Saw Filing - The Actual Filing

The Jig
Saw Setting
Saw Filing
Some Theoretical Issues
Sloping Gullets
Copyright (c) 2002-15, Brent Beach



If you are filing a saw and retaining the existing filing angles, then you already know what rake, bevel and slope you are going to use.

Some of the time you are filing a saw for a specific purpose so need to select rake, bevel and slope angles appropriate to the wood and the type of sawing you will be doing.

Recommended Filing Angles

The recommendations in this section come from books and pamphlets I have read over the last 10 years. Not all are easily found.

Hodgson on Saws

Fred T. Hodgson wrote a book Hand Saws ... published in 1909 but copyright 1888, includes some angles for cross cut saws:

wood type rake bevel
soft wood 20d to 25d 40d to 45d
hard wood 5d to 10d 20d to 25d

He adds a few of his own drawings, but those drawings are impossible.

Charles L. Johnson

Charles L. Johnson produced a small pamphlet in 1907 which he called:

Charles L Johnson

For regular hand saws, he gives at least 8 different filings (10d means 10 degrees):

Use Rake Bevel Points
Fine Cut-off 10d 30d 10-11-12
Coarse Cut-off 15d 40d 7- 8- 9
Buck Saw 20d 30d 5 1/2- 6- 7
Rip Saw 3d to 7d 2d to 5d 5 1/2- 6- 7
Heavy Ripping 3d 2d 5-1/2 pt
Light Ripping 5d 5d 6 pt
Long Miter 6d to 7d 5d to 10d 5 1/2- 6- 7
Short Miter 10d 10d to 20d 7-8-9-10-11-12

A Buck Saw is a saw for bucking up fire wood - cut quality not an issue, cut speed more important.

A Long Mitre is a cut angled to the grain, but closer to a rip cut.

A Short Mitre is a cut angled to the grain, but closer to a cross cut.

Johnson got around the problem of people not having protractors by including a number of celluloid protractors with his filing manual. The protractors were marked with the various angles. Although copies of the pamphlet survive, I have not seen a protractor.

Johnson says that the file should be level at all times - without saying it, he denies the possibility of sloped gullets. His pamphlet was written at about the time that the Disston books no longer appeared to be suggesting people use sloped gullets.

I use Johnson's Light Rip angles, but find that his recommended bevels for crosscuts are too big.

Leonard Lee

As usual, Leonard Lee's book A Complete Guide to Sharpening includes an excellent section of saw filing.

His directions for filing cross cut handsaws are contradictory on the topic of slope. In the instructions he says: "With your file aligned with a bevel guide, drop the handle the same number of degrees as the amount of bevel you are using ...". However, in a sidebar on he says: "In practical terms, this is not worth the effort."

His recommendations for filing angles:

Saw type Rake Bevel Slope
Cross-cut 15 - 30d 10 - 20d 10 - 20d
Rip 0 0 0

Saw Filing

With the right file, the appropriate slider and rake angle, saw filing is pretty easy with this jig. Here are a few tips.

  1. Arrange your saw vice so that you are in a comfortable position while filing. I sit while filing, with the saw vice just below shoulder level. I can comfortably reach over the saw to hold the far brace, but am close enough to be able to see the tips of the teeth, and the flats left by jointing, clearly.

  2. Set your work lighting to make the flats clearly visible. You have to be able to see the jointing flats while filing, so find a position that gives the best reflection of light off the flats when you are in filing position.

    I have found that in my shop, the flats are more visible with the saw handle on the right end of my saw vice. The flats are easier to see on the first pass, since they are wider. It is much more important to be able to see them on the second pass, when you are removing them completely. So, always do the first pass with the saw in the direction in which the flats are harder to see.

  3. Joint the saw. Most of the loss of metal in a saw occurs during jointing, with only a small amount being worm away during use. I find the existence of the flats produced by jointing very helpful during filing, so I joint very lightly almost every time I sharpen a saw. If you do a lot of saw filing, you might be able to sharpen without jointing. You will be filing by counting - using the same number of strokes on each tooth. Getting an even filing without jointing requires considerable talent.

    Jointing should put a small flat on the top of each tooth. If the teeth are not about the same height, some of the teeth will not get a flat after a light jointing. If there are just a few short teeth, don't worry. They won't do much cutting now, but will eventually. It is not worth it to remove a whole lot of metal here just to get a flat on every tooth. If most of the teeth have no flat, joint a little harder.

    Charles Johnson recommends making a special jointing file by removing most of the teeth on one side of a file: "place it on a large true faced carborundum whetstone, and grind one face only, until it is nearly smooth". He recommends jointing every time, but is able to joint very lightly with this almost dull file.

    Most saws benefit from a little breasting which produces a roach back or salmon back saw. With the teeth pointing up, the middle of the saw is a little higher than either end. The difference is very small - as little as 1/32". You create this shape during jointing - put a little more pressure when jointing near the saw ends than in the saw middle. The arc shape produces a smoother sawing motion. As well, it compensates for a natural tendency of the teeth to become concave rather than convex. Most sawing is done in the middle of the saw, very little within 6" of either end. So, the middle teeth get dull while the teeth on the end remain sharp. Sharpening only the dull teeth produces a concave tooth line - the opposite of the breasted form that works better. (There is no need to get all the breasting the first time you file a saw. Aim for some breasting over a number of sharpenings.)

    After the jointing step, the tips of all of the teeth are their final height.

  4. Set the saw, if there is currently not enough set or the saw is badly set. I only set about every third to fifth sharpening, if I noticed the saw was binding during the last use. If the existing set is really bad, you might do another light jointing after this setting.

    Most saws I have bought have too much set. They may have been set for wet softwoods and I mostly saw dry hardwoods. A tapered saw requires very little set for use in dry hardwoods. All saws will need more set in wet softwoods. More on setting saw teeth using a drawing from an old Chas. A. Strelinger & Co. catalog.

    There is a possibility that even though you have set the teeth well one or two teeth are set slightly wider than the rest. As well, on tapered saws, the blade near the heel is thicker than the blade near the toe. Even if set perfectly, the teeth near the heel will be wider than the teeth near the toe.

    The teeth now are all of exactly the right height and width/set - too bad they are all dull.

    More on setting the saw teeth.

  5. Select a file appropriate to the size of the teeth. I use Extra Slim for most large saws, switching to Double Extra Slim for the finer toothed saws (over 10 ppi). Generally the file should be large enough so that the teeth cover less than half of the file face during filing, theoretically allowing you twice as many filings before the file becomes dull.

  6. Select an appropriate bevel angle. You are actually selecting the slider appropriate to this saw. I usually use a 5 degree bevel for Rip Saws, 25 to 30 for crosscut saws.

  7. Assemble the jig with the appropriate slider and file. Make sure the braces are level and the file is parallel to the back dowel. Tighten the braces to the back dowel.

  8. Select a sharpening strategy! There are sharp disagreements on which gullets to sharpen from which side of the saw. I have my own strategy, which I describe here. I work from the heel to the toe, with the tang of the file always closer to the handle than the toe. This means that the tooth on the handle side is pointing away from me, the tooth on the toe side toward me. Put another way, I am sharpening the front of the tooth with the set, the back of the tooth against the set.

    You can do it other ways. You can file with the file tip pointing toward the saw handle, or toward the saw toe. You can work from the handle toward the toe, from the toe toward the handle, or alternate. Make sure your strategy gives you the best possible views of the teeth tips on the second filing, when you have to just remove the jointing flats. Find the filing strategy that works for you, lets you produce the most consistent filing, and stick to it.

  9. Set the rake angle using a second slider.

    This sketchup model shows the fully assembled jig with the file resting on the teeth. The slider on the back dowel has the appropriate hole angle to provide the desired tooth bevel. The file is on the saw handle side of the jig.

    A second slider with the front face having the desired rake angle is placed on the teeth.

    Loosen the screw holding the file dowel. With the slider on the teeth and the file in the gullet, rotate the file dowel so the file has the correct rake angle.

    As you move this second slider toward the file you rotate the file so the file face aligns with the slider face. This face of the file will produce the front of the tooth which must then have the same angle as the front of this second slider.

    Once the file face aligns with the second slider front, tighten the machine screw, locking the dowel position.

    The second orientation of the sketchup model, looking from the other end of the file, shows the file face almost aligned with the front of the second slider. As you adjust the file face by rotating the dowel, you push the second slider toward the file until the file face is flush on the front of the slider.

    Hard to explain, easy to do.

  10. Shape the teeth, if necessary. If there are few gullets that are much smaller than their neighbours, do these teeth first. You would file this gullet until it is as big as its neighbours, even if that means filing away more than half of the flat. This step is only necessary if the saw has been badly sharpened. Again, don't worry if a couple of gullets are bad. Clean them up a little each time you sharpen.

  11. File alternate gullets from the handle to the toe. This means that the slider will be on the toe side of the file.

    As you start each gullet, check both neighbouring gullets. If the gullet you are filing is noticeably bigger/smaller than its neighbours, file away less/more than half of the flat. [When using my jig, take a couple of practice filing motions to make sure you are lined up correctly. If you are not lined up the dowel will jam in the slider. If the dowel does jamb in the slider, let go and the jig will move to the correct position. Take a light grip and slide back and forth a couple of times, then resume filing.]

    When you near the toe, the slider will drop off the end of the saw. Flip the back dowel to the other side of the file, reset the file rake angle, continue to the toe. If you move back to a gullet you just filed, you can use that gullet to set the rake angle. Having to reset the rake angle for the last few teeth is a real nuisance - you are down to a few teeth on a part of the saw you don't really use. I do it anyway, so the saw is uniformly sharpened.

  12. Switch the saw in the vice. Working from the heel to the toe, file alternate gullets with the tooth closer to the heel facing away from you. On this pass you must remove the flats on all teeth.

    Before you start in a gullet, check the size of the flats on either side. If the flat on one tooth is larger than on the other, you will have to file the tooth with the wider flat harder. It is actually quite easy with my jig to work only one side of the gullet. If you do not force the corner of the file into the bottom of the gullet, you can file only the tooth face of interest.

    When the flats are almost gone, rotate the file away from the teeth so you can check both teeth to compare the size of the flats. When you are almost done, ease up on the filing pressure a little. The advantage of this jig is that you can move the file away from the teeth to get a good look at the flats, then bring it back with exactly the same file position.

Quiet Version

My jig allows you to consistently file one side of a gullet differently from the other consistently. As mentioned above, if you keep the lower corner of the file a bit above the bottom of the gullet and move the jig in the direction of the tooth you want to work, you can easily do just one side of a gullet.

This allows you to perform an experiment on noise and effort depending on whether you are filing with or against the set. I think you will find the difference considerable.

Based on this observation, you might want to try a saw filing system that takes 4 passes along the saw, rather than the usual 2 passes. Given the extra saw handling, the gains in noise and effort would have to be considerable.

Pass 1 - Start at the heel, file in the gullet with the tooth set away on the saw handle side, jig set with the file handle toward the heel, and file just the front of away tooth, removing half of the flat.

Pass 2 - Working from the same side of the saw, from the saw toe, with the file handle pointing away from the saw handle, work back toward the saw handle filing just the back of the teeth done on the first pass (the away facing teeth). Since you are finishing the tooth, on this pass you file to just remove the flat. Now both front and back of the away facing teeth are done.

Pass 3 - Working from the other side of the saw, file handle toward the saw handle, file in the gullet with the away tooth on the saw handle side, file just the fronts of the away facing teeth. On this pass you are removing only half of the flat.

Pass 4 - working back from the toe to the heel, file handle facing the toe, do the back of the teeth facing away from you. On this pass you are finishing this set of teeth, so you file to just remove the flat.

This takes twice the passes, but the filing of each tooth goes much faster and is much quieter - you can easily do it this way without hearing protection. Since you are only filing one tooth at a time, you need only look at that one tip. As well, you put the best possible surface (filing away surface) on both the front and back of each tooth.

So, there are three filing systems to try. If you detect an advantage to any one of them, let me know.

Quick Touch Ups

As mentioned in the How To's, it is easy with this jig to file just one side of a gullet. This means you can touch up a fairly sharp saw with a much reduced set of steps.

Set the jig up as usual, then without jointing or setting, work from heel to toe just filing the fronts of the teeth. One smooth filing stroke to each tooth. Such a light filing will change the height of each tooth by at most a few thousandths of an inch. Such a small change will not put the saw seriously out of joint.

Reshaping Teeth

Reshaping teeth, even just to change the bevel angle, is a difficult process.

Changing the filing on a mitre-saw from 15 degree rake and 20 degree bevel to 10 degree rake and 40 degree bevel took me two complete filings to get most of the teeth into the new shape. The first filing was very difficult, as the file kept getting stuck on the edges of the old bevels. Many of the teeth looked pretty ugly after the first complete filing.

On the second filing, the teeth were more uniform, but some tooth backs were still not flat.

Removing set is more difficult than tapping the teeth back with a hammer. After several passes down each side there were still teeth that were uneven. Side filing seems to be required to make the teeth uniform. I suspect that there is a lot of spring back since the teeth are not being hammered against an anvil. Do this before reshaping.

Filing FAQ

  1. Why do saws have different numbers of teeth per inch?
  2. How do you get consistent gullet size?

Why do saws have different numbers of teeth per inch?

The main reason is that as the saw cuts the kert, the sawdust collects in the gullet. If the gullet fills, the sawdust starts to press against the bottom of the kerf and lift the tips of the teeth up. The teeth no longer cut.

The second reason is that finer teeth have less set, so tend to leave a smooth surface.

People wanting to have a very smooth cut face may opt for a saw that is too fine. They end up spending a lot more time sawing.

How do you get consistent gullet size?

What is the problem?

This drawing is from the Disston Lumberman's Handbook 1919, Figure 13, page 159.

I doubt anyone ever produced a filing that bad. To do that you would have to never joint the saw and also get the rake, bevel and slope wrong on each tooth. If you use my jig, you would not make rake, bevel and slope errors, but you could still have a problem with uneven gullet size.

A filing machine can control depth based on teeth per inch by filing until the file reaches that depth. With hand filing you do not have that option. You must actually look at the teeth and make a decision on every filing stroke. Actually multiple decisions. How hard will I press? Will I press more against the tooth on the right? More against the tooth on the left? Equally on both? Am I finished this gullet?

How do I solve the problem?

Even with the jig, you will still need to acquire the skill to ask and answer these questions in a split second as you lift the file at the end on one stroke, look at the gullet as you bring the file back for the next stroke.

The key to learning is feedback. You asked those same questions before the just completed filing stroke. You decided on how much pressure on each tooth. So, your first goal on lifting the file is to decide what happened. How much of the flat on the right disappeared? How much of the flat on the left? Did you previous plan work out as you expected? Are you putting too much pressure on one side or the other consistently? Having learned, you go on to make your plan for the next stroke.

This is the learning sequence - make a plan, execute the plan, observe the outcome, adjust future plans based on this feedback.

The jig means that when you execute the plan there are fewer variables that can vary from your assumptions of what you are doing. The rake angle is pretty much what you expect, as is the bevel angle and the slope angle. Really, the only variables you have to control for are the pressure you are applying to the tooth on the left and the tooth on the right.

A really good filer, appraising a gullet, may be able to decide exactly how much total pressure and how to divide that pressure in a single glance, and execute perfectly. When she lifts the file and sees she has done it again, another perfect gullet, she immediately moves to the next gullet. Exceptional filers do this with no jig while watching baseball and drinking beer in a dark room. The people who say you don't need a saw filing jig use this exceptional filer as the justification for their argument. I hear about this guy a lot but have never met him.

Me, I expect to take three strokes on each gullet.

  1. The first finishes most of both flats. (Finish means about half of the flat on the first side, and the rest of the flat on the second side)
  2. The second finishes the rest of the flat on the left of the gullet.
  3. The third finishes the rest of the flat on the right of the gullet.

Some times I only take two strokes for a gullet - the left or right flat is perfect. Rarely one stroke is enough - both flats are perfect.

Why do I need to solve the problem?

Aside from looking better, a saw with consistent gullet size cuts better as well. The reason is related to the first question. The ability of a saw to cut is limited by the smallest gullet. Once a gullet fills with sawdust, that region of the saw is lifted up and does not cut until the sawdust leaves the gullet when the gullet leaves the board.

Selecting the right saw for the thickness is important. If all gullets are big enough (the number of teeth per inch is small enough) then you won't have this problem even with considerable variation in gullet size.

One step at a time.

The decision you have to make is: How much of the flats on the adjacent teeth will I remove while filing this gullet?

This decision can only be made during the filing from the first side of the saw when the entire flat is present. When you file from the other side, you must remove all of the adjacent flats. (See Systemic Error below.)

When you start filing a gullet, you look at the gullets on either side of that gullet.

If only one neighbour is larger, remove a bit more than half of the flat on that side only.

If only one neighbour is smaller, remove a bit less than half of the flat on that side only.

Unequal Pressure

How do you remove a bit more than half of the flat when you are filing the gullet? What is he talking about? Crazy man!

Well, when you are filing, you can push forward with equal pressure on both sides of the gullet. This removes metal about equally from both teeth. The flats will disappear at about the same rate.

If you want to remove more on one side, you press down and a little to that side. What a little is you will learn with time. What you actually learn is how much of the flat disappears when you file with pressure almost entirely on one side (and down).

Filing then becomes a sequence:

Systemic Error

There are two sorts of problems: occasional and systemic.

If you look at a saw and see a couple of smaller gullets you have an occasional problem and can fix it as a part of your regular sharpening sequence. With each sharpening the consistency of the gullets should improve a little. You may create a new problem now and then, but expect to fix it the next time you sharpen.

If you look at the saw and see that every second gullet is smaller, then you have a systemic problem. You are filing the first side differently from the second side.

This problem is more difficult to solve because the feedback only comes after you have finished filing rather than as you file each gullet. You look at the saw after lifting it from the saw vice and notice the problem! Egad! But, by the time you next file a saw you have forgotten. How do you learn from your mistakes?

To get the feedback you need you have to look at the saw before putting it in the vice and you have to make an overall assessment. Is there any systematic flaw in this saw? Are the gullets filed from one side consistently smaller than those from the other side? If the answer if yes, you will have to adjust when filing the first side. Either take a little less than half the flat, or a little more than half the flat. Once you start the second side, you have no room for adjustment.

Some saws are pretty bad. Hopefully this is a saw you bought, not one you have previously filed. On a saw with few problems, you can fix the problems as you work along the saw.

If the saw has many serious flaws, you may consider a fix up pass. You joint the saw a bit more heavily than usual then file only the worst gullets. This will involve filing from both sides. You can then do a second light jointing if necessary and a standard filing.

Questions? Comments?

You can email me here.