Monday, September 15, 2014

Saw sharpening

La tercera es la vencida, we say in Chile. Third post has finally some tools on it.

This is saw I been filing for a few days. I know the angles are not perfect but that's the idea: practice.

I got the saw from ebay, in a really poor state, rusty and with several teeth missing on the cross cutting side. After some cleaning and a few passes it can cut relatively well ash and beech. Still need to make a handle for it.

So, the set up.

My saw vise is a piece of wood hold in my metal vise (5 euros in Leipzig, gotta love DDR). I file with a Slitting japanese file from fine-tools. The Barrette file on the back sharpens the saw next to it (the file was 1.50 euros in Leipzig, same guy, I also got some tool steel, and carbide tips from him.) The setup functions like this:

The saw rests on the piece of wood and I file with one hand. Yes, I could make a saw vise but then I have to take it to Chile next year and I got already too many tools. It works and helps me to train my hand to follow a straight line. Besides, in japan they do it too:

This guy is slightly more professional tho

How do I know if I am doing it right? I found out that (in almost a cheap-zen-koan-way) the best way to see the saw is not to look at the saw. Look at the reflection of a far away light. The rays come parallel and if all the teeth are at the same angle, their reflection is the same. Light scattering we call it in physics. As you can see on the first picture the first 10 teeth or so are quite parallel.

This is the ripping side:

You should focus on the tips of the even teeth, they all reflect light in a similar direction (these are the teeth with set going inside the screen). I saw this type of filing on my Azebiki saw and decided to try it. I liked it.

Here is the reason for the piece of wood in the vice. To make these nicks, or the third side on the cross cut tooth,  I file a small channel on the wood where the file can move, this way I don't file other teeth. It works most of the time. This guy has another, probably better, way of doing it: 

That's it. Focus on the reflection of the light, keep a steady hand, and use a support to file the third face of the tooth.

I still need to file down 3mm or so to bring back all the teeth on the cross-cut side, so plenty of practice left.


  1. Excellent! Finally someone is writing about Japanese saw filing! The small bit of sawing that I do is generally using a disposable blade Z-saw, and each time that I use it I feel a bit of sadness. The saw will never be as sharp as it is, right at that particular point, and will only degrade from there. I want to make the leap forward (back?) to handmade saws, that show evidence of the makers hand....... And that I can sharpen.

    I have an intermediate quality Ryoba that is new, but doesn't work well for me (slow and poor finish). It looks like the factory sharpening is a little bit rough, so a simple touch-up file work might help the cross-cut side. I also have a higher quality Ryoba that is in the mail, so it will be interesting to see how the saw tooth designs differ. My question pertains to the rip-side. Would you consider writing a blog post detailing the rip tooth pattern that you are using? I have been reading about the ChoMasaru tooth designs, but haven't tried anything new myself.

    You seem to be one of the few persons actually doing this sharpening. Thank you for exploring this, and giving me the encouragement to try this myself.

    Jason Thomas

  2. Hi Jason,

    I sharpened a small dozuki Z-saw in Germany, Opa has 3 blades there. The one he was currently using, a brand new one still in the box, and a very very old one that had lost all the coloring from the impulse hardening. I gave it a try with a normal file and it worked. At the end, it was cutting better than the used blade, but not as sharp as the one from the factory. Not so bad for a disposable blade.

    I don't know if the sharpening will last since I guess the steel is not hardened beyond the teeth, but it was nice to see that you actually could do it. Next time I'm back I will try to blunt it with some rosewood and see how much they last.

    Do you have the link for the ChoMasaru tooth? I cannot find it online.

    And yes, sadly there is not many people who's doing this at the moment, I would really like to hear from someone who actually knows what he's doing.

    I write right now a bit more on the rip pattern. And soon I hope to have back my other ryoba from Oliver, a jeweler friend of mine. I broke the saw and he's welding it... let's hope it will work again, if so I will also report on the welding process.


    1. Hi Sebastian!

      The "ChoMasaru" tooth (I'm not sure that the name is correct, because my information is filtered through Google translate, haha!)...... I keep seeing this guy referenced in the Japanese tool blogs that I read. It essentially is a crosscut tooth, except with a more vertical orientation, as opposed to being tilted forward. He also has a pattern that has "windows" that seems to work well for cutting on a bias. Fast, too. People seem to really like how it works. I'm going to try it out, when I decide which saw to modify (but I bet that you will try it before I can!).

      Here is the kanji I use for the search:


      Here is some of his tooth design philosophy:

      He claims that his design is faster, makes cleaner cuts, and requires sharpening less often..... Good, no? I don't know if he does anything unusual for the rip teeth, though. I would love some ideas, so I'm very interested in how you are shaping yours.

      You are already learning more than most people ever will, because you are actually DOING this! You will quickly learn what works, and what doesn't. When you do things for yourself, some mysteries are solved, but you will always find new questions to ask. Fun stuff!

      Thank you for your time,


  3. Gosh! that sounds like fun.

    I'm starting to study it already. As I said before I have still 3 mm to file to get rid of the broken teeth, so nothing lost if it doesn;t work.

    I need to make myself a drawing to understand why this geometry should cut faster. And if I am too bored at work, I could even make some simulations of it with my DEM particles and check which is the best geometry for cross cut and rip.

    Thanks for your words, it's really encouraging to know there is someone else interested on these magical matters.


  4. ok, it does seem to cut faster.

    Tomorrow I will try to take some pictures.

    You should have bet a beer, or even better, a sharpening stone :)