Reducing the set on my tenon saw – Part I (Becoming a butcher)

Posted by

This post forms part of the “Reducing the set on my tennon saw” series.

This started out as a attempt at minor surgery. Actually it should have been more like taking a couple of paracetamol; perfectly safe and simple enough that anyone could do it. By the end of the process however, it had gone beyond major surgery to what could best be described as savage butchery.

It all started simply enough. I don’t own a good tenon or dovetail saw, and have been vacillating between spending the money on a good new saw (which I’d have to sharpen down the track anyway), or get a used saw and learn to sharpen sooner rather than later, or build a saw from scratch.

Anyway, back to my current saw. It’s a simply Bahco tenon saw with impulse hardened teeth. For those that don’t know what that means, the teeth (not the whole saw plate) are hardened so that the saw stays sharp longer, but when it goes blunt, there’s nothing you can do.

I’ve been using this saw for a year or so, and while I’ve been able to successfully cut rather good joinery with it, the wide set makes the blade wobble in the kerf whilst trying to cut a straight line. In this case, the 0.75 mm plate was cutting a 1.2 mm kerf. You can see from the photos below how much play there was with the saw in the cut.

I’d read about this fantastic technique where you fold a piece of normal printer paper over the blade and squeeze the teeth between two steel plates in a vice. The idea being that the teeth pierce the paper as the vice closes but the plate won’t, so you end up with a set the width of a piece of paper. I did refer to several sources and all appeared legitimate (unlike fixing broken porcelain by soaking it in milk… but that’s a story for another day).

Because my vice jaws are rather the worse for ware, I’d need something a little more uniform. Looking around the shed I found some shelf brackets that I’d welded up from RHS. I thought that would be strong enough to do the job. After tightening it nice and hard, at the heal, mid section and toe, I unwrapped with an air of anticipation to see what the procedure had delivered. The tooth perforations on the paper were nice and uniform, so assumed sufficient pressure had been applied along the length of the blade.

Hmmmm, subsequent measuring showed 0.00 mm reduction in set. Not what I was expecting and a little disappointing if truth be told. Hmmm, what to do? And this is where things started taking a turn for the worse. Rather than meticulously researching my next options, as is my natural disposition, I recalled that I’d read somewhere that you can reduce set with a pair of hammers, and threw caution to the wind.

I dutifully clamped my large claw hammer in the vice, and gently, using my ball-peen hammer ran across all the teeth of my saw. One pass on each side.

The first test cut was great. I can’t tell you what the set reduction was, because I considered it pointless measuring as any teeth I did measure would provide an unrepresentative sample of the results due to the uncontrolled process. The kerf was noticeably thinner when viewed next to some initial control cuts.

And this is where the sad tale starts it’s downward spiral. There was still some set, and started wondering what it would be like to use a no-set saw (I’d been reading about cutting saw plates you see, and some people like having no set). So back to the hammers. Whoah! Once the cut was started (at the correct angle, hmmm), I could follow a line with eyes closed. This was great! The only issue was that if I did a deep cut, after a while it would try and track to the left and bind. No problem I thought, it just needs a touch up with the hammer on the left side to bring what errant teeth there are into line. And just to be safe, I went over the other side again too. It was about half way through the second side when I noted that rather than hammering to the same plane as the saw plate, I was angling the hammer into the plate. Surely I wasn’t creating the first tenon sword? As it turned out I was. A serrated blade is great for dispatching ones enemies, not so much when trying to cut wood however. Yep, within 5 mm of starting a cut, the saw started to bind. The teeth were now cutting a slot narrower than the plate.

Every downward spiral has a point where it becomes a vertical free-fall. This was that point. Realising I now had no saw that would work, and not owning a setting tool I realised that with nothing left to loose, it was MacGyver time!

Getting out my centre punch, I started the task of running down each side of the saw plate, and punching every second tooth. To be hones I really had no idea if I was salvaging my saw at this point, or just making noise for no purpose.

Once that was done, running my fingers down the side of the saw suggested that I had indeed introduced some set back into the teeth. Highly inconsistent no doubt, but there was some set. Back to the test piece.

To say I was astonished at the result would be under selling my reaction. This thing cut very nicely as long as I stayed within the front 60% of the blade. There was a horrid tooth that was projecting way out. Trying to knock it back into place simply snapped it off, but at least I can now use the whole length again. Now I should point out that I consider this a monumental fluke and “phew” moment. There was no skill involved, just blind luck. And I only have a temporary solution. I suspect this is a temporary state of my saw’s terminal lucidity. I doubt it will work well for very long.

Test cuts in spotted gum

There is however good to come out of this. I now have a saw that is in dire need of work, so plan to cut off the hardened teeth, and follow the process of filing in new teeth from scratch as a learning process (unless someone tells me that’s a dumb idea). If that goes reasonably well, I’ll progress on to making my own saw from scratch. So my emotional state went from anticipation to devastation to find a resting place at excitement.

Lance

May 2019

Leave a Reply

Your email address will not be published. Required fields are marked *