I am now selling Grobet Swiss needle and Bahco taper saw files, which are among the best files made today. While I don’t carry every size (and certainly not as many as were available decades ago), the range is sufficient to cover most tooth pitches you are likely to encounter. The following post will, I hope, give the novice filer the confidence to get started.
A common source of angst for new sharpeners is choosing the right file for sharpening their new (or new to them) saw. What follows is nothing new, but a very modest compilation of background information that may allay some of that trepidation.
We will concern ourselves in this post with only triangular tapered saw files (hereinafter to be simply called files). These are the files used to sharpen western handsaws and backsaws. The files required to shape and sharpen two man crosscut saws, Japanese saws, and others are a different topic altogether.
FILE GEOMETRY: All saw files share some basic geometric similarities. Examine the file in the drawing below. The cross section is an equilateral triangle with radiused corners. The three faces and three radii are all toothed. The cross section gradually increases in fullness as one moves away from the point, reaching its maximum size approximately one fourth to one half of the way down the file body. At this point, the cross section remains constant until the tang is reached. Since a file sharpens both the front and back of adjacent teeth simultaneously, the cross section of the file is imparted on a saw’s teeth. For this reason, the points of all teeth and the base of the gullets are always 60° angles, no matter the rake (see footnote at the end of this post).
Very basic taper saw file geometry.
The role that the taper plays in a file’s usefulness is the topic of some debate. I have never found or used a non-tapered saw file, but they are shown in some older catalogs. Nicholson (and perhaps others) sold both blunt (non-tapered) and tapered saw files, noting that they were often preferred by carpenters and other expert saw filers. This implies that tapered files may be easier to use; not having had the chance to compare the two, I cannot speak with any authority on that topic.
Scan showing taper and blunt saw files. From File Filosophy, Nicholson File Company, 1949.
FILE NAMING AND SIZING: While all saw files share the basic geometry outlined above, they are differentiated by their length, thickness (or width of a face), and cut.
- Length: Saw files come in at least seven different lengths, 4″, 5″, 6″, 7″, 8″, 9″, and 10″ being the most common. The length is always measured from the point to the beginning of the tang.
How the length of a tapered saw file is measured.
- Thickness: The nomenclature for this dimension is a bit odd. Each length of file comes in four different thicknesses (listed in order of decreasing width): regular, slim, extra slim (X-slim), and double extra slim (XX-slim). When considering a particular length of saw file, the regular taper will have the widest face, and the XX-slim the narrowest. For a given thickness (say, slim), the longer file will have a wider face.
- Cut: Saw files are either single or double cut. Single cut files have a single row of diagonal teeth, while double cut files have two rows of diagonal teeth at opposing angles. A single cut file cuts somewhat slower, but leaves a slightly smoother surface than a double cut file. These differences are usually slight.
Single vs. double cut files. From File Filosophy, Nicholson File Company, 1949.
CHARACTERISTICS OF GOOD FILES: Much has been written of the decline in availability of good saw files. Having used a wide variety of both recently manufactured and new old stock (NOS) files, this decline appears very real to me. The quality of files available today ranges from very poor to very good. Short of actually trying out a potential file, there are a couple of things you can look for as an indication of quality.
- Surface finish: The teeth in files are not cut by removing material, but by raising the metal from the surface of the file blank with a chisel. As a result, any irregularities in the surface of the file blank will be present at the tip of the tooth. Examine the body of the file at its tip and above the tang to see how the blank was finished. The smoother and more polished this area, the more care was taken in preparing the blank, and the smoother the file will cut.
How a saw file is made. From File Filosophy, Nicholson File Company, 1949.
- Straightness: Inferior files are often bowed or twisted along their length, which makes them more difficult to use in a controlled manner.
- Durability: This quality, unfortunately, can only be determined after use. A good file will dull gradually as the teeth slowly lose their edge. A common mode of failure in inferior files is fractured teeth along the radiused edge, where the pressure is greatest. After losing just a few teeth, the problem cascades quickly, and the file rapidly becomes useless. Files are particularly prone to this failure when filing in new teeth by hand. As teeth are lost, the file begins to jump around, making it very difficult to maintain an even spacing. In very short order, the file loses ever more teeth, and soon ceases to cut at all.
For cutting in teeth, new old stock (NOS) files are an excellent choice. With one exception, I have found no modern saw files that will stand up to this task as well as they do. Grobet Swiss needle files are the exception; these files are just as durable (or even more so) as any NOS that I have found. They are, unfortunately, not large enough for teeth larger than about 10 or 11 ppi. If you are simply maintaining sharpness and not doing major reshaping, then the range of acceptable files made today broadens, as this is not nearly as strenuous a test of a file’s mettle.
CHOOSING THE CORRECT SIZE OF FILE FOR YOUR SAW: The most common advice given for choosing a file to use on a particular saw is to select one whose face is twice as wide as the edge of the tooth being filed. When one edge has dulled, the file can be rotated to expose fresh teeth. In this way, all three corners and faces of the file can be fully utilized. Guided by this advice, it is seen that filing from both sides of the saw requires a larger file than working all the teeth from one side.
Showing the traditional method of selecting the proper size of saw file.
While this advice is a decent simplification, it quickly becomes apparent that strict adherence to it requires that many sizes of files (requiring considerable expenditure) be kept to sharpen a well-stocked till. Fortunately, further examination and refinement can lessen the required outlay.
Implicit in this rule of thumb is the assumption that a file wears evenly over its faces. In my experience, it does not. Rather, the corners of the file dull or fail first, even while there is still life left in the faces of the file. Rotating the file to a fresh corner restores the cutting ability, even though a portion of the face that was previously used is asked to continue cutting. This overlap allows one file to work for several different tooth pitches.
Another consideration in selecting a file is the corner radius of the file. This radius increases as the length and width of the file increase. In extreme cases, the radius of the file can actually be larger than the front edge of a tooth. In this case, the actual rake of the tooth will be greater than expected, leading to unpredictable and unsatisfactory performance. This phenomenon is much more pronounced in smaller teeth (about 13 ppi and finer).
Furthermore, since the gullet takes its shape from the file, a larger file (with its larger radius) leaves a smaller gullet. With less space to carry sawdust, the saw will not clear the cut as well as one with a larger gullet. In general, gullet size and corner radius are of greater concern with small teeth and files, while face width is the more important consideration on larger teeth and files.
Largely due to these two observations, I now use three square needle files for filing small teeth (11 ppi and finer). Their small corner radius makes for precise and predictable results, while maximizing gullet volume.
Showing the effects of different files on gullet formation.
One last point: a file with a narrower face gives better visibility of the tooth when filing than does a larger one. It is a minor point, to be sure, but any advantage that can be had when straining to see small teeth is appreciated. There are any number of tables and charts out there that match saw files to the ppi of the saw you are sharpening. There is some variation between these recommendations, but by and large they can be trusted to yield good results. The best bet is to follow the recommendations of the vendor, since not everyone sells the same range of files. The table below is what I tend to use. Bear in mind that, for the reasons outlined above, I usually prefer a slightly smaller file than most recommend.
Saw file recommendations.
FILE LIFE: Some tools are a lifetime investment. Saw files are not. Chances are that if you are wondering if the file is dull, it reached that point some time ago.
Flippant answers aside, this is a tough question. Broken teeth along the edges of the file are easy to see and feel, but dulling of the teeth is more gradual and less obvious. If you need to press down too hard to keep the file from skating across the tooth, you diminish the accuracy of your filing and frustrate yourself needlessly.
RECOMMENDED FILE BRANDS: If you can find NOS saw files, these can be a great value. Nicholson, Simonds, Disston, and others once made top-notch files. They no longer do. The supply of the NOS is, unfortunately, unpredictable. Restricting the discussion to new files, the field narrows considerably. I use Grobet Swiss three-square needle files for smaller teeth, and Bahco files for all others. Pferd files are nearly the equals of Bahco. I have sampled a few files made by Baiter (sent to me by a friend), and was pleased with their quality.
Footnote: Technically, these angles are only 60° when the teeth are filed without fleam or slope. When fleam and slope are introduced, this angle changes.