Category Archives: Saws

Bookshelf: The 60-Year Dovetail

Posted on by
Reply

I’m slowly making my way through the prototype bookshelf project. After realizing that my widths were all wrong, I had to spend some quality time with my rip saws and jointer plane to fix all of it, and I was able to do that during what little time I had this weekend.

So then I decided to get started on the dovetail joints for the top, and let’s just say that my dovetails are not the famous five-minute variety. I don’t get a lot of uninterrupted time in the shop these days, so I just cut whatever I can, and stop when I need to. On Monday, I laid out the tails and cut them. Yesterday, I only had time to finish cleaning the waste between the tails and mark out the pins. This morning, I had enough time to cut the pins and finish the joint (and mark out the tails for the next joint). So it may not be the 60-year dovetail, but sometimes it seems like that.

However, the story is not a sad one. For one, the joint is perfect; the only thing left to do here is plane off the proud excess at the ends of the tails and pins:

Another bright spot is that it went off without a hitch, even on pain-in-the-butt yellow-poplar sapwood (the lighter bits on the right). Saw, chop, and test-fit. The pins fit right off the saw; no paring was required on the sides of the pins. Thinking back to the first dovetail joints I made, it took much longer to cut joints that did not turn out as well. My speed is improving.

I’m back to using my cheap Crown gent’s saw, still with the same touching up of the teeth that I gave it when I first got it (slight jointing, filing with a needle file, slight stoning on the sides to remove some of the set, and wax). It works fine, though I can’t imagine how bad it would have been if I hadn’t tuned it. I am still plotting out the dream dovetail saw that I will make one day, but I’m too busy with furniture projects right now to get tied up with making another stupid tool, and honestly, it’s just not that important.

Sometimes I think of what could possibly make things move a little faster, and I came to the conclusion that some operations are actually going quickly but some not so quickly.

The fast ones are:

  • Sawing down the tails and pins
  • Removing the bulk of the waste
  • Test-fitting

The stuff that seems to take a little longer:

  • Laying out the joints
  • Paring down the final little bits in the tail and pin troughs

Now, I know that the paring could go a little quicker if I just bothered to make a pair of skew chisels. I really should get on that case some day.

What about laying out the joints, though? I do this mostly by eye now, marking out “what looks good” (to me, at least) for the tail spacing, then using a square and T-bevel to mark the lines. This works and it is St. Roy- (and others) approved. But it’s not fast for me. I can mark the spacing quickly enough, but lining up the T-bevel to the mark on top always seems to take a little extra time. A dovetail template could save some time, because you can register it to your marks at the top quickly.

But then again, we’re talking about a savings of only about two or three minutes here. And I don’t think it’s worth getting yet another tool for that at the moment. Furthermore, I can’t just arbitrarily pick the angle I want, as I do now. I know that’s been said a million times before, but I do think it counts for something.

Frankensaw; Saw Sharpening Guides

Posted on by
Reply

I’m finally done with sharpening the last saw on my to-do list. This and the one in my previous post were both acquired at an estate sale over in the Sunset district last year. They’re both Disston No. 7s, but both have been rehandled. The first one I worked on had some sort of modernish handle on it. The one I just finished has a No. 12 handle.

Someone must have liked these saws. They were both rust-free and had pretty good visible etches. Unfortunately, the sharpening on them was crap. The No. 7/12 Frankensaw needed serious jointing (and therefore, serious tooth reshaping) before it could be pointed.

But once at the pointing stage, things went pretty quickly in spite of it being a 26″ saw with 8 teeth per inch. I used a 10 degree fleam angle this time, and used a roughly 10 degree slope on the gullets as well. Here’s a shot of the sharpening in progress:

This picture shows one of the issues that often comes up when sharpening crosscut saws that I’ve been talking about recently, and that is, that the shape that the tooth appears to be can be misleading. Look at the teeth on the right side of the saw in the preceding image. See how they look kind of spindly? It’s a trick of the reflected light. Here’s a close-up, where you can see the reflections and the full tooth profile:

Obviously, there are some uneven spots here, too, like the second-to-left gullet, but those were taken care of on a second pass. but the point is that you have to be careful about what you see. Just be consistent with the angles you work at, look at the tops of the teeth to see when you’re done.

Here is the saw in its finished state:

You can see my fleam guide in the photo of the work in progress. I worked a little on the code for this before starting this saw. My previous version required you to use two different cutouts to sharpen each side of the teeth. That was kind of stupid, because you can see only one side of it at a time. So I reworked it.

Update: I now recommend that you use the versions that are on the Plans and Guides page rather than the following, but I’m not going to remove these any time soon.

Here are PDF versions for:

  • 10 degree fleam (fleam10)
  • 15 degree fleam (fleam15)
  • 20 degree fleam (fleam20) (whoops, need to regenerate that sometime)

If you know how to use PostScript, ask me for the source code; you can put in any fleam angle you like. (I haven’t figured out how to trick WordPress into letting me upload a .ps file without doing something stupid like archiving yet, sigh.)

Also, I have a PDF tooth pitch gauge (toothgauge) that you might be able to use at some point.

My Sloped Gullets

Posted on by
Reply

Leif over at Norse Woodsmith did a recent post on sloped gullets. Strangely enough, I’ve been working on restoring a couple of saws lately, and about the time he posted, I was starting to sharpen one of them, an old Disston No. 7. I finished this morning and decided to take a photo of the result. Unfortunately, I don’t have a macro lens, so I can’t get much better than the following shot, and also unfortunately, I don’t think the angle is very good, because you can’t really see the bottom of the gullets too well:

In any case, take a look at the bottom of the brightly reflected edges, and you’ll see that the rear is higher than the front, especially on that tooth all the way to the left. This shot was taken straight from the saw coming off the file, so you can see some burr here and there. You can also see how difficult it is to judge the height and shape of the teeth here. In this shot, it looks like the teeth set towards the camera (the all-dark ones) are a little bit lower than the ones set away, but in reality, they’re all the same. It really is easier to tell by jointing the teeth and filing until the flats are gone.

I’d mentioned in a comment on Leif’s post that the angle I used for fleam and slope was not that great. On further inspection, it seems to be moderate. The fleam angle is 20 degrees and the slope is about the same. I think a 10-15 degree fleam and slope would be easier to file, and I plan to try it out on the other saw I need to sharpen.

All of this theory is useless if it doesn’t work in practice. Here is a test cut in beech:

It’s about what you would expect from a 6TPI saw that’s 26″ long. The important part is that it belches sawdust profusely when sawing, and with a controlled cut, you don’t get much tearout. And it’s always nice to get a 110+ year old tool working again.

Frame Saw: Resawing panels, repairs

Posted on by
Reply

One of the projects I’m starting now is a small bookshelf prototype that I’ll use to guide my way through building later versions. I’m in the process of milling the wood, and to make efficient use of the wood, I decided to resaw 4/4 boards so that I can use one slice for the sides and shelves, and the other slice for a panel in the rear of the bookshelf.

So it’s the first real-life use of my frame saw. It does fairly well; here’s a roughly 1/4″ slice that came off one board (this is Yellow-poplar/Liriodendron tulipifera):

The upper left looks slightly ugly in this shot, but it’s actually just two passes of a plane from totally flush. The result is actually quite good–very flat, no wandering of the saw. I seem to run into difficulty at the end of of the board, and I’m still trying to figure out ways to make that easier. Unfortunately, what didn’t help was the tension on my frame saw being too great for the hardware that I made to hold the blade in place, causing it to tear apart on one side:

Yikes. Well, okay, so I just chopped it off and remade the piece. I think this part is a little weak because of the recesses I filed for the itty bitty screws that I no longer use. I’m still not entirely happy with the arrangement, but if it holds, I’ll change my attitude.

Speaking of attitude, mine towards my front vise has been one of complete irritation for the past several months. I’ve had the endcaps held on at the very tip with screw inserts in the center with a machine screw to hold the whole thing in. Unfortunately, the insert kept coming out (because inserts don’t work well when inserted parallel to the grain, duh), the cap would come loose, and the handle would fall out of the vise. Of course, it would always fall out at an appropriately inconvenient time.

This was happening on only one side of the vise, and I finally fixed it for good today by relocating the insert to the side of the handle instead of the tip (you can see the hole in the tip at the right where the screw used to be):

I suppose that it’s fixed “for good” until this happens to the other end of the handle and I have to do the same thing there. Sigh. Why it took me so long to fix this is beyond me.

Frame Saw: Important tweaks

Posted on by
2

After a bunch of sessions with the new frame saw, I determined that it wasn’t working the way it was supposed to. It didn’t cut quickly, the kerf was too wide, and the blade wandered all over the place. I suspected that more than one thing was wrong, and I had a few ideas.

First, I caught on to the fact that the blade wasn’t really sharp, although it had the appearance of being sharp. I really need to learn that if the guy who made your saw doesn’t have a name like Wenzloff, it probably isn’t sharp, so you should save yourself a lot of trouble and sharpen it before using it. Groan.

I reshaped the teeth to have a fairly aggressive zero-degree rake angle, and took a considerable amount of care when sharpening to make sure that the tips were all very close to the same height. The result looks like this:

When doing this, I realized two additional things about this blade. First, it had way too much set, and second, the saw plate is a little thicker than I thought it was. This latter point was a big deal, because it seemed like the tips of the points originally were chamfered or slightly rounded. I couldn’t see this originally, even with reading glasses. And obviously, it makes a big difference in use, because, as I find over and over again, sofa-shaped blades don’t cut wood very well. (I wonder why.)

The difference in sharpening alone was really remarkable. Because the process removed most of the set, it made for a wonderfully thin kerf, and therefore, it tracked a line much better, even though the blade wasn’t terribly taut. And the more aggressive and sharp teeth cut much faster and smoother.

Now, the second problem I was having was that I couldn’t increase the blade tension too much, because the little screws that I was using to hold the blade in place were snapping due to the tension:

Yikes. So I cut and filed a few brads for this purpose:

Now that there’s enough tension, I don’t have a problem with the blade twisting around (thanks for the pointer, Dan). The only issue I have now is that it’s difficult to keep the blade straight when tightening it up. Christopher Swingley uses a wrench on the flats, which seems like an idea that might work.

After making these two changes, this saw really seems to be on the right track, and I think I’ll be able to do decent work with it. I already sawed out some 3/16″ slices from a couple of smaller boards (without even marking!), and they came out great.

There always seems to be a lot of discussion about frame saws, and what kind of blade is appropriate. Tom Holloway’s saws use thin blades, and I can attest to how effective they are, having played with them. Bob Easton uses thicker blades from old Disston saws and that seems to work too. It seems to me that there’s more than one way to skin a cat.

A case in point here is what I edited in as a note in my last post, that there’s a version of a Japanese saw, the “oga kobiki” or something, that has a gigantic blade, but seems remarkably easy to use. Check out the pictures in this link if you haven’t yet. You see how that tiny woman is using that huge saw? Look at how beautifully the cut turns out. Oh, how I would love to examine that saw that they’re using.

So it seems that if the kerf is even enough–not too wide, not too narrow, of constant width, and straight–it doesn’t matter how big your blade is, as long as it’s slippery and sharp.

I don’t think this is going to be the last blade for this frame saw. I’ve got some ideas that might make it faster. Let’s just say that the gigantic teeth on the Japanese saw got me thinking.

I also got to thinking that I might need to do something about my saw vise. It works pretty well for small saws, but when you start to file the big teeth, it shakes too much. What to do here? Finally make my own? Cave in and get one of those new Gramercy saw vises? Find someone who has a good one and mug them?

Frame Saw: Push or Pull?

Posted on by
4

One of the interesting features of a frame saw is that one can use it as a push saw (with enough tension on the blade) or as a pull saw. Which is more appropriate, if there is such a thing?

I thought about this while in the process of resawing some beech, a bear of a process when the board is a 13.5″ wide slab. I had the saw in at roughly 45-80 degrees and it seemed to me that it was much faster pulling than pushing. What could have caused this, other than not enough blade tension?

As pretty much anyone will tell you, during ripping, your saw teeth are acting as little chisels and are hopefully shaped as such. But then wouldn’t the factor of grain orientation come into play?

  • When pushing into a board with the tooth side of the blade at an acute angle to the board, you are generally sawing with the grain, as you would with a plane.
  • When pulling out of the board, you are sawing into the grain, invoking tearout.

If you tear out, is that a bad thing? It seems to me that it would be faster because your saw would be continuously digging itself into the fibers. Of course, it would be tougher work. Would it also help avoid tracking the grain?

Think of the Japanese timber saws (stuff like the Kobiki Nokogiri or whatever its name is) — these are notoriously quick pull saws.

Well, I think I’ll do some more sawing to see what I like in practice. Probably with something other than a 13.5″ wide chunk of beech, though.

[Edit: Check out this blog post that a friend dug up for me. They’re ripping chestnut logs in a demonstration of traditional techniques, letting lots of people try. The big saw is called “Oga Kobiki” (大鋸) or “Daigiri” or something like that. Check out the angle they’re using in the wood versus the rake angle of the teeth.]

Frame Saw: Almost Finished

Posted on by
4

To make the handles into one end of the frame saw, I first marked out the length-wise dimensions with a pencil and the cross section limits with a marking gauge (hard to see in this photo, but they’re there):

Then I set to work shaping the handles. I used my Shinto saw rasp first. I really like that tool: cheap, innovative, and highly effective.

To finish shaping, especially at the edges, I used my Gramercy saw maker’s rasp. Then I smoothed it out with progressive grits of sandpaper on a dowel, starting at 120 grit. Finally, I hand-sanded to 320 grit, and called it ready:

Then I assembled the saw with the newly-waxed blade:

So far, so good. I tried it out and quickly found that I will need to do some version of the trick that others have done to keep the blade square. It does tend to slip around. It also seems like it’s necessary to practice technique, because this thing does have a tendency to slip out of the kerf because the blade is so thin.

And then there’s the issue of workholding. Hmm. Clearly there is more work to be done here.

Frame Saw: Joinery and Hardware

Posted on by
Reply

After looking at lots of frame saws that a lot of other galoots have made (not mentioning any names here, Mssr. Isola), I decided to make my own for resawing. Well, I had decided to make my own a long time ago, but never got around to it until now. About a year ago, I ordered the blade (a Wilhelm Putsch blade, about 5 teeth per inch), and set aside a piece of beech. Then I proceeded to do nothing else on the project until last week.

After cutting the pieces out and milling them to size, I cut the joinery. There have been many styles for the frame joints, but most people do mortise-and-tenon joints that aren’t glued (the blade tension keeps the saw together), so I did, too. I decided to make the joints haunched, because in theory, that kind of joint resists twisting better than a plain blind joint. Also, I’ve never made one, so I figure I’d better get down to the business of screwing them up.

And I screw up I did, slightly, on the first two. On the first one, I blew out the side of the mortise when chopping it (it was too close to the end, 1/4″). Nothing a little glue won’t fix. On the next one, I cut the tenon too loose, which doesn’t matter on a joint that I’m not going to glue, but still.

The next two came out perfectly, though:

Then it was time to make the hardware to hold the blade. I didn’t deviate from the way others have done it. I used a 5/16″ carriage bolt for one side (bottom in the picture) and a section of threaded rod (top in the pic) for the other:

The steps I took were as follows:

  1. Filed the threads flat and four-square at the end.
  2. Drilled the hole that will hold the blade-holding pin with a Millers Falls #5 eggbeater.
  3. Cut the blade slot with a Bahco junior hacksaw (slowly, to keep the hacksaw blade from wandering around). The kerf is almost a perfect match for the blade.
  4. Cleaned up the areas around the slot with a small tapered file.
  5. Cleaned up the slot with some folded sandpaper in the kerf.
  6. Cleaned up the tip.
  7. Discovered that the little bolts that I was using to pin the blade were a little too short for the little nut to fit.
  8. Filed recesses around the hole in the big bolt so that the little nut could reach the little bolt.
  9. Cleaned up the tip again.

One little tip when you’re filing any kind of thread: Keep a nut on the inside of the filing area. After each stage of filing, take off the nut and put it back on again. This cleans up the threads, although for this project, it doesn’t really matter for the carriage bolt side.

The saw is functional now (I’ve done a test cut), but I’m not quite finished yet. I’m currently waxing the blade and I also need to shape handles into the ends.

[Edit: I have ditched the wingnut in favor of a regular hex nut tensioned with a wrench. The wingnut does not provide enough tension.]

Saw Till: Finished

Posted on by
2

After the glue-up, I planed the sides of till flat. The dovetails looked fine, and the through tenons turned out a lot better than I thought they would:

The front stretcher is slightly proud of the edge, but I decided not to bother planing it flush. I had the last significant step ahead of me, one that wasn’t illustrated in the plan.

The saws have to rest in some sort of slots in the back of the till. I didn’t know how I was going to make these, but I knew that I wanted this part to be replaceable. There’s too much potential to want changes in the future, and what if I screwed up on my first shot anyway?

I don’t have any photos of this (and I’m not going to take it apart to take them), but the next step was to put some screw nut inserts into the upper two stretchers so that I could fasten saw rests.

Then I took a couple strips of the worst part of my cheap mystery softwood, screwed them together, and set out cutting the slots where the saws would rest. I clamped the two rests together so that I could saw the slots as a gang. My first attempt was just a simple kerf with my biggest saw. Here it is, on the left:

After a lot of frustration (and work with my keyhole saw), it was finally big enough, but I decided that I would need to do it differently. On the next one (at the right in the preceding image), I laid out an 1/8″ slot that I would take with two separate cuts. Despite the sloppiest sawing I’ve done in a while, this worked really well:

The slots are 1.5″ apart, and 1/8″ wide. The depth depends on the saw (the bigger saws get deeper slots).

Notice that the structure here isn’t very strong because we’re cutting across the fibers. I figured that this was okay, because the rests weren’t supporting anything. Then I saw how Derek Cohen used certain space in his saw till, hanging backsaws instead of propping them up. We’ll get back to that in a bit.

After cutting the second slot, I did a test-fit:

So far, so good. You can see how the rests are attached to the rear stretchers with machine screws.

I cut the rest of the slots, with some extra attention at the right side for backsaws, and prepared for installation. Most people would just screw something into the wall in their shops, but since we rent this place, I was looking for a way that wouldn’t require me to spackle and paint later. Luckily, I had this to work with at the left of my workbench (where my lumberyard is):

My plan was to be daring and hang the saw till from these rods. I bored a couple of holes into the sides of the till at the top, then strung up some twine from the rods, inserting it into the holes. Getting everything level was certainly a bit of a challenge, because the back does not rest on the wall. However, the total weight is an advantage; it doesn’t swing around much when you grab something.

Does the twine make the end result look more rustic? At this point, I’m just happy that it’s done:

I wish I had known about that hang-the-backsaws trick just a little earlier, because I probably would have planned to hang all of my small backsaws. This isn’t bad, though, because I was not originally planning to put any of those little saws in the till. Now they all fit in there, and I can always reconfigure it if I want to.

Saw Till: Glue-up

Posted on by
Reply

So I had all of the parts ready for glue-up, the clamps were in reasonable shape, and I had the time this morning. It was time to glue and assemble.

I decided that I’d use my Workmate as an assembly table again, but that I wouldn’t use its jaws as a clamp (because I figured that the till would be too wide. So to start, I cleared it and pulled it to the center of the shop (where I would have access to all sides), and laid out the clamps set to their approximate settings:

I put wooden pads (made from the tenon waste of the shoe rack project) on the big clamp pads and between the pins on the tail side of the dovetails where the other clamps would go. Magic tape was my weapon of choice here, since I figured that the bond only had to last for less than a half-hour.

So far, so good. I laid out all of the parts in order (right side as step one, then the bottom, rear and front stretchers, then finally the left side would be banged into those pieces all at once). I decided that it was time to go.

Applying the glue and putting the pieces together wasn’t so bad, though I almost forgot to put in the front stretcher, I got everything together to a certain degree.

Then I moved it to the Workmate and put on the clamps. It was a little quirky, and of course I felt like I could have used a few more clamps, but it all went fine and everything drew up well.

At this point, it was time to drive in the little wedges on the through tenons. At first, it went just fine–I used the strange plastic-faced hammer that I use as a plane hammer (it has a specific purpose, but I don’t know what it is). It was kind of fun to bang those little things in, and they seemed to be holding up fine despite some of them having checks. Thump, thump, klonk. I had glue all over my fingers, but I was making good progress.

Halfway through this process, I realized that I was in trouble. You see, despite having a math degree, it seems that I still don’t know how to count. Remember how I said I needed ten wedges and made thirteen? From where did I get that number? There were four stretchers, with two tenons each, two wedge kerfs per tenon. 4 x 2 x 2 = 16, not 10. I should have made closer to 20 wedges.

Remember also how it took me a stupidly long time to make those? How could I possibly make three more in the limited amount of time I had? (Remember that I’m using liquid hide glue here!) Well, I still had some wide wedges that were in a reject pile. I decided to grab a chisel and chop straight down on these “rejects” at the proper width and hope that the checks and grain wouldn’t make me pay.

The results weren’t exactly perfect, but in a minute, I had made three wedges that would fit, and that’s what I needed. I banged those in, looked over the joints, checked stuff for square, and then started to take pictures.

It would have been better if I would have had a clamp for that center dovetail, but I don’t have one. Perhaps a really big block as part of the cauls.

I found a check in the upper left corner (in the preceding photo, it’s on the right side at the bottom). Oh, drat it. It was a problem I was fearing somewhat, and I don’t recall it being there before. I may have accidentally banged up that corner when I was attaching the left side. It doesn’t make any difference as far as the functionality is concerned, and in fact, it’s likely that no one will even see it (it’s on the far side of the till), but I wish I’d been a little more careful. Actually, what I wish the most is that I could find out when it happened.

Now the glue is curing, and I’m waiting for the final steps, which will be to level and plane off the joints, insert the blade rests, hang it up, and put some saws in there. I will not apply any finish to the till, at least for the moment. This is one of those projects that needs to be complete.