directly to the glass with some sort of brackets (like you do when using acrylic) is not going to work with the thin glass you're using.

The glass can be attached to the frame using aluminum angle that screws into the frame and then the bottom portion of the angle reaches under the glass pane and holds it up against the frame (The Archivist used this method, I think?). You can also glue the glass to the frame using silicone sealant. When using the scanner the force on the glass pushes it up onto the frame so whatever you use to hold the glass against the framework only needs to overcome gravity. You could probably use some strips of double-stick tape (the high-strength 3M stuff) and that would work. Note that using the aluminum angle method would allow you to change the panes more easily than having to cut through glue.

The Internet Archive's early scanners from a decade ago (before they started using the Archivist) are similar to the Landin design. They of course use glass panes in their platen. There are videos on youtube where you can see the mounts on the side of the framework and get an idea of how they are supporting the glass. Here's a small photo: https://www.digitalnc.org/wp-content/up ... 80x300.jpg. The lifting arms are connected to a trundle and you raise/lower the platen using your foot.

Finally, don't worry about overdesigning the counterbalance. You can use a plastic water bottle and fill it with sand or even water to get the exact weight you need. A wide-mouthed Nalgene bottle works well for this because the plastic strip that holds the lid to the bottle can wrap around the plastic pipe then screw the lid on and you don't need to tie/tape it to the pipe.

cday
Posts: 300
Joined: 19 Mar 2013, 14:55
Number of books owned: 0
Country: UK

Re: Archivist Quill Kit Plans vs. Archivist Book Scanner Plans?

Post by cday » 21 Apr 2021, 13:26

dpc wrote:
21 Apr 2021, 12:15
You need to set the saw cut to 40 degrees, not 50. It's the opposite of what you think it is. The smaller the angle, the more "open" the platen V will be when the two cut pieces are mated together.

Think about it this way, if you set the saw angle to 0 degrees and mated the two pieces together you'd have a 180 degree platen. If you set it to 45 degrees (the max angle most saws will cut) you'll end up with a 90 degree platen when the two pieces are mated. So you want something between the two. That "something" is a 40 degree angled cut which will result in a 100 degree mitered joint.
Thank you, I follow your clear explanation!

Cutting at 40º with the acrylic sheet simply clamped to a flat surface would certainly be a bit easier than cutting at a greater angle, but I'm not sure how well it work in practice. I had imagined the saw blade titled towards the clamping board, but if it were tilted away that would assist in clamping the sheet close to the edge without interference with the saw blade. Some more thought and a sketch would be useful if that route is shortlisted.

I have established that epoxy can be a satisfactory adhesive when used with acrylic, and of course it should have good gap-filling properties. On the other hand solvent welds are said to be quite weak as David Landin said, and gap-filling is very limited.

I have also had a further insight into possibly filling the large open 'v' when sheets are position as in my drawing immediately above: I checked whether square section acrylic rods are available and was surprised that they are readily available on eBay and are inexpensive:


Square Acrylic Rod.png

For reference £2.15 is currently about $3. The 3 mm square size is fractionally smaller than 1/8" so should fit into the gap at the bottom of the 'v' with a little space for some epoxy, one edge of either the rod or one sheet would probably need to be trimmed very slightly either by abrasion or shaving with a sharp blade, or maybe a sharp plane? I'm not sure if 1/8" square might be available in the U.S. but if both are available the 3 mm size would probably be the better anyway?

If consideration of using glass for the platen continues, I have established that drilling glass shouldn't be too difficult, using either special drills for glass or the drills used for drilling ceramic tiles. Glass can alternatively be bonded securely to metal, so it is said, I'm not sure about well that would work in practice, or in principle, of course, clamped securely, although suitable presumably spring-loaded clamps might not be easy to source or devise.

cday
Posts: 300
Joined: 19 Mar 2013, 14:55
Number of books owned: 0
Country: UK

Re: Archivist Quill Kit Plans vs. Archivist Book Scanner Plans?

Post by cday » 22 Apr 2021, 07:15

Although the way ahead is not well defined at the moment and we are still in brain-storming mode, I have had two further thoughts on my last suggestion above.

1. Considering the possible acrylic construction option illustrated in my last 100º Platen_3.png image above, the edges of both sheets could probably easily be cut with a circular saw as described above, at a 10º (dpc?) angle so as to create a rectangular space where the sheets come together, into which a rectangular bar could be cemented. Those circular saw cuts could be easier to make than cuts at a 40º angle.

2. In principle a rectangular section of the same width as the acrylic used could be cut from the same acrylic sheet using a circular saw, avoiding the need to locate and purchase a suitable section to fill the space. Not sure how well that would work in practice, though...

Finally, if any of the above options that involve profiling acrylic sheets are planned, it would probably be as well to purchase the sheets slightly oversize in the relevant dimension to allow for more than one attempt if necessary, on the basis that any surplus width could be cut off later.

Doranwen
Posts: 45
Joined: 13 Aug 2014, 04:19
E-book readers owned: Onyx Boox (dead), Astak EZ Reader (dead)
Number of books owned: 2662
Country: USA

Re: Archivist Quill Kit Plans vs. Archivist Book Scanner Plans?

Post by Doranwen » 22 Apr 2021, 20:13

My brother confirms that glass can be drilled (he asked someone) and there's a glass shop only a minute from my house that I can ask if they can do the drilling. But whether that's too much weight to hold at that angle or not, that's something I'm not as sure about.

I like the V-frame idea but how to create that… I'll point my brother to read this again, as I think he's back from the trip he was on.

cday
Posts: 300
Joined: 19 Mar 2013, 14:55
Number of books owned: 0
Country: UK

Re: Archivist Quill Kit Plans vs. Archivist Book Scanner Plans?

Post by cday » 23 Apr 2021, 14:29

Confirmation that the edge of an acrylic sheet can easily be profiled at a slight angle using a circular saw... :D

A test cut on a spare piece of 3 mm acrylic using a circular saw fitted with a fairly new TCT blade with 48 teeth, with the saw blade angled at 10º to the normal vertical cutting angle, produced an excellent clean cut edge. The blade was a standard 190mm (7.5") diameter blade with a tooth width of 2.6mm (0.10"). A blade with fewer teeth might also produce an acceptable result, a trial cut would be necessary to check, as might a blade designed for cutting laminate.

A circular saw blade cuts at the leading edge with a rising action, so eye protection is particularly advisable when cutting a material such as acrylic due to the possibility of small and possibly sharp chips. I laid the acrylic sheet on a clean board serving as a workbench, and wishing to minimise flexing of the acrylic sheet as the blade rose into it, and with the possibility of chips in mind, clamped a piece of thick board over the sheet with the edge of the board extending right to the edge of the acrylic sheet, so there was no possibility of the acrylic sheet flexing upwards while being cut.

To make it easier to move the sheet and board together while setting the cut, I used some small pieces of doubled-sided tape to tack the sheet to the underside of the board. When positioning the sheet on the workbench, it is necessary to allow a little more overhang than would normally be needed due to the circular saw blade being angled slightly towards the worktop edge. Better to anticipate that.

On top of the thick board I screwed a strip of planed timber to act as a straight edge to guide the saw while making the cut. When cutting small pieces of board clamping everything in place can be sometimes be difficult even with a good supply of clamps, and in this case screwing the straight edge to the board provided the easiest solution. As extending the edge of the thick board to the edge of the acrylic sheet results in the board being trimmed slightly, profiling a second sheet requires moving the straight edge slightly away from the edge. I don't know how much difference to the cut extending the thick board to the edge of the acrylic edge actually makes, but it seems worthwhile to ensure a good result.

The acrylic sheet would normally, of course, be profiled with the protective film still in place, and if necessary additional protection such as a sheet of office paper could be used, although in my test using a spare piece of acrylic the sheet wasn't protected by the original film.

The result of my test was such that, with care and the saw set at the 0º rather than 10º, I think it might actually be possible to cut a square rod section with a side close to the sheet thickness, as illustrated in my previous post above, as a possible way to fill the resulting inverted 'v' space.

However, with the option of profiling the edge of acrylic sheets now established, the optimum solution would seem to be to butt the sheets together as in the David Landin design, but at 100º.


Platten 100º_5.png

The left-most sketch is David Landin's design for a 90º platen angle with the sheets butting neatly together, the middle sketch recalls the result of butting normal square-edge sheets at 100º with a small space to be filled and a potentially small sharp edge, and in the right-most sketch the feint blue lines are the sheet edges that have been profiled by 10º, butting together flush as in the David Landin design. Butting the sheets together neatly should also slightly simplify platen assembly if triangular blocks with 100º inclusive angle are used.

If acrylic sheet is used my suggestion would therefore be to profile the sheet edges as described, and assemble the platen using either an internal triangular block at each end, or if preferred, mountings extending outside the sheets. It should actually be possible to use 3D printing to make very neat one-piece plastic external platen end plates with raised guides for the acrylic sheets if that were available, I'm just thinking aloud!

If glass sheets are still considered an option, in addition to the need to drill holes for the support T-bracket, you might inquire if your local glass supplier is able to profile one edge of each piece at a 10º angle, or whether they can suggest someone who could do so at a reasonable price. The amount of glass to be removed would be very small.

cday
Posts: 300
Joined: 19 Mar 2013, 14:55
Number of books owned: 0
Country: UK

Re: Archivist Quill Kit Plans vs. Archivist Book Scanner Plans?

Post by cday » 29 Apr 2021, 10:01

David Landin's Instructables article (revised in April 2015) evidently doesn't represent his final thoughts on the design of his scanner.

In the reader's comments at the end, I noticed a reference to his use of a 110º [not 100º] platen angle, and decided to try to find a reference to that in his forum posts, and found it in this 2016 post:
davidlandin wrote:
11 Sep 2016, 14:23
3. Platen Although it is easier to make a 90 degree platen, there are some problems with 90 degrees. Most important of these is that it is a bit difficult to avoid reflections from the other side of the platen and also to take images that go right into the gutter of the book.

So I have started using a 110 degree platen.

It would seem that even using anti-reflective acrylic didn't completely overcome the problem with reflections, so I think you can finally eliminate that option.

He also provided the following detail:
I asked the supplier of the acrylic to cut the long edges at 55 degrees, so that when they are butted together they form a total internal angle of 110 degrees.
I don't know how widely available that option is likely to be, but it certainly might be worth asking if your acrylic supplier is able to profille one edge of each sheet, to save the need to do it yourself and to guarantee a good result. If that option is available to you, I would suggest being very careful in defining the cut angle you need to produce a 100º platen angle, and then provide the acrylic supplier with a sketch to avoid any possible misunderstanding. Preferably as a final check run your calculation by dpc!

David Landin's use of a 55º profile angle implies that rather than butt the two acrylic sheets together, as in his orignal 90º design and as I indicated in the sketch in my immediately previous post, he assembled them in a symmetrical 'v' design, as proposed as an option in one of my earlier diagrams. The method used of course has a big effect on the profile cut angle: if it is necessary to use a circular saw I think the 10º profile angle I tested could be easier to make, although the larger angle might not prove a significant problem in practice.

I don't intend to read all of David Landin's now very long thread, or to search for anything else relevant that he may have written, but given that the Instructables article didn't represent his final thoughts, it might be worth your having at least a quick look through everything that is available.

Edit:

Another David Landin post giving a little more background.

It would seem that a number of changes were made after the updated 2015 Instructables article, some possibly requiring a slight change in the tube lengths. In that respect, the Schedule 40 tubing is described somewhere as being a close fit in the couplings used, so if possible it might be best to avoid cementing parts together any earlier than necessary, and possibly cutting them slightly oversize as shortening if necessary is easier than replacing them. The slight increase in tube diameter could also require a slight increase in tube lengths if more tube is inserted in the couplings.

I also read somewhere a good suggestion that had already occurred to me as an alternative to cementing parts together, to drill a small hole and insert a small screw, which could easily be removed later if the tube needed to be shortened or replaced.

dpc
Posts: 347
Joined: 01 Apr 2011, 18:05
Number of books owned: 0
Location: Issaquah, WA

Re: Archivist Quill Kit Plans vs. Archivist Book Scanner Plans?

Post by dpc » 29 Apr 2021, 13:47

I don't know what the reason is behind Landin choosing a 110 degree platen angle over the more common 100 degrees. Perhaps it was a requirement that the cameras needed to be mounted closer to the platen with his plastic tubing arm design and that would require an increased platen angle to prevent a reflection off of the adjacent platen pane. And/or it could be that his scanner was built to support larger page sizes and with the increased Field of View angle the platen angle needed to be increased?

In any case, as the platen angle is increased, the light position will also need to be moved higher to prevent a reflection of the light source off of the platen surface and into the camera. Sometimes it's easier when you're designing a scanner to mount the cameras in a convenient location for the page size, then play around with the platen angle and light height until you don't see any reflections in your images. The downside of moving the light farther away from the platen is the light's intensity drops off at a 1/d^2 (one over the square of the distance) rate and may require brighter/additional lights to adequately illuminate the page.

I've also found that camera position can be key in eliminating zoom lens aberrations like pin cushion and barrel distortion. Some of this can be dealt with in a post-processing step, but I prefer to capture as clean an image as possible so that sort of thing isn't necessary.

Doranwen
Posts: 45
Joined: 13 Aug 2014, 04:19
E-book readers owned: Onyx Boox (dead), Astak EZ Reader (dead)
Number of books owned: 2662
Country: USA

Re: Archivist Quill Kit Plans vs. Archivist Book Scanner Plans?

Post by Doranwen » 29 Apr 2021, 14:17

My brother says at this point between the videos, diagrams, and everything he's looked at, he's pretty much got nearly all of it worked out. (He's actually got a lot of construction experience with saws and tools, and is very good at math and problem solving, so he's feeling pretty confident at this point.) Thank you all for all the brainstorming ideas and everything! The suggestions have been very helpful and helped him think through a few potential problems.

The one thing he's still contemplating is what one can put between the glass panes (we are going with glass for sure, and he's got the platen design worked out now, he says) so he can butt them against each other in a way that would be removable so that one could change out a pane if necessary at some point.

And I need to actually sit down and get the Open Camera app installed on the phones (it's been a hectic week or two) but that's on my priority list for this weekend if I don't get to it sooner. Hopefully the 'net will work OK when I do that…

Post Reply