WoodCentral Forums

Est. 1998 — 27 years of woodworking knowledge

Fixing a Dovetail Saw

Posts

Fixing a Dovetail Saw

#1

Fixing a Dovetail Saw

Robby W.

>Hi, All -

Quite a few years ago, I purchased a Freud dovetail saw. It came sharpened crosscut with a moderate set to it. I used it for a bunch of dovetails with decent results, but, after reading a Krenov book, wanted it redone with rip teeth and no set. Sooooo, I took it to a saw shop that was fairly well respected, explained what I wanted to them and even got it written on the receipt. They said: "Come back in a week and we will have it just like you want it". Life is good.

When I came back a week later, they had sharpened the saw crosscut with a huge set. I was almost in tears. I asked the guy what they were thinking and even pointed out the requests documented on the receipt. He had the audacity to ask me to pay for it, but I just walked out with my wounded saw. Since it wouldn't cut a straight line any more, I put it in the drawer and forgot about it.

While cleaning my shop the other day, I came across the saw again. I would like at least take some of the set out of the teeth and get it back where it was.

Is there an easy way to do this? I think I remember someone saying you could clamp the teeth between two lathe bits to take out most of the set, but I have be leery of trying it. You help would be appreciated.

Robby Wright

San Marcos, CA

Re: Fixing a Dovetail Saw

#2

Re: Fixing a Dovetail Saw

Adam Cherubini

>Hi,

I think you have several problems, set probably being the easiest to deal with. I don't know how well you can clamp out set- the bigger problem is reshaping the teeth. I recommend a heavy jointing, followed by reshaping (just get the fleam and rake right) then joint again, and then file from each side. After all that jointing you will have removed much of the set.

How long is this saw? How deep is the blade (teeth to spine)? What is the current pitch? Assuming you haven't done much dovetail sawing, I think I'd recommend filing 5 degrees rake.

As for removing all set, its is my belief that people who advocate dovetail saws with no set TO HELP YOU SAW STRAIGHT don't know how to use a hand saw. So if you want to remove to set to crreate really small kerfs for really fine work, you're on the right track. If you think removing set is going to help guide the saw, you've received some bad advice. Know that people do use saws without set, place the saw flat on the end grain and go for it and actually do nice work. But this technique is a tremendous handicap.

Adam

Re: Fixing a Dovetail Saw

#3

Re: Fixing a Dovetail Saw

Robby W.

>HI, Adam - While I have done quite a few dovetails, I have to admit I am rusty now. The saw is about 8 inches long, is about 3" from teeth to spine and has 18 teeth per inch. It is a brass backed saw with a full handle.

I was not counting on no set to saw a straight line - I can do that on my own - but right now there is so much set that it makes a kerf close to 1/8" wide, and is very hard to control. I wanted a nice, narrow kerf.

All of that said, I do have to ask what fleam is. Rake I understand, but have never heard of fleam.

Thanks for your answers.

Robby W.

Re: Fixing a Dovetail Saw

#5

Re: Fixing a Dovetail Saw

Adam Cherubini

>yeah that's a great website. Fleam is the skew angle of the front of the teeth and the difference between a x-cut and a rip. You'll need a real small file for that pitch. I've never worked with teeth that small. You may need a needle file. Smallest saw file I have is double extra slim 4" and I'm pretty sure that would be too big for 18 teeth per inch.

I enjoy this sort of work. Too bad you live 2500 miles away. It might be fun for you to meet up with some West coast neander galoot centralian (whatever we call oursleves) WOODWORKER (how's that?) and do this job together and make a new friend.

Adam

Re: Fixing a Dovetail Saw

#6

Re: Fixing a Dovetail Saw

L. Hanson

>You've gotten some good advice here - I'll just throw in a couple pennies...

For 18 tpi a 4" Double Extra Slim file is small enough, IMO - if it's a good quality file, such as Nicholson or Simonds brand. Nicholsons I know are available from Lee Valley, among others.

The first thing I would do is put it in the saw vise and re-shape the teeth to a rip profile, all from one side - it's easier to re-shape them that way. It should only take a couple of swipes with the file to get them there. Then, when you have all of the teeth re-filed, hit every other one from one side with a light swipe of the file, then turn the saw around and give the remaining teeth a swipe, the same as you did the other side.

By the time you've accomplished this, you will probably have removed enough of the set that you won't have to worry about it...

If you only want to remove the set, you could probably side-dress the teeth with a fine grit stone (a cheap one, like a carborundum stone - because the saw teeth will wear grooves in it). Do a light pass on each side, then try the saw to see that it tracks properly, then repeat as necessary, testing with each step. Go too far, and the saw will bind, something you don't want. If the saw consistently tracks to one side, then dress the side it tracks to only and test, until it tracks straight.

I agree with Adam about no set - I'm not a big believer in that. I think rather than "no" or "minimal" set, a better term to use would be "properly" set. A properly sharpened and set saw will sing through the wood almost effortlessly.

If you want to read more after you're done reading the Vintage Saws primer, there's a page or two on my web site:

Sharpening hand saws

It may or may not bore you to tears, but might have something useful there too. I directly address tuning the saws after sharpening here:

Tuning hand saws

HTH

Leif

www.norsewoodsmith.com

Re: Fixing a Dovetail Saw

#7

Re: Fixing a Dovetail Saw

paul womack

>Smallest saw file I have is double extra slim 4" and I'm pretty sure that would be too big for 18 teeth per inch.

Pete Taran, who has worked on pretty much every size of saw, recommends 4" 2XSlim Taper as suitable for 15-20 TPI.

On a separate note, I recently sharpened a (rather thick - 30 thou) saw for dovetail use.

As an experiment, I use 8 degree fleam, 8 degree rake.

This worked out rather well, being easy to start, and easily aggressive enough to cuit a DT rapidly (it's a half inch cut; DT's are about control, not labour).

On reason for the fleam is that with very small teeth sharpened rip, and lightly set, it's rather difficult to see which was the set goes, so the fleam acts as a handy visual reminder.

A DT cut is NOT a normal rip cut, since a DT cut has the saw more or less straight across the end grain, whereas normally ripping is done around 45 degrees to the end grain.

The saw worked very well, and I will most likely convert my other DT saw next time I file it.

Oh - and the 30 thou saw leaves a 35 thou kerf.

BugBear

Re: Fixing a Dovetail Saw

#8

Re: Fixing a Dovetail Saw

Don McConnell

>Adam Cherubini wrote:

"As for removing all set, its is my belief that people who advocate dovetail saws with no set TO HELP YOU SAW STRAIGHT don't know how to use a hand saw. ... If you think removing set is going to help guide the saw, you've received some bad advice. ... this technique is a tremendous handicap."

I'm sorry I'm coming to this topic so late, but have been out of town for a few days.

Needless to say, everyone is entitled to their own opinions, but I don't believe stating them in such a prejudicial manner is very conducive to a thoughtful discussion.

I have found a couple of nineteenth century references to dovetail saws needing no more "set" than that left by the files when sharpening. Now, of course, it is possible that the authors were voicing uninformed misinformation (it does happen), but I have reason to believe otherwise. While learning cabinetmaking, I had a number of opportunities to talk with men (mostly British) who had undergone traditional apprenticeships in cabinetmaking. When the conversations turned to dovetailing they invariably mentioned "knocking out the set" immediately after purchasing their dovetail saws. Of course, maybe they never learned how to use a handsaw, but I was inclined to give them the benefit of the doubt.

Additionally, this was in line with what I was being taught. The idea was that the saw was not expected to make very deep cuts (minimizing the issue of binding), and the resulting kerf would be quite smooth and straight (if the saw was properly sharpened). It was thought that this maximized the possibility of producing pins and tails which fit straight from the saw, *if* the cuts were accurately started. Of course, one needs to become quite familiar with their dovetail saw and acquire a certain level of skill in order to utilize this approach.

Which is why I recognize that it isn't for everyone. Nor do I assume that someone doesn't know how to use a handsaw if they choose to take a different approach. But I do believe people deserve to know that the "no set" (actually, very minimal set due to sharpening and residual set) approach is practised by some people who actually know how to use a handsaw. All I will say is that I have not found this approach to be a handicap.

Don

Re: Fixing a Dovetail Saw

#9

I Agree

Todd Hughes

>I have heard many times that a dovetail saw doesn't need much or any set. Also On most of the small dovetail saws that I have found that were in very good original condition they basicaly didn't have any set to them and I believe that was how they were used

As you point out if you think about how a dovetail saw will be used a set is not really needed....Todd

Re: Fixing a Dovetail Saw

#10

Re: Fixing a Dovetail Saw

Adam Cherubini

>This is not a hill I'm willing to die on, but suffice it to say, and with my profoundest respect, I offer the following;

Needless to say, everyone is entitled to their own opinions, but I don't believe stating them in such a prejudicial manner is very conducive to a thoughtful discussion.

Strong statements start good discussions and mine has worked as intended!

I have found a couple of nineteenth century references to dovetail saws needing no more "set" than that left by the files when sharpening.

Like you, I saw entirely by hand. My experience has been that old saws with absolutely zero set produce kerfs wider than their blades' thickness (try it yourself everyone). I think this is caused by the natural waviness of hammer straightened saw plates (blades). We must be very careful about comparing these old blades with modern LN or Adria blades which to my eye are excessively straight. It could be that the old timers (be they British or otherwise) never truly removed 100% of their set. Also, a saw filed but not stoned would pull to one side if its teeth were not filed from alternate sides (or both). It could be the tooth's burr was sufficient set. Keep in mind LN sets 3 mil per side, so even set saws aren't set a lot. So maybe its easy for them to recommend removing set, but I'm not sure we should apply this advice to modern saws.

It was thought that this maximized the possibility of producing pins and tails which fit straight from the saw, *if* the cuts were accurately started.

Producing pins and tails which fit straight from the saw has nothing to do with set. As my friend Dean says, you could cut DT's with a chain saw and have them fit. The trick is only sawing to a line- nothing more.

Reducing set is a great idea. Maybe not for a beginner, but helpful for some specific joints including the popular "london pattern" dt. Its also helpful for thin woods and reduces clogging (or at least theorhetically).

Removing all set to stop you from turning a good start into a terribly bad cut is really quite silly when you think about it. In the same way it guides the good cut, it prevents you from fixing the bad. But most importantly, it conceals the consistently demonstrated mystery of "long makes straight". To fix a wayward cut or create a straight cut, lay the saw down until the work sees more saw. Very simple. Relying on the kerf to guide your saw is a recipe for frustration. That people do it or have done it successfully is not in question. Removing set is helpful, but not (in my opinion) a commendable (recommnendable) approach to sawing accurately.

Adam

P.S. I consider it a great honor to be read by Don McConnell. He is without a doubt my all tiime favorite galoot. I mean, anybody know owns an 18" Gabriel backsaw has got to be someone special!

Re: Fixing a Dovetail Saw

#11

Re: Fixing a Dovetail Saw

paul womack

>it[minimal set] prevents you from fixing the bad[cut].

As Ian Kirby sensibly points out in his excellent book, if you deviate from the initial direction of your cut, you must perforce create a faceted (or even curved!) cut face, which cannot possibly form a well fitting joint.

BugBear

Re: Fixing a Dovetail Saw

#12

Re: Fixing a Dovetail Saw

Adam Cherubini

>As Ian Kirby sensibly points out in his excellent book, if you deviate from the initial direction of your cut, you must perforce create a faceted (or even curved!) cut face, which cannot possibly form a well fitting joint.

Seriously Bugbear? Does this really make sense to you?

Said differently, rather than correct a miscut resulting in a tiny splinter sized gap (the facet), we should continue the miscut and create a tapered gap and shim it?

Adam

Re: Fixing a Dovetail Saw

#13

Re: Fixing a Dovetail Saw

Don McConnell

>Adam Cherubini wrote:

"Strong statements start good discussions and mine has worked as intended!"

Unfortunately, I don't find "the ends justify the means" line of argument to be very compelling. Very often, it amounts to justifying pretty much any kind of behavior/statement - in this case, maligning generations of anglo/american cabinetmakers. Unless, of course, you are prepared to argue that none of them "know how to use a handsaw."

"... My experience has been that old saws with absolutely zero set produce kerfs wider than their blades' thickness (try it yourself everyone). I think this is caused by the natural waviness of hammer straightened saw plates (blades). We must be very careful about comparing these old blades with modern LN or Adria blades which to my eye are excessively straight."

I have tried it, and my experience tells me that the kerf is slightly wider than the gauge of the blade, even on the new saws. If the older saws had any "advantage" in this matter, I believe it more due to the fact that they were often forged and/or ground to a slight taper (i.e. were sligthly thinner at the back than at the edge with the teeth).

"... It could be that the old timers (be they British or otherwise) never truly removed 100% of their set."

Already acknowledged as "residual set."

"... Also, a saw filed but not stoned would pull to one side if its teeth were not filed from alternate sides (or both). It could be the tooth's burr was sufficient set. Keep in mind LN sets 3 mil per side, so even set saws aren't set a lot. So maybe its easy for them to recommend removing set, but I'm not sure we should apply this advice to modern saws."

I would not recommend filing from one side only, except, possibly, for initial defining of teeth when retoothing a saw. Even at that, I *very* lightly stone the sides of the teeth after sharpening. If the "tops," or cutting edges of the teeth are uneven side-to-side, that will also result in a saw which tends to pull to one side. But this isn't just an issue with minimal/zero-set saws. I don't know why you would want to use a saw which didn't tend to cut straight, regardless of the set.

As I said, I have applied this advice to "modern" saws and have been quite satisfied with the results.

"Producing pins and tails which fit straight from the saw has nothing to do with set. As my friend Dean says, you could cut DT's with a chain saw and have them fit. The trick is only sawing to a line- nothing more.

...

"Removing all set to stop you from turning a good start into a terribly bad cut is really quite silly when you think about it. In the same way it guides the good cut, it prevents you from fixing the bad."

As Bugbear has already pointed out, if you are in the position of needing to "fix a bad cut" after it is well underway, then you've already left material where you don't want it, or have removed material which will leave a gap. So, yes, the idea of producing dovetails straight from the saw does have to do with set. Not sure how a chainsaw fits into this discussion, but if your friend can produce passable cabinet grade/scale drawer dovetails with his, then I'll have to take that line of argument a little more seriously.

"But most importantly, it conceals the consistently demonstrated mystery of "long makes straight". To fix a wayward cut or create a straight cut, lay the saw down until the work sees more saw. Very simple. Relying on the kerf to guide your saw is a recipe for frustration. ... "

Not sure how something can be "consistently demonstrated" and a "mystery" at the same time. (??) As previously indicated, the key to cutting dovetails with a minimum/zero set saw is an accurate start. In other words, any "correction" needs to be made at the outset. And it can be, as demonstrated by the Edwardian era "London dovetails". Both angled cuts which define the socket into which the pin will fit are started from the same line/kerf. This is accomplished, when starting the second cut, by registering the saw at the start of the already established kerf, then lightly beginning the second cut while holding the saw at the desired angle (or vertically if you've shifted the drawer side so the second cut is defined by a plumb line). Once started, the saw wants to track the newly established kerf in a predictable and satisfying manner. I only mention it in this context, because I believe it demonstrates the principle of this approach in a fairly dramatic fashion.

I belive starting a cut with the idea that it can always be corrected mid-kerf, and/or can be remedied by paring later, is the recipe for frustration. Yes, I have needed to correct cuts by paring, but I try to keep that to an absolute minimum. Experience informs me that the more paring/fiddling one introduces, the more likely one is to be dissatisfied with the results.

"P.S. I consider it a great honor to be read by Don McConnell. He is without a doubt my all tiime favorite galoot. I mean, anybody know owns an 18" Gabriel backsaw has got to be someone special!"

I genuinely appreciate the kind words, but apologize if I've said anything to lead you to believe that I own an 18" Gabriel backsaw. I have owned an 18" Disston backsaw, so mis-spoke if I said otherwise.

Regards,

Don

Re: Fixing a Dovetail Saw

#14

Re: Fixing a Dovetail Saw

paul womack

>Kirby teaches that when cutting the tails (which he does first) one should not try to "correct" the slope, for the reasons already mentioned.

He suggests that extensive practise (and correct technique) should render angular errors on the *vertical* pin cut most unlikely.

This continues to sound reasonable to me.

BugBear

👍 This page answered my questions

Your vote helps other woodworkers quickly find the answers and techniques that actually work in the shop.