Radial Arm Saws
No video - but an entire book
Scroll down to: Download the book
Principles of Precision - excerpted from the book
If you own this book and have a question, or if you just what to say what you think about the book, add to "Your Blog comments" at the bottom of this page.
If you are not sure you want to buy this book, please read the BLOG at the bottom of this page. You will see what other woodworkers have to say about the usefulness of this little publication.
Anyone who has worked a few days with a radial-arm saw knows that the manufacturers' manuals do not show you how to get the machine to work as precisely as needed for good woodworking. The text books on radials tease you with accessories, jigs and projects but still don't give you much help in setting up your machine.
In the early '80s I presented seminars on the use of the radial-arm saw to craftsmen and hobbyists across Canada. The first hour of each seminar was dedicated to these problems of line-up and precision. No, you will not find videos on the Radial Arm Saw here -- I never shot any video for that back in 1980. But I did write a book.
This book is, in fact, the refined product of all those teaching sessions. In plain language, it gives detailed instructions on making the notoriously sloppy radial arm saw cut dead square and to a tolerance of one 128th of an inch with a simple saw blade. This book is the necessary companion to all other information on your radial-arm saw and I like to think that it is helping craftsmen everywhere to keep their solid old saws working.
Since 1987 when this book was first published by Lee Valley Tools, it has been recognized as the bible for setting up and using radial arm saws in a precision fashion. But with the advent of the sliding miter boxes, and the fact that few people ever used their radial arm saws as precision instruments, the use of this wonderful machine has declined. Some manufacturers have dropped out of the market; others don't take their saws very seriously. So, although we sold 5,000 copies of this very specialized book, sales were not strong enough to warrant a new press run.
But then along came cyber space and the potential for web publishing or at least electronic distribution of books. This has opened the door to cost effective publishing of very limited quantities of a given book, and hence the revival of out-of-print books. Essentially all I have to do is the layout, and you print it yourself, all of it, or only the pages you want. You will find that if you can print in 600dpi, the photos will come out as clear as the original publication. Since then a few people every week ask for the book -- so here it is available as a download pdf for purchase.
How to best print the PDF
The one publishing bug is that the book layout was designed for left and right pages (even pages on the left - odd pages on the right), rather than consecutive pages out of your printer. Hence the text and graphics make good sense when you see two pages side by side. Knowing that will help if you get lost in the text. The best copy of this book can be had by printing both sides of thick paper stock and then recreating the left/right layout of a book by putting it into a three ring binder. The blank pages were left in so that such a layout would work. -- but see the note below on Version 2.
I read through this book twenty+ years later and found that nothing needed to be updated; those old saws just keep on cutting. Hence the rest of this book is as originally published. For a while I sold it on the web, then as part of a DVD, and now here it is back on the web.
Well "nothing has changed" is almost true. Thirty years after the first printing, readers pointed out a problem with modern PDF readers that required that I adjust a page, so that when viewed in two-page (side-by-side) view, as the original printed book, the graphics lined up properly. (If you have a PDF program that will allow you to delete pages in the first PDF version – identified as the Electronic reprint 2000 - delete the second page (a blank page) in the version 1 book, identified as page “0”. Things will now line up as in Version 2.) And on page 64 I added in a great critique/suggestion from a wood worker. So as of November 2021, there is a version #2 – identified as such on the credits pages.
Buy the book
Just click here to go to PayPal to pay for it ($14.95).
After paying, you will be sent an e-mail with the download link. We don’t give you the download directly because some browsers have problems with downloads, pop-up windows and the like. If you don’t get an e-mail quickly, check your spam for a letter from firstname.lastname@example.org.
An electronic book is easy to copy and pass around, or even post on the web. I appreciate that most woodworkers recognize the labour that goes into the production of something, be it a chair or a book, and that they respect the copyright on this book. Thank you for copying the link to this book, not copying the book itself or taking rip off copies found on iressponsible British or Russian sites through Google. I get a great deal of satisfaction hearing that this book has helped woodworkers and continues to do so since when it was first published in 1987. Somehow that seems to be a statement against the planned obsolescence so common in tools today.
There are many forums on the web related to the Radial Arm Saw, one of the best is Paul Snotzalot’s DeWalt Radial Arm Saw forum – which covers more than DeWalt. http://forums.delphiforums.com/woodbutcher
Chapter 2 of the book has some information that applies to many tools beyond the Radial Arm Saw, particularly to the miter saws and sliding miter saws. I offer the whole chapter to you here. The rest of the Radial Arm Saw book deals with specific nuts and bolts adjustmements and jigs to attain this precision.
Accurate results can only come from precise work. In this sense the "precision" of the machine and of your working techniques means the ability to work to close tolerances. Cutting wood to close tolerances at the correct dimension will result in accurately made, well fitting joints.
Although craft books dealing with such delicate manual tasks as dovetailing and marquetry frequently deal with the details of precision, the subject is rarely discussed when dealing with power tools. Yet how can we expect to understand the importance of detailed adjustments and techniques unless we understand their role in obtaining precision from our radial arm saw? My radial and I work with six principles of precision. I will outline these principles here and then apply them in detail throughout the rest of the book.
Principle 1 -- Snug Movement
"Things that move must move smoothly and firmly."
Whenever wood and a machine come together they need to be in firm constant contact. Vibrations in the machine will mean a clattering contact between the cutting edge and the wood. A hand plane with a thin blade digs in and out of the wood grain. A jig saw who's shoe is allowed to rise off the wood will not give a clean cut. Grime, dust and rough surfaces make consistent contact equally as impossible.
On the radial arm saw the biggest offender of the principle of SNUG MOVEMENT is the roller head. It must not move freely, nor even easily but rather it must move firmly. That is not to say that it should bind, but a loose roller head is the major cause of the "running forward" feeling with most radials. When the roller wheels are adjusted loosely on the overhead track, the entire motor assembly has a tendency to lift up during a cut, but because of the offset blade it does not lift straight up nor does it necessarily stay up. The entire motor comes rattling forward, with the saw blade climbing up on fibers it should be cutting. The cut is ragged and your right arm is quickly exhausted not knowing if you should draw the saw forward or hold it back. When it is adjusted for "firm" travel you rarely have this problem. You must however clean the track several times a day (exactly 23 seconds of work) as the least bit of dust or even grime accumulation will block its movement and make drawing it forward firm but jerky. We want firm and smooth.
The column must move up and down smoothly and firmly in the base casting. Too much freedom here can cause the arm to pivot up or sideways or both. Too tight and it binds and becomes jerky in movement.
Similarly, the arm to column connection must be firm and free from slack.
Equally important is that when the wood moves, it too must move firmly and smoothly that is why I include hold-downs as part of the fine tuning of a radial arm saw.
Principle 2 -- Positive Clamping
"Things that should not move, must not move."
Almost all adjustments on woodworking machines hold their positions by means of friction clamps; two surfaces that bind together relatively solidly by pressure alone. Hand planes as well as power planers hold their blades by friction. Table saw depth adjustments are locked by friction. The radial arm saw relies on friction all over.
Whereas bolts and pins lock things absolutely into position, friction clamps hold them "positively". This means that they should clamp sufficiently solidly to prevent the pressures that the saw will exert on them from moving their position, but at the same time they should not be over tightened to either harm the machine or tire your hands.
The prime example is the arm to column clamp (the miter clamp). If the arm is set at 30° on the miter scale and the clamp is engaged, it will always be possible to force the arm to 35° despite the clamp because of the tremendous leverage you have on the end of the arm. To try and tighten it so that it would be impossible to move would rapidly wear the cam parts or break a pivot somewhere along the linkage. It must be set tightly enough to effectively resist the lateral forces of the cutting action of the saw. These forces are in fact rather minor except when ripping. This exception can be dealt with by proper use of the fixed 90° indexing as will be explained in the alignment procedure.
People tend to tighten the bevel clamp so tightly that the handle eventually breaks and gets replaced by a pair of vice-grips. The common cause of this is that they never did understand that the 90° index point can be calibrated to be dead on, so they kept trying to set the bevel to 1° or 2° in an effort to get a square cut. But at 1° it clamps very poorly (the beveled indexing pin itself is fighting the setting) and overkill becomes the standard procedure.
The use of sandpaper in certain circumstances can help to achieve positive clamping without undue pressure as is explained when considering the sandpaper miter fence and the wooden feather boards.
Surprisingly enough, even chipping on the bottom side of the cut can be eliminated. It's a question of clean back-up and eliminating vibration between the wood and the table as with my "fence hold-down" jig .
Principle 3 -- Use of the End Point of Slack
"Slack always has an end point, use it."
Ignoring slack in screw drives and indexing pins leads to sloppy results. You must take up the slack right to its end point to maintain precision.
The arm to column indexing pin must have some slack in its adjustment or the pin simply will not pivot in and out of its indexing slots. This means that when the pin is indexed to 90° the saw may very well be cutting at 88° or 91°. In fact many people find it hard to get the same result twice. If, however, some clear point within that slack is defined and calibrated to exactly 90°, then the saw could be set to exactly 90°every time despite the slack. I seat the pin into the column and before clamping it tight I shove the arm to the right until it comes to a full stop against the indexed pin. It is this "end point of slack" that I calibrate to exactly 90° and it stays there, waiting, dead on every time.
When a screw drive is used in any machine there is always slack between the screw threads and the nut threads. Although the threads may be an exact dimension (such as 1/8" per revolution), changing the direction of rotation will throw the slack into any effort to use screw pitch for depth measurements. If turning is done only in one direction you will be sitting firmly on the end point of slack all the way and depth adjustments can easily be made to 128th of an inch (if the column is sliding smoothly and firmly) by simply watching the crank handle position. With the blade in the horizontal position this allows for cutting surprisingly accurate box joints without either measuring or marking the wood.
Principle 4 -- Squaring to Actual Work Surfaces
"Align a machine at its production end."
Too often instructions say to square the saw up to the table rails or some other reference point that is not the same surface that will be holding the wood. Even lining up the table and then attaching a covering can change the angle of the cut in the wood. It is the final table surface and the final fence that should serve as reference points for alignment of the saw.
If temporary tables or fence additions are added, the actual angle of cut in the wood should be verified before trying to produce accurate results, since the precision of your machine may have been compromised by inaccuracies in your temporary additions or their installation.
Principle 5 -- Verification on Cut Wood
"Rough static alignment is done with instruments on the machine; dynamic fine tuning is done with a blade cutting wood."
It doesn't matter if a jointer, for example, looks good. What counts is that it produces straight square boards and this is always the final test of alignment.
There are two reasons for using wood cuts, not metal squares, as the fine tuning instruments for a radial arm saw.
Firstly, by using manipulation tricks when cutting wood, you can adjust the saw square to the table for a 6 inch depth and square from the fence for a 26 inch cut -- twice the actual cutting capacity of the machine. This means that within its actual cutting capacities it will be twice as precise as can possibly be measured with metal squares.
Secondly, the end product is really what counts so I don't care if something may look out of line on the machine as long as the results of the blade passing through wood gives me the perfection I demand. This difference in apparent alignment and actual cutting results can be due to the fact that the metal square by necessity gives a static picture of the machine's alignment. The blade engaged in wood gives a dynamic measurement identical to actual working conditions.
Principle 6 -- Definition of Entry Before Cutting
"Know where the blade will cut before cutting."
We often scratch up the edge of a piece of wood or create considerable scrap trying to determine exactly where the point of entry will be for a cutting edge. Whether a drill press or a table saw, we are used to a lot of trial and error.
The fence on a radial arm saw is usually a sorry looking chopped up mess, often hiding down below the wood to be cut. The gaping hole in the vicinity of where the blade passes through provides no backup against splinters on the back side of the cut and certainly is useless in guessing where the blade will strike the work piece.
You must get used to the idea that the fence can be shifted an inch or two to the right or the left without really affecting its function at all as a back stop. This shifting allows you to put it in its normal position during line-up and general work, and then, when you want to get down to the business of precision, slide the fence over to a clean spot and make a nice fresh kerf cut through. This new cut defines quite precisely both sides of the cuts you are about to make and backs up against splintering on the backside of the work piece at the same time. If in addition the fence is about 1/4 inch taller than the wood you are to cut, this slot in the fence becomes a precise index against which cut marks on the work piece can be accurately placed before ever coming near them with the blade. See the graphic.
(Scroll a little to see blog comments below)
Work safely -- Jon
Learning Curve 75
Add A Comment