I recently decided to act on my long-festering bandsaw itch. I’ve always thought it would be nice to have one, although I’ve gotten along OK without it, cutting curves with a jigsaw or scroll saw, or coping saw. But now I have more space, and a bunch of big dead trees, and would like to be able to do some resawing. So I poked around online and found a reasonably good-looking used 14″ saw. I probably overpaid for it (“or best offer” means you’re supposed to offer less, clearly), and got in way over my head with this thing, but I’m beginning to think it’s going to work out just fine, phew.

My unusually patient wife came with me to an industrial site in central Massachusetts where the saw was. I knew it would be heavy and expected to be breaking it in two pieces (base, where the motor is, and the top, which is a heavy cast-iron frame plus all the actual sawing bits), so I packed up some tools and we headed over there in our sturday but not very large Subaru Crosstrek. As it turned out, I didn’t need to worry about the tools as there was a well-equipped shop there, including a forklift and a roof-mounted crane that we were able to use to get the ~140 lb saw into the car in its two pieces. The seller did plug it in so I could verify that the motor ran and the saw would operate, but later I realized there were so many other things I could have and maybe should have also checked out.

We got this to our house, but it was too heavy to do anything about until my strong friend Brian showed up. The two of us managed to move the two pieces in a wheelbarrow out to the barn and set it up in a likely spot. Brian took a look at this thing and basically said I’d need to strip it down and see what might need some work.

I have to admit to some naive optimism – sure it was kind of old and beat-up looking, but who knows? Maybe it would all just plug in and run? So I spent a little while reconnecting the saw to its base, carefully tuning the table to be at precisely level and the blade guides to their 1/64″ clearance, reattached the belts connecting the motor to the pulleys on the saw, and then turned it on. Well, the light lit up, and the motor turned the wheels and blade, but only slowly (it was set up for metal cutting), and then it started to make horrible groaning noises! I spent a while cursing the people who had sold me this albatross, and then got to work trying to diagnose what could be wrong.

This saw has a shifting mechanism that enables the motor to be connected (via a belt) directly to the wheel to run at high speed (3000+ feet-per-minute) for cutting wood, or by way of a gearbox to achieve one of several lower speeds (40-335 FPM) for cutting various kinds of metal. Or at least that is what the manual says. When I tried to maneuver the clutch, I could not get it to move. At this point I started casting about online and discovered a community of enthusiasts, many of whom had posted photos, videos, and detailed descriptions of their own processes of saw rehabilitation. Many described jammed clutches, similar to mine. Some said it was due to pins being sheared off and getting stuck. Others talked about gummy oil residues. I honestly hadn’t expected this machine to have oil inside! Once I started removing pieces from the saw, taking the blade off so I could rotate the wheels independently, disconnecting the motor so I could check it separately from the wheel, I realized that the lower wheel was verry hard to turn. It seemed like it really ought to spin more freely. This, together with the evidence of the various rehab videos convinced me to open up the .guts of the saw and see what was stopping the clutch from shifting and the gears from turning.

The first thing was to drain the oil out. Annoyingly the drain plug on this saw is inaccessible when the saw is attached to its base?! So I had to disconnect these again, and hoist the upper part onto some sawhorses to get at the drain. It weighs a lot; more than 100 pounds I’m pretty sure, but I rigged up some ropes and a pulley and managed to swing the thing over. Then, after opening the drain, a drop or two came out. Then again, someone in his video had said it took more than a day for his to drain — I guess this gear oil is pretty thick? So I left it alone and bought myself a set of gear pullers for getting the various components off the axles or shafts on which they were mounted, and some fancy solvent called Kroil because I had misplaced my WD40 and someone said this stuff was better.

After a few days I came back and the oil seemed to have finish dripping out. There wasn’t much: maybe a quart, and I had added a little motor oil, while the saw was supposed to be filled with 2-1/2 quarts. You can see the inside (and the outside) was pretty filthy:

Check out those drilled-out impressions on the gear you can see in the picture. I think those must have been used to true the gear. There wasn’t a proper gasket, just some silicon caulking, and I suspect oil was leaking out of this thing all the time. The stuff inside smelled terrible too – honestly it smelled like baby poop after the baby ate spinach for the first time, plus it seemed to have a lot of grit in it, maybe metal shavings or something. So I cleaned all that stuff out as best I could, adding motor oil, rocking it back and forth, getting in there with newspaper and my fingers. I took the gearbox apart too, just to get its parts cleaned up as long as the whole thing was opened up.

The gear puller is off-center because the gearbox isn’t symmetrical, but it seemed to work out OK. Then I tackled the upper shaft, which has the clutch assembly on it that wouldn’t move, and is directly attached to the lower wheel around which the saw blade turns. It took me a while to figure out that the nut holding that wheel on is threaded backwards and needs to be turned clockwise to remove. Thank you internet! In the meantime, I sincerely hope I didn’t damage it too terribly by wailing on my crescent wrench with a hammer, sigh. After some precision bashing with a hammer and a piece of wood, I got this out:

If you look closely you can see these two cylindrical pieces with rectangular teeth that look like battlement crenellations. Those are called Lovejoy couplings, apparently, and are supposed to mate with the big gear on the inside (the one that has the divots drilled out of it, shown above), or, in its other position, a pulley (not shown here) that mounts on the outside of the box and connects to the motor with a belt. But these suckers just would not move. At first I could not see what was preventing them. My initial theory was that one of the roll pins that goes in the holes in those Lovejoy couplings, and then into a slot in the outer shaft, and finally (I think? I couldn’t see) into the inner shaft that is supposed to slide up and down with these couplings, and holds the whole thing together, more or less, was broken off somewhere deep inside. This theory was prompted partly by various internet stories of similar things having happened, but also by the fact that one of these pins, unlike the other, was not flush with the surface, but appeared to have been driven deeper in. After some fruitless minutes spent attempting to drill out one of these pins, I started looking for other theories. Luckily my drill bits were not up to the task. Then I spent a while trying other various things, staring at the object, turning it around, scratching my head, returning to the internet to look at other peoples’ disassembled assemblages of parts, and it finally dawned on me that mine had an extra part! It looks so natural and unassuming, I hadn’t even noticed it, but there is a bushing, or spacer, in between two of the parts that just does not belong!? Some truly sadistic person had gone to the trouble of taking this whole business apart — they would have had to remove the aforementioned roll pins, not an easy task (although it would have been easier when the whole thing would slide properly since they probably could have been driven out with a punch when properly lined up with the holes on the opposite side), and adding a new piece which made it literally impossible to undo. Maybe they intended to disable the clutch so the machine could only and ever be used for metalworking? Or so it seemed to me. I find it difficult to conceive what this was all about. Anyway I managed to cut away the offending bushing with a hacksaw.

I did only minor damage to other parts in the process, cringe. After all that, and lots of Croil and TLC and oil, things did start to slide, halleluiah. Then I put the whole upper shaft back in, with its inner gear, and the lower saw wheel, because I just had to see if things were going to spin at high speed. I tried to share the vertiginous victory video I captured on my phone, but I am not sufficiently skilled to rotate, crop, rescale, transcode, and compress it sufficiently for your viewing pleasure! But please take my word for it – it seems like it’s going to work. I’ll try to post some more satisfying clips and pics later.

Some super-helpful resources I found while trying to figure out what the heck was going on:

• Video of someone taking apart this saw and fixing the clutch
• Thread on owwm (“old woodworking machines”) about rehabbing this saw
• Another thread with pics about the same model
• A sketchy scan of the manual

It took me a couple of weeks to get the finish done. Coating the canvas with butyrate dope was pretty easy, although I learned that the ventilation in the converted garage room was not all that great, and ran around opening all the windows in the house to flush out the toxic fumes; not before I was discovered though. The thing that really took a while was the varnish I layered on top of that. The book proposed coating the transparent undercoats of dope with more layers of pigmented dope – paint, basically. But I wanted the my dye job to show through, so I was committed to use a transparent finish. I guess I could have just use more layers of dope, but I used a whole gallon on three layers, and this stuff was pretty expensive – around $80 for the gallon.

So for the final layers I used spar varnish, which is a heavy-duty traditional oil-based varnish. It’s also about $80/gallon at my hardware store, but I was able to apply three coats using only a pint of the stuff. Unfortunately the first coat I applied using some old stuff I found in the basement. It seemed OK underneath the top layer of hardened semi-gelatinous varnish, if a little dodgy. It was kind of thick, so I thinned it with some paint thinner and applied with a rag. It went on like runny snot and didn’t seem to want to dry. Maybe because the temperature dropped into the 40s and it was raining. Varnish likes it warm and dry. I ran the heater and the air conditioner (to pull moisture out of the soupy air), and kept on checking. After three days it was still kind of tacky. I rubbed it down with paint thinner, scraped off the worst bits of goo and bought some new varnish and Japan dryer, an additive that is supposed to help varnish dry in colder wetter conditions.

I think in the end it came out nicely, although it’s far from professional standard as such things go.

Here’s a full view from the top. I’m pretty happy about the slightly irregular dye color.

And finally, it’s a she, and she floats!

How does she float? Well, I am really pleased. No leaking, that’s good. Much tippier than my old plastic tub, but I got used to this pretty quickly while paddling. The fit is kind of tight, but it seems right. I can securely brace my thighs against the Masik and paddle comfortably. After 45 minutes of paddling though my left foot was going numb. I have the same problem when running sometimes; bad circulation I guess. I hope that I will get used to the position, but we’ll see.

Generally speaking, she tracks quite well. Turning is a more deliberate process than with the flat-bottomed 9-footers, which spin like tops. There is a nagging tendency to pull to the left though, that really does need to be corrected. I think the stern piece must not be quite straight. So I’ll probably add a skeg, an external fixed rudder-like attachment, to get that worked out.

All-in-all a success! I look forward to many more days of paddling.

Steam-bending is a voodoo woodworking art. If you heat up wood, and it has enough moisture in it, the fibers will loosen enough that a previously stiff board can be bent into a new shape, which it will retain after cooling. At least some wood will. How well this works depends on a lot of factors including the species of wood, whether it is clean (without knots), whether the grain runs nicely parallel, how dry it is, whether it has been air-dried, or kiln-dried, how long you steam it, whether you pre-soak it, in water, or in water plus fabric softener! Some youtubers will tell you how easy it is, but even they keep extra pieces on hand for when they break one. Many years ago I had tried bending guitar sides over a hot pipe (no steam required), and that went OK, so how hard could it be?

I went to a specialty lumber store and bought some ash (kiln dried, as everything is that you buy at a lumber yard). I had read that ash was a good species for bending, even kiln dried. The gold standard for this kind of work is white oak, but supposedly only when it is green, or lightly air dried, and I just didn’t have access to anything like that. Following the book, I concocted a steam box from a sheet of styrofoam insulation and a lot of duct tape, and hooked it up to our old hot pot using some PVC pipe left over from a bathroom renovation.

I did some trial bending runs to get an idea how long I would need to cook the ribs and settled on something like 15 minutes. The idea here is to work efficiently by staging all the ribs, loading the steamer with a batch of 4 or 5 inserted at one minute intervals, and then inserting a new dry one at the left every time you take a cooked one out of the right.

I won’t go into the whole bending process itself – there are some great videos from pros out there you can watch if you care to. I’ll just say that it was nerve-wracking and fun, and it came out OK in the end, and even if some of the ribs did split out the grain here and there, nothing too terrible happened. One thing I might do different if I try this again though, is  to pre-soak the wood. I decided not to this time, following the advice of one knowledgeable-sounding site I found. That guy had nothing but scorn for the idea. Now I’m not so sure – it might have helped. I had to apply a *lot* of pressure to bend these ribs, and I get the idea it should have been easier. You can see one of the several failed attempts at the bottom of the photo. The room smelled like hot styrofoam the whole time.

The ribs outline the shape of the hull. Now we have really defined how the kayak is going to eventually sit in the water. Its hydrodynamic characteristics will all derive from this shape.

The next step is to fit out several longitudinal pieces: the keelson, which runs the length of the boat at its lowest point, and the chines, which run in parallel on either side. These all come together at the bow and stern where they meet two other pieces that form the pointy ends. The book had a fancy name for them, but for the life of me I can’t remember what it was. In this photo you can see the nifty lashing that holds these pieces to the ribs; the keelson is done, here, and the chines are being held in place with clamps while they are positioned for optimum boat shape (they will hold the skin just off the ribs).

Here’s the pointy end I was talking about.

Now it really looks like a boat! Just needs a few little odds and ends, and then a skin!

Oh I forgot to mention this cool part of the Greenland kayak design, the Masik. This is a curved piece of wood you brace your thighs against when paddling. I made this one out of a curved maple branch that fell off the tree in front of my house. You can see that tree’s shadow on the bricks there. It’s a giant, starting to die, but it still explodes in a riot of red, orange, and yellow every fall and buries us in its leaves. The piece I made the Masik from was starting to rot a bit, and it has some neat figure in the grain that I liked a lot. Shaping it was a fun challenge. I roughed it out with a chain saw and then went at it with a variety of hand tools. In this picture it’s still in a pretty rough state, but you’ll have to take my word, it worked out nicely.

 Oh no! Somewhere along the way this gunwhale developed a nasty-looking crack. More fallout from forcing them into too dramatic a shape? Oh well, it doesn’t go too deep – I hope it’ll be OK? Better get the skin on soon, that should help to hold it together, right.

I decided to skin the boat in canvas, and coat it with butyrate dope like an old airplane, in keeping with the generally old-fashioned gestalt. Sealskins seem to be out of fashion, and anyway, unattainable. Most people seem to use nylon at this point. It’s lighter and probably easier to coat since you can use something easier and cheaper like epoxy or urethane or some such thing. Cutting it is annoying though: you need a hot knife to melt the edges to prevent it from fraying. Canvas was fun to work with. I learned how to sew. A thimble was essential to get the big needle through several layers of the stuff. The thread is dental floss!

The whole process of stretching the canvas on, trying to keep as much tension as possible, was pretty fascinating. The traditional system is to lace up a very rough seam, like a giant shoelace, along the centerline of the deck, draw it tight, and then sew up with fine stitches. The book I was following had a different idea, which I tried, stapling the fabric to the gunwhales to tighten it, and then stitching along one side of the deck.

Oh I almost forgot to mention. There was this thing that drove me totally crazy right at the end of the project. Another steam bending step was required in order to create the hoop that forms the cockpit, drawing all the fabric together around the place you sit. The idea was to make this hoop by bending a single six-foot stick of ash around a form. I tried this three times and broke the darned thing every time. Maybe I didn’t steam it enough at first? This piece was bigger and thicker than any of the ribs. One time I think there was a weak spot in the grain. The bending itself is tricky – you are holding this thing under a lot of pressure, working under time pressure since the wood is only pliable for a short time after you take it out of the box. Finally, even the fourth time it broke, so I finally decided to make it out of two pieces, and glue them. Sigh, the only glue on the thing. But it got done

Once the stitching was done, and I had made the cockpit hoop, I died the fabric. You can see my first crude stitches on the stern before I figured out how to get the seam right on the corner.

Getting the skin stitched into the cockpit hoop was an exercise in trying to tension up the loose flaps in the middle of the deck. Following the system proposed in my book was pretty straightforward and a little magical; I drilled holes all around the hoop and stuck nails through the fabric just below the holes, levering them up and pushing them through in order to tension everything. Then I laced a length of cord through all the holes, removing the nails, and the twine draws everything up nice and tight.  Here’s a picture of the hoop with the nails holding the fabric to it:

The last thing is to coat the whole job in cancer-causing chemicals to make it watertight. As I said before I’m using butyrate dope. This is the stuff you may have sniffed if you ever made a model airplane. It is a toxic stew of VOCs (volatile organic compounds) – basically every chemical ending in “-ene” is in there. I bought a respirator, don’t worry, my glue-sniffing days are over.

Next post should be launch day. I’m super excited to see if this thing actually floats, how tippy it will be and whether it will track straight. Oh and hmm I guess I’ll need to name it (her?). That will be the hardest part.

>> Launch Day

OK it’s 2018 and it is officially time to stop hosting blogs in your basement. I’m moving this blog and various other stuff out of my laundry room server and into the cloud.

The move promises to replace the nagging fear of attack by bitcoin miners, death by power outage, hard drive failure, or flooding with a $3.50/month bill. That’s what AWS Lighstail hosting costs today for its very lowest tier. I also plan to restore my vanity email domain using this server, since for a few years now I have been unable to maintain that at my house due to Verizon’s SMTP port blocking. I’m also planning to move some larger archives (music, pictures, etc) to S3 and relay them here using S3-FUSE. I think this micro-host might just be enough.