Cutting New Saw Teeth

When I remark that I’ve cut teeth entirely from scratch on many of my saws, some people think that it either involves stamping, a machine, or some sort of magic trick. It’s nothing of the sort. If a klutz like me can do it on the abomination of a saw vise that I have, anyone can.

It’s actually quite simple because it’s derived from normal saw-sharpening practice. You start with a file with a handle and flat block of wood as described in Lee’s book and any saw sharpening site out there. To get the tooth spacing, make yourself a little guide. I wrote my own PostScript program to do it (check the Plans and Guides page for PDF versions ugh, gotta restore that some time), but I imagine that there are plenty of ways you can do it with several packages.

Get some reading glasses; they help a lot.

Fold the guide over the front jaw of the vise, put the blade in, just peeking over the top, and make a first pass with the file, just filing in a little notch over the top of each line (sorry about the fuzziness, but I just don’t have a macro lens):

Notice that I’m not really getting the spacing spot-on; you can tell from the flats at the tips of the newly-formed teeth. Don’t worry about this on your first pass–you’re going to refine it later on. You don’t even have to worry much about the file itself on the first pass. The one I’m using here is way too big for the final tooth size (this will be a 16TPI saw that I will reveal later). I’m doing this because I don’t want to put unnecessary wear on a relatively expensive small saw file.

After you’re done with the first pass, do a second pass to go deeper. Make an attempt to correct for uneven spacing by putting a little pressure left or right as you’re filing, but use a very light touch; don’t make any extra strokes with the file just to even it out, and don’t press harder than you normally would. You still want to be fairly consistent in the newly-cut tooth valley size. The unevenness will disappear as you make more passes with the file.

In the saw above, the teeth are so small that I went most of the way with my second pass, this time using a file that’s appropriate for the pitch. You can see that the teeth are slightly uneven, but not by extreme amounts:

I then set the teeth, jointed them, and did a final filing. At that point, the teeth were quite even. It’s important to joint and file after you set the teeth because the act of setting the teeth dramatically alters the orientation and shape of the cutting edges. In addition, you’ll often accidentally set the teeth more than you need. A final pass at sharpening helps reduce the set. For the saw above, I used a fine-tooth saw set at its minimal setting, and it was still too much!

Frame Saw: Endless Tweaking

Just when you thought it was safe to go out again, here I go again with the frame saw. This time, I wanted to fix some of the problems I’ve had with the blade-holding hardware. The basic problem is that the bolts I was using were too small for anything reasonable as a cross pin to secure the blade. I bought some 3/8″ bolts and threaded rod to go into the frame, then some stainless steel #4-40 machine screws and nuts to serve as cross pins. I used the same procedure to adapt the hardware as before, so I won’t repeat that.

I thought that I had to enlarge the holes in the frame for the new hardware, but it turns out that I needed to do it to just one of the sides–the other was already big enough (I don’t remember doing it this way). But enlarging the holes means that there’s even more of a weak point in the wood at the very point where it is getting the most stress. Beech is very strong for this application, but I didn’t really want to take any chances, so I resawed and shaped some scraps to bolster that point:

Then I glued them to the frame (with liquid hide glue, of course):

A couple of days passed (while I was working on other stuff), and I figured it was best to let the glue cure a fair amount anyway. When I came back to it, I decided that I’d also do something about the difficulty I’d been having keeping the blade straight while tensioning. It turns out that there’s a simple solution. I grabbed a cutoff from the stack (looks like this came from saw handle project) and sloppily cut a kerf halfway down the center:

To use it, just slide it over the blade when you’re tightening it up:

With these changes in place, I can get the blade much tighter with less work.

Oh, in case you’re wondering why there is a the hole in the blade securer, it’s an experiment in keeping everything together while in storage:

Rip Panel Saw: Finished

Progress on my new little saw was puttering along, but with the nightstand finished, I found the time to proceed. The handle was done, the sawplate was pretty much to size but not toothed or sharpened, and then there was the matter of the hardware. I decided to tackle this last matter first.

I wanted to do something a little special for this project because the handle is a more traditional style. However, I still wanted to use the furniture connectors that I’d used for my other saws because, well, I like them. Some time ago, and somewhere in the dark corners of my mind, I began to think of a plan to spice up those things. And out of whatever means, I came up with a ridiculous scheme to make sort of a pseudo-lathe out of an eggbeater drill.

The following is what happens when you do unholy things to a Millers Falls #2:

A few notes on this hand-cranked bit of absurdity:

  • I removed the side handle (seen in the back above) and bolted this through a spacer and a dog hole with a threaded rod and a wingnut.
  • I used the hacksawed-off tip of the sawplate for this project as a washer underneath. Turns out that the hang hole is useful for something!
  • The connector bolt is chucked directly into the drill.
  • I used a pretty coarse file for sizing the head, and a finer one for polishing.
  • The calipers were really handy for getting a consistent size.

I don’t know why this worked, but it was surprisingly effective. Here’s the transformed hardware (with an original nut next to the handle in the background):

It turns out that the “antique bronze” finish had a bit of copper plating underneath, and when I put some very fine-grit sandpaper to the head while it was turning in the drill, it exposed a decent-looking surface.

Notice that one of the nuts and one of the bolts have bigger heads. Those are my “medallions” to match the original. Although, I suppose that if I were being ultra-correct, I’d size the bolt down to the other ones because the bolts on saws are all the same size. Meh, too much work now–I should have thought about that sooner.

I sprayed a protective coat of lacquer on the hardware and then set out to finish the sawplate. For whatever reason, it took me longer to sharpen this one because it was difficult to get past the remnants of the original teeth. Maybe I wasn’t paying attention. Then I sized up the heel of the plate, drilled the holes for the hardware (with that same eggbeater), and put it together. It was finished:

It’s a little difficult to get a sense of scale from the preceding photo. This is a small saw, having just a 16″ blade. Here’s a photo of how it compares in size to my 28″ No. 7:

And here’s a close-up of the handle, showing the ray pattern of the nearly-quartersawn pacific madrone that it’s made from, as well as how that goes with the hardware (if it actually goes):

I guess this is all fine and good, except there’s one question that I might not have answered so well: What’s this silly little thing used for?

In some earlier posts, I’d mentioned that I’d been having a little trouble ripping panels and drawer bottoms (I use my Pax saw for crosscutting them). Although this is not a frequent operation for me, every time I did it, it was annoying. And I’m happy to say that it works wonderfully for this type of thing, as shown in the test cut below:

All of this said, I would not have made this if not for two things. First, I’d wanted to make that saw handle for as long as I’ve had the original saw. And second, I had that “Kobalt” saw sitting around the shop doing nothing but taking up space. If neither of these had been true, I may have just bought a ryoba for the task.

I will mention, however, that if this saw were filed crosscut, it would make a nice kid’s saw.

Rip Panel Saw: Sawplate Modifications, Taper Grinding Musings

Sometimes, you gotta learn the hard way.

I started work on the sawplate by cutting off a bit of the tip so that the hang hole was off and there was a reasonable depth to the blade at the tip. So far, so good–no trouble encountered there.

Then I removed the lacquer from the blade with this clamping arrangement and Norton 3X sandpaper:

That, too, worked pretty well, and to my surprise, it removed .001″ of the thickness of the blade. That got me thinking that, hey, maybe I could remove some of the steel with this sandpaper, too. Maybe I’d even be able to taper it!

Well, that turned out to be an idiotic idea. Spring steel is still pretty hard, m’kay? After wasting who knows-how-long to remove maybe .0005″, I decided that maybe I’d try drawfiling it:

That sort of worked. After expending a ridiculous amount of effort, I’d knocked the plate down to about .032″ (from .035″ originally), and I got the tip of the top tapered down to about .029″ (maybe). I decided to call it finished.

Please don’t try this yourself. It was an incredibly stupid idea, and I’m never going to do it again. If you want a nice blade, I suggest this approach:

  1. Start with the blade thickness that you want in the first place.
  2. Use something like a belt sander with Norton 3X belts if you want to taper grind it. (And remember to use proper safety equipment–a respirator mask, safety glasses, gloves, and something that will hold that blade in place with no chance of letting go.) Not that I’ve tried this, but it’s hard to imagine any way that this would be worse than what I did.

Speaking of taper grinding, I discovered something about it when working on this project. There is a fair amount of smoke-and-mirror action going on with some of these old taper-ground blades. Most full-size handsaws seem to go from a full thickness of about .040″ to around 0.030″ along most of the significantly-tapered tapered part of the top, then it drops down to about 0.025″ at the very tip at the top (maybe the last half-inch). So that’s about 25% thinner along most of the top and I can see where that could be helpful in use. But the teeny bit at the very end? Meh. You hardly ever even have that part in the kerf, anyway.

My Pax saw has similar taper grinding as on the old saws except at the very tip. It appears much thicker than the old saws, but measuring reveals that it’s about the same. What’s going on here?

The original saw that I’m copying has less significant taper-grinding along most of the top–its cutting side is about .031″ and it’s ground to about .028″ along the top. That’s about a roughly 10% difference. However, it looks a lot thinner up there, and after a long time, it dawned on me that they’ve rounded off the top of the saw a little to make it look more elegant. It was the same thing with the older saws, and that’s why the Pax doesn’t look like it’s been taper-ground as well, even though it has.

At this point, I was really starting to lose my mind, so I did the same rounding operation on the top of my new sawplate, and guess what? It now looks like it’s been “seriously” taper-ground. Continuing the absurd voodoo practice, I filed a nib onto the thing:

Now I’m almost ready to form and sharpen the teeth. I was originally going to file off all of the teeth and start new ones. However, it seems that the original teeth on this saw were set at 8TPI, which happens to be about what I want on this thing (it also matches the original), so maybe I don’t have to go the whole way.

The handle varnishing is almost done, too.

Rip Panel Saw: Choosing a Sawplate

While the handle is in the varnishing stage, I’ve been worrying about how to approach the sawplate for the new saw. Being the cheapskate that I am, I have been considering using the plate from this thing that I picked up for a song and a dance some time ago:

The basic shape is right, but of course, I’d cut the tip off to make it look halfway normal, mill off the teeth, and cut new ones, and do whatever else is necessary to get it fit on the handle.

The trick to all of this is the thickness of the plate. It’s kind of a tricky business, because all of the old saws were taper-ground, and you won’t find a cheap saw these days that is. So I’d be giving that up, but I don’t have any old saws that I can cannibalize anyway.

Furthermore, until now, I really didn’t know the thicknesses of any of the plates that I was working with; I was only able to sort of guess by looking at them. Well, that all changed when I finally got off my butt and bought a micrometer. Why I didn’t get one before, I’ll never know. It’s really taken a lot of the guesswork out of a lot of stuff.

I’d been wondering about this because I’ve got saws that turned out in certain unexpected ways due to this, and I really didn’t know about it before. For example, my crosscut carcase saw works wonderfully, but a rip version of it didn’t turn out the way I wanted it to. The plate on those two saws is .0263″, but you’d really want something more like the 0.020″ thickness found on most dovetail saws.

What’s .0063″ of difference? Well, one way to think about it is that it’s 31.5% thicker, which is pretty significant. In theory, you’re cutting 31.5% more wood, and so it’s 31.5% more work to do it. (This does not hold for a crosscut saw, where you are slicing fibers out instead of scooping shavings.) You can also look at it visually; take a look at the following image that I just cobbled together as a PostScript program:

The difference between .020 and .0263 looks pretty striking here. These thicknesses are at the cutting edge. I have the measurements for the tops of the saws that are taper ground, but I’m going to set that aside for the moment.

As you might have noticed, I put labels on this image. Of course, after I got my micrometer, I obsessively measured all of my saws. I don’t have a Kenyon saw, but it’s listed here as a representative for a big tenon saw (the Wenzloff versions have .025″ plates).

There are some surprises, such as that Stanley dovetail saw with the gents handle. This is a relatively new one that I picked out of a basement. I never bothered with it because I already had the Crown equivalent (.0205″), but you can see that it’s quite thin! This is the saw featured in Korn’s book, and it should be pretty obvious that if you’re willing to learn how to sharpen it, it will give you experience with a thinner plate and should work just fine. It’s no wonder he has no trouble recommending it.

So getting back to the question at hand, is the lovely “Kobalt” plate going to work? I think I’m going to give it a shot. It’s about 16% thicker on its cutting edge than the original that I’m working with, but that doesn’t seem that bad (and some of this thickness may be lacquer). The thickness of the original is a little unusual anyway; full-size saws are about 33% thicker. The only blade I have that’s close is my frame saw blade (0.029″), and, er, that’s not gonna work.

Then there’s the matter of taper-grinding. Should I try this? I’ve got sort of an idea of how I can do it without a power tool (although if I were sane, I’d ask to use someone’s belt sander). Would that warp the plate? Hmm.

Rip Panel Saw: Handle Shaping

Available time to work on this saw comes and goes, but it’s getting somewhere. After coming up with the template, I grabbed the piece of madrone that I’d worked on before (and made the nightstand drawer pull from) and thought that there might be enough of a solid section to make this handle. I started by ripping down the sides (you can’t split madrone as you can oak, because it does not split terribly straight). I came up with the following ridiculous-looking, yet surprisingly effective workholding arrangement:

I thought that it might be a pain to saw, but it wasn’t that bad. I also have a newfound respect for my front vise. Before long, I had a little board milled.

The big question still remained about the checks. I suspected that I would be able to position the handle cutout where the cracks would not extend, so I traced out the pattern:

Yes, this would be unaffected by the check on the top of this photo, but I didn’t know how far in the one at the right end extended. When it was all said and done, a small amount does remain, but its location is inconsequential (details of that might be in a later post).

I cut out the rough template with the coping saw:

As I was doing this, I started to suspect that the shaping process would be somewhat more difficult than my previous handles. There was something about this wood that just felt different.

I made quick work of getting down to the lines on the outside with my saw rasp and sawmaker’s rasp, but the inside was a little trickier because there wasn’t much space to work. I started to wish that I had an inchannel gouge, and after a minute or so, I realized that I did have one, that Taiwanese one I’d bought during the holidays.

I don’t have a proper mallet for it, so I just banged at it with a marginally not-uncomfortable chunk of the madrone cutoff, and this got the job done (though I think I want a proper mallet):

It turns out that this is a really good gouge! (The ones with the red at the ends are supposed to be the better ones.) After all of this chopping, there were no nicks and it seemed just as sharp as ever. I’ll have to pick up a few more of those on my next trip.

Then it was off to shaping. That, as I suspected, was kind of a pain. Madrone cuts easily enough but its sawdust is really fine and clogs everything quickly. I ended up using a lot of 60-grit Norton 3X sandpaper to fine-tune the shape because that stuff doesn’t clog so easily. I used a 1/4″ chisel for a lot of the nooks and crannies and imagine that some outchannel gouges would have been handy. But I did eventually get it done, and then cut the kerf for the blade using my trusty gents saw:

I also tapered the front with this saw, but made some modifications to the original. Here’s the new handle next to the original:

The new one is tilted slightly counterclockwise to the original in this shot, and there are some perspective issues, but you can see some subtle differences, some intentional, some not quite so. The bottom is a little thicker and the cheek has a little bit more surface area. The front is tapered a little differently. The original’s tapering was uneven, so I decided that I would smooth it out a little on mine. The unintentional change is how the back of the “lamb’s tongue” isn’t as deep as on the original. I can’t figure out how this happened, but I’m not going to bother changing it. (If I didn’t like the way that it looked, I’d change it, but I do like it.)

So now, I suppose I should put some holes in there for the screws and varnish the thing. And I happen to be already varnishing the nightstand at the moment–how convenient.

Rip Panel Saw: Handle Template

I’ve got the nightstand drawer in the clamps, put the first coat of varnish on the frame, and sized up the top. Now with just varnishing left on that project, I have been given license to do whatever I want for the next two months while I do the finishing (and wait for the varnish to dry) on that project.

Aw yeah. I’m gonna make me some more tools.

The list includes three saws. One of these is a little panel saw filed rip that I’ve found myself wanting several times in the past. There’s another reason for this–I’ve got an old Disston panel saw, probably a No. 7, that I really like, but won’t use because the handle is loose (due to a crack), the blade’s slightly broken at the end, and it’s kind of rusty. The handle on this thing feels incredibly comfortable in my hand and I’ve always wanted to make one.

I had a little time today to work on the template for the handle. I dug out an old PostScript program I wrote that draws a grid, barbarically converted it from metric to imperial units, and printed it out. Then I placed it on a wall next to a shelf, put the saw on a shelf, and took a photo (dumb note: this is one of the very few on this site where I used a flash):

Hey, let’s look at that medallion really quick:

The Disstonian Institute says that this is from around 1874-1875.

Getting back to the task at hand, I loaded a cropped version of that photo into Inkscape, scaled it so that the grid matched the scale of the page, then made that layer transparent and laid out the pattern on top as vector graphics:

This was a lot easier than I remembered it being in the past. It only took about 20 minutes to get it set.

Normally, I would try to simplify old-fashioned lines on a handle because I like to have a somewhat more contemporary design (for example, ease out that angle on the inside), but I like this saw so much that I’m going to try to clone it as faithfully as possible. That’s not going to stop me from using my usual furniture connectors as the hardware, though.

With the lines set, I printed out the template:

Now I have to cut it out and trace it onto the piece of wood that I’ll use for the handle. The trick is that I don’t yet actually have a piece of wood. Okay, gotta work on that.

[Update: This template is now available on the Plans and Guides page.]Aarrgh.

Nightstand: Resawing Legs With The Frame Saw

Even though I have not yet completed the cutting list for the nightstand, I do know that I will need four 24″ long legs that are 1.25″ square. I had meant to get the stock milled for this sooner, but it’s been a busy week. I finally completed the flattening of the 7″ wide 8/4 board segment that I plan to use today, and then faced the task of resawing this thing into the piece for the legs and a panel for one of the sides.

It’s been a few months since I last did any resawing, and I wasn’t particularly looking forward to it, given how much work it’s been the last few times. But this time was a little different (otherwise I wouldn’t even be writing about it).

I wanted to try out the techniques shown in a few videos of those magnificent Japanese sawyer’s saws. They use them mostly perpendicular to the grain, letting the huge area of the saw register a very smooth cut through the wood. I can’t do this completely with my frame saw because its blade is not deep enough to register so well, but I could at least try to do it partway.

I started by the diagonal cuts along marked lines that are necessary in order to get a square cut started. Further cutting along each side at a very slight diagonal established a somewhat shallow kerf on side. Then, to connect the two kerfs, I clamped the board upright in my front vise and went at it almost perpendicular to the grain (I’m pushing it here):

As noted before (by several folks), it is really important to tighten up the blade to a very high tension in order to keep the blade from binding up and cutting a curve. From time to time, I would reach the end of the kerfs I’d cut on each side and I’d have to extend them. I think that this actually took more time than doing the bulk of the sawing. In fact, most of this went very quickly because the saw followed the kerfs better than before.

I sawed all the way down to a spot about an inch away from the end, and then wondered if there was any better way to finish off the cut with the vise. I then did a very silly thing: I clamped the board in upside down, with the saw upside down, and pushed the saw up from underneath to finish it off. Don’t do this; it’s ridiculous. Afterwards, I realized that it would have been a lot easier to clamp the board flat to the bench with the uncut end hanging off the end, and just work the saw sideways for that last little bit. There are other setups that I’ve seen that do a better job still, but I don’t have the resources for that yet.

So I was finished with this bit:

Not bad, not perfect, though. There’s one saw mark on the top center-left, but it’s very shallow (one pass of the fore plane will remove that).

I’m starting to figure out that my frame saw is pretty good for boards that aren’t so wide. This was a 7″ wide piece of beech, and that’s probably about the maximum that’s comfortable with this saw. If I ever build another frame saw, it will have the following changes:

  • Longer frame and blade to handle wider boards better.
  • Wider blade for better registration.
  • Much larger teeth.
  • Improved hardware for attaching the blade. I’m not sure how to go about that just yet, but I do think that it ought to be bigger than the stuff I currently use.
  • Yeah, a saw bench would help a lot, too.

That said, what I have is workable for what I do now, and I do have to finish this project in a reasonable amount of time. (I think it’s supposed to be done by the end of August.)

Stool: Plan, Milling Stock

Now that the bookshelf prototype is glued up and ready for varnishing, I’m able to move on to my next project, which will be a short stool, essentially a single-step stool similar in spirit to the one that my great-great grandfather made. It won’t resemble his in size, wood type, or joinery; it will be adult-size and not use nails. Here is the general plan:

[Update: The plan is now available on the Plans and Guides page.]

I’ll make the whole thing from a wide piece of beech that I got last year. The top has already made an appearance here, although back then, it wasn’t flat, uniform, and smooth, as it is now. This board is 8/4 stock, but unfortunately, part of one side of a 2.5-foot section isn’t usable. After looking at the requirements for this project, I determined that this section contains enough usable wood, though I have to be careful where I cut. The legs will be 1.5″ square, and the stretchers will be 1.5″ x 1″.

Milling the legs and stretchers is a little more work than I expected. Because the board is quite cupped, and also because I’m trying to avoid some nasty parts of the board, I decided to mill each piece to thickness individually instead of the entire board at once. I’m first ripping out sections like this (notice the wedge in the kerf at the end):

You can barely see the nasty stuff I’m trying to work around in this shot; the discoloration at the upper right and a previous (aborted) attempt at trying to slice that stuff off.

After ripping, I have to flatten one face, square an adjacent face, saw some more, and plane some more. So far, I have one leg and one stretcher four-square. If I didn’t have the big saw, I probably wouldn’t even be that far.

Bookshelf: Making Panels

The final components in the bookshelf are panels to go in the back. Recall from earlier that I used my frame saw to separate slices from the rest of the stock; those slices were to become the panels.

I’d been dreading this part somewhat because milling those panels down to size always seems to be kind of a pain. It’s not that they have to be flat (they don’t; stuff that thin bends to a certain degree), and it’s not that they have to be the same precise thickness (they don’t; you need only line up things on the face). It’s that holding the work in place without obstruction had always been a pain.

So I put in yet another dog hole and made this planing stop out of two dogs, a clamp, and a piece of hardboard:

It turned out to be not so bad. I think it would be even better if I put in some tiny brads at the end of the hardboard to grip the end of the panel, but I haven’t gone too nuts yet.

Another thing I did finally was to put some serious camber on the blade of one of my jack planes. Since I have four, it would be kind of silly not to try. I did this by just letting my Norton 220 waterstone dish out as I was grinding out a new bevel angle. That stone dishes notoriously, but in this case, it produced exactly what I wanted it to:

This really made the panel-milling process a lot quicker and a lot less work. Yay.

Once the pieces were milled, it was time to cut them into various kinds of strips and parts. This process is just like with any other kind of board; you first joint and mark:

Then you get out some sort of saw, rip alongside the mark, and joint that straight. Check out the snazzy reflection of the wood in the saw here, the one that supposedly lets you know if you’re sawing straight:

When first approaching this task, I thought to myself, “A rip panel saw would be awfully nice right now, wouldn’t it?” But it turns out that the bigger tenon saw was actually a lot of fun to use on this thing.

With everything cut to width and approximate length, I put the pieces in the arrangement in which they’ll be glued. It looks interesting, to say the least:

Whether it’s a good idea or not remains to be seen. I haven’t glued them together yet. Maybe tomorrow. I still have to plow the grooves in which these things will rest, and therefore, I still have the part where I wish I had a plow plane ahead of me.

(And yes, all of this because I’m not using plywood. No, I don’t have anything against plywood.)