365 Bowls

“You’ve got rolling pin problems”.

“I knowthat”, he said.

Am I right?

< crickets >

Hmmm. OK. Fairly obscure reference to a joke about a car bomb in a fairly obscure bookthat came out almost a quarter of a century ago.

…guess that’s not the best way to open up a blog post.

Anyway.

It was a simple lunch of tacos, so I was making tortillas from scratch (yeah, I know… see my other blog), and one handle on my rolling pin broke.

So, of course, down to the shop.

My favorite wood is cocobolo, and it may not surprise you to find out that I’ve got ten or twelve drawers dedicated to wood scraps of various exotic species, so I went to the cocobolo drawer and got some.

A little bit of work later, and I had two new rolling pin handles.

One-off art bowls are nice, because no matter what screwed-up shape you end up with you can claim “that’s not lack of craftsmanship – that’s just my artistic vision!” ( appending “….man!” at the end of the assertion is strictly optional ).

Even the simplest treenware, if it’s done in sets, can be a better test of skill because when you turn out multiple of something, they all have to match, and match pretty darned well.

I’m happy to say that my skills are at the point where two identical handles was pretty much child’s play.

This tells me one thing: I need to make a matched set of salad bowls…because my skills are certainly notup to making four (or eight! (eek!)) identical bowls.

yet.

Onward!

My youngest brother gave me a long-term loan of his c. 1988 hotter-than-hot guitar effects multi pedal (a GSP 21 Legend).

It’s pretty cool, in a retro I-could-run-this-on-an-iPod sorta way.

One problem: the sound is all scritchy, even in bypass mode.

Dollars to donuts it’s got bad capacitors.

So today I took it apart and unsoldered all the power supply caps.

Because I’m a lunatic anal retentive I use Swiss watchmaker trays whenever I take something apart to corall all the itsy bits

And, because I’m a lunatic anal retentive, I write down the specifications on each part as I desolder it.

• c-106: 100uF 25V
• c-107: 100uF 50v
• c-127: 10uF 16V
• c-128: 10uF 16V
• c-129: 10uF 16V
• c-130: 470 uF 25v
• c-131: 470 uF 25v
• c-132: 4700 uF 16v

Now I’m off to octopoart to buy some capacitors!

It’s been two weeks since I last worked on the CNC project – last weekend was taken up with Easter stuff.

Today I got down to the shop for just three hours. I took apart the left gantry foot (the one nearer the camera here) and propped the gantry up on 4×4 blocking. I then used a big honking drift pin (originally purchased as an aid in drawboring mortise-and-tenon joinery) to drive the screws out. Each of the 10 bolts got filed on one side with a bastard file, then the whole thing was reassembled – ahh! – now there’s some clearance instead of binding!

Once the gantry was reassembled – but with the bolts loose – I was able to put the foot together underneat and get tight registration on the cold rold steel way.

The mechanicals are 75+% constructed at this point. I don’t want to do anything with the motor mounts until the NEMA 34 motors arrive later this week.

Here are the mechanical parts that aren’t yet bolted in:

Not many left!

(My G_d there were a lot of bolts, nuts, washers, flanges, pads, lock washers, and plates when I started this project!)

The machine build is going quickly, despite the facts that

1. there are no instructions
2. because there are no instructions I often have to build an assembly, tear it apart, build it a second time, tear it apart, and then only – on the third try – get all of the subtle details right
3. the extruded aluminum is out of specification and I have to hand-file every one of 80 or so bolts that tie into the gantry

An overview of the 5′ x 4′ machine bed, the two X-axis ways, and the gantry with Y-axis sled installed:

The gantry. See the 9 bolts holding the rack on the top, the 20 bolts holding the Y-axis ways on, the 10 bolts holding the Y-axis-way plates? I had to hand file every one of them. Oh, and those five Y-axis-way plates? I had to hand file those suckers too.

Also, the gantry rides on those two big red feet. Those feet each have two sleds. Each sled had six ball bearings that are adjusted with set screws, lock washers, and more.

Also, the Y-axis sled likewise has two sleds, each with six ball bearings. And on and on and on.

Here’s the back view of the gantry. Note the eight bolts in the lower left corner (and another eight off screen to the right). All of those bolts had to be hand filed.

Oh, and note the rack up top? And the four rack mounting plates? Yeah, you’ve got it: installed them in the wrong place, hammered to fit, then had to pry out, file off galling, file to fit the under-specification T-track, etc. etc.

An end view of the gantry:

Not pictures: the Z-axis slide (90% assembled), the NEMA-34 motor mounting plates (30% assembled), and the axis end bumpers.

Woof.

To do when I next get down there: the gantry feet are spaced about 1″ too wide. I need to cut a 4×4 to length to prop up one end of the gantry, unbolt the gantry from the red gantry foot, then use a mallet and wooden dowel to pop the 10 bolts out of their overly tight T-slots, file them, then reinstall them.

Woof.

Lots of work, but it feels great to know a machine inside and out…and when 1,000+ part machine doesn’t come with instructions, you know it inside, out, * AND * Biblically.

(or, rather, it knows you Biblically…)

I’m launching a new company.

Details are a bit hush-hush right now, but it involves CNC.

CNC, for those not yet in the know, stands for Computer Numeric Control – it’s machine tools that are run by computer.

The idea is that hyper complex shapes can be cut out of metal, wood, etc. using computers to control all the fussy little details.

With out sharing too much, my plans involve a LOT of software on top of the standard stack – I want to have computers designing things, not just making them.

…but baby steps. For now, I need to either buy or build a CNC machine.

Building from a kit is cheap. I can get a mechanical kit for $2,800 and an electrical kit for another $2,500. Ubuntu is free, most of the CNC software I need is free, or cheap, and I’ve got a half dozen computers sittinga round.

So, for under $6,000 I can have a complete CNC setup for machining wood that has these specifications:

• X travel 49.75 inches
• Y (gantry) travel 50.25 inches
• Z (vertical) travel 6.5 inches (minus spoil board thickness)
• Rapid movement up to 1000 inch per minute
• cutting feedrates 400 inch per minute

All I’ve got to do is…six boxes containing 300 lbs of metal, and build a machine tool from blueprints.

Yeow!

Opening the first box

The base completed

X travel bumpers installed, rails installed, rack and pinion and gantry parts stacked on top

ONWARD!

My Nissan Xterra is now 12 years old and getting a bit balky.

Today the driver’s window rolled down but wouldn’t roll up.

Sure, I can buy a new window switch cluster, but in the mean time, it’s the winter and I need that window rolled up.

What to do?

Take the armrest apart, break out the multimeter, find power and ground, figure out that the switches must be DPDT reversing switches…

…start jumpering connections power from likely-looking lines, then probe for the otherline that’s now 12 volts away from ground. Find that and jumper it to ground…

and voila.

“Back in my day we programmed our computers using front-panel switches.”

“You had switches ? Bah. Back my my day we used wires.”

It’s been a while since there’s been progress in the workshop. I had a major server meltdown at work which resulted in building a machine from scratch, and that triggered lots of long-overdue upgrades in technologies, libraries, and techniques…and since that was already being done, what better time to move the front-end webserver for the two businesses into the cloud over at Liquid Web’s Storm On Demand?

Then I decided to rearrange my workshop, clean out the kitchen pantry, reorganize my personal finances, and start making room for the new CNC machine that I’m building from a kit.

And, of course, I’ve been cooking up a storm.

Exciting times.

Yesterday, though, I made some time to work on the Limbert table.

The first leg that I pattern routed was a bit of a mess, because (a) I’d never pattern routed before, and (b) I accidentally skipped the step in the process where you rough cut your workpiece of size on the bandsaw.

It turns out that if you use the router to remove ALL of the stock, instead of just the final 1/4″, the bit is cutting on both sides, and can bog down and suddenly lurch out of your hands. So, the leg got scarred in several places where the router jerked, tilted, and bit too deeply.

That was sort of depressing, and probably contributed to the long break in between working on this project.

In the interval, though, some new toys arrived from Lee Valley, including (but not limited to) a set of spokeshaves.

So, I took a half hour or so and used spokeshave, hand-plane, rasp, and file and cleaned up the mess I’d made in my eagerness.

And the result was…actually, it was pretty decent.

That emboldened me to work on the other three legs.

I took the template off of the backer board and traced it onto the other three leg blanks.

But, recall, I want to cut the blanks a bit oversized. So how do you make a tracing a bit bigger than the thing you’re tracing?

In The Gimp or – I imagine – Photoshop, you grab a selection and then increase the selection area by a pixel or two.

When tracing you use a much more analog technique: you grab a small washer, put your pen tip through the center hole, and then trace around the object, letting the washer ride on the perimeter.

So, I did that, then bandsawed away the waste (carefully reserving the larger chunks of quartersawn white pine – I’m sure I can use at least half of the scrap in some small project like a lamp or a clock), and then pattern routed.

It worked wonderfully.

The legs are held down to the pattern via screws driven through extra tabs (the woodworking equivalent of a small scale modeler’s plastic injection molded “sprue”). After routing I cut those tabs off on the bandsaw, and then started cleaning up the edges where I’d cut them off. The nearest leg has been cleaned up, the others have not yet been finished.

Next: fab the tabletop, clean up the rounded corners of the mortises to be square, and dry assemble.

While I was swapping the dado stack in I dropped the arbor nut.

Oh noz!

I had to crank the mechanism all the way up to allow my arm to get past, then root around blindly in the vast pile of sawdust inside the cabinet.

The good news is that I found the nut in short order.

The bad news is…why is there a vast pile of sawdust inside the cabinet?

Is the dust collection that bad?

Come to think of it, didn’t dust collection get a lot worse, all of a sudden, about a year ago?

And, come to think of it further, isn’t the blade tilt mechanism really stiff these days, as if the cabinet is full of eight cubic feet of sawdust?

I rooted around in the pile of inside-the-saw dust, and found…a weird flexible hose, about 2.5″ in diameter.

…connected to…

nothing.

Yep, it turns out that the blade rotates in a shroud under the table, and this shroud catches 99% of all the dust generated, and then funnels that out a hole in the bottom to a flexible hose, which exits out the back of the saw and plugs into one’s workshop dust collection system.

…except, in this case, it seems that maybe a year ago that hose fell off the shroud.

…and all of the sawdust since then has been filling up the cabinet.

…until it reached the level of the blade tilt gears and pack into the teeth.

Argh!

On the bright side, this is pretty easy to clean up.

I turned on the workshop dust collection system, then reached into the tablesaw from above, reached deep into the sawdust pile, grabbed the hose, pulled it away from the wall, and let it eat away the mountain of sawdust from below.

For five minutes I bent deeply over the tablesaw, one arm entirely swallowed by the beast, vaccuming it out from with in.

Eventually the job was done and I re-attached the hose to the shroud.

Yay!

…then I went out to empty the dust collector bag, which I’d managed to fill up to the 115% mark.

First I selected three quartersawn white oak boards to build the table from. One of the boards had a bad bit of bow in the last foot, so I cut that bit off. Most boards were 10″ wide, but one was 10.5″, and I picked that one for the shelf.

I ran the three boards through the thickness planer a few times, sneaking up on the final dimension and verifying it with calipers. Then I used the jointer to create on reference edge on each board. I tidied up this reference edge by hand.

I mounted my templates to backing boards, then used double-sided tape to stick the oak blank to the template

before using a router with a ball-bearing guided bit to trim the oak to size.

The table legs needed a rabbet in one edge, so I loaded up the dado stock, created a sacrificial fence from MDF, and tuned this with calipers before creating the rabbets.

…then I screwed the oak down to the leg template and routed it.

I skipped a step: I should have traced the leg templates on to the oak, then roughed the oak to shape on the bandsaw. My failure to do this, and my choice of plunging the router into deep cuts meant that I got really crappy surface finish on some places of the first leg.

Also, the ball bearing guide on the cheap Grizzly router bit unscrewed itself at one point and the router bit plunged 1/16″ or so into the pattern.

I think that if I pick up some spoke shaves and work at it, I can still salvage this one leg (with about two and a half boardfeet in it, this is $15 worth of lumber).

Even if the day ended on a slightly frustrating note, I can look on the bright side: I’ve emarked on my first ever piece of real furniture (putting aside white pine bookshelves, workbenches, and shop shelving), and so far I haven’t really !@#%$-ed anything up. Everything is recoverable, and I’m learning lots of little subtleties about routing templates.

It’s going decently.

Onward!

I needed to cut some dados (actually, some rabbets) on the tablesaw today, and this necessitated swapping blades out.

The Forrest dado blade is hard as heck to get on, and even harder to get off.

Now, if this was some no-name brand, I’d blame the blade manufacturer, but with Forrest, I’m sure that they’ve got the specifications down to the ten thousands of an inch…and they probably don’t just drill the holes, but actually bore them with some micro-crystalline carbine boring bar that’s been blessed by the High Elvish Council under a new moon.

…so, if the arbor is just a wee wee bit oversized, what am I going to go?

Expose it.

Inspect it.

Hole a well lubricated Arkansas stone against the running spindle to knock it down a fraction of a ‘thou’.

…and after that, the blades slipped on and off with just a bit of finger pressure, but still ran perfectly true.

#FTW !