New Trays for the Pandora

As mentioned yesterday, the trays I made felt rushed, and subsequently I wasn’t happy with the results, and so a remake was in order. One of those things – less haste, more speed.

I decided that I really did want it to have dovetailed sides – the wall height was around 20mm, so it would involve a single pin and two half-tails, with a wall thickness just sufficient for the Gifkins Dovetail Jig. Once the Silky Oak was machined square and to size, the dovetails were cut and the sides dry-fitted together.

Tray sides dovetailed, ready for slots to be cut for the base

It was only a dry fit at this stage, as I wanted to have a captive base, as I had done for the main box. With a clamp holding the box together, a groove was cut all round the inside of each tray, about 5mm from the bottom and 5mm deep.

The inside dimensions of each tray were again measured, and 10mm added to give the size for the base. In reality, I tend to cut it about 0.5mm-1mm undersized, so there is no chance the base will stop the sides coming together completely during the glue-up.

How I measure this is with a rule, and in this case I regularly turn to the Woodpeckers Rules, which are particularly easy to read. By setting the reading under one of the teeth (and ensuring the desired dimension is on the correct side of the tooth), I set the fence position (or the stop on the mitre gauge, depending on the cut – rip or crosscut).

Setting up the Incra Miter Gauge

Setting the rule to measure to the side of the blade tooth

The photo doesn’t show an actual measurement, but in any case accuracy is always something both difficult to achieve, and worth pursuing. Even measuring to the edge of the tooth is not an assured result. All blades (and all tablesaws for that matter) have a degree of runout. The only real way of determining a measurement is with a test cut. You can take some steps to actually get accurate measurements, but it still involves a test cut, and measuring to the side of a specific tooth, and measuring to this tooth each time. So long as the blade does not slip on the arbor, and you do not change blades then this will then remain reasonably accurate.

In practice, this degree of accuracy is rarely needed – wood is reasonably tolerant in any case, and there are other ways of ensuring accuracy. One is gang-cutting. If I want two sides to be cut to exactly the same length, you can either use a fixed stop that each side butts up against (such as the Incra Shop Stop), or cut both sides at the same time.

Back to the bases, once they were cut to size, it was over to the router table to cut the rebate around the edge. To set it accurately so the base sits flush with the bottom of the sides, I use the same router bit as used to cut the trench. It needs to be dropped an accurate amount, and I have a reasonable way to achieve that, and it doesn’t involve a rule.

Setting accurate router bit height

A router bit is a power chisel, so I use it as such. Without turning the router on, I lightly scrape the endgrain just enough to reveal the exact height of the router bit. This leaves a mark to line the router bit up with when dropped to the lower position.

Scoring the exact chisel height

Tray base and sides, ready for glueup

Each tray got glued and clamped. One interesting aspect of dovetails, is you primary clamp the tail sides, which pulls the pin sides in. I still use a clamp to ensure the actual joint is not loaded up until the glue sets – you don’t want the wood fibres getting compressed unnecessarily. You may note that I used pine for the base – given I planned to cover the working surface with felt, I didn’t see the point wasting top quality timber in that situation. It doesn’t look bad from underneath, and will rarely get turned over in any case.

Once the trays were glued, and sanded, I tried the fit to the main box.

Testing for fit of the tray inside the box

You know you have the fit pretty right when the tray struggles to sink into the box – not because of friction between the sides but because of air pressure in the box! With a little more sanding, it slips down nicely, still with a little resistance, and a very satisfying “shhhh” as the air escapes. Love it!

I had another detour at this point. After the trial a week or so earlier of the dividers, it was time to make them for real.

Jarrah interlocking dividers

The dividers were cut with the thin-kerf CMT blade, and again the Incra Miter Express proved invaluable.

FWIW, Incra and Woodpeckers gear all comes from Professional Woodworker Supplies, and the CMT blade from Carbatec. Thought I’d mention it if you were looking at what I use.

Main tray with dry fit of Jarrah dividers

I was happy with the main tray with the dividers made, but when I fitted them into the smaller tray, it looked too hard to get the individual charms out, and too much like a iceblock tray.

The “ice bock tray”

Again, when not being prepared to accepting something not quite right, I decided there was no option but to remake the dividers for the upper tray. This time, I chose a wall height of 6mm. When working with power tools, that is small, and risks putting fingers too close to blades.

So it was time for handtools. Yeah, I know – shock, horror.

The sides were cut close to the height required, and then it was time for the handplanes.

There was no point trying to bring a handplane to the individual piece – too hard to see what is going on, let alone controlling it, so I reversed the situation, and used the plane in the same way as it’s power equivalent: inverted!

Inverted HNT Gordon Trying Plane

So I took my HNT Gordon Trying Plane, and mounted it in my Veritas twin-screw vice. The individual sides (the dividers) were then run over the top of the blade. The blade was set for a very light cut – there is no rush! If you haven’t set a traditional wood plane blade before, there are no adjustment screws, it is all done with a careful tap tap of the wooden mallet you can see in the top right of the photo.

The new, 6mm high dividers

So the new dividers in comparison with the original ones – chalk and cheese, and right.

I haven’t mentioned how I cut the slots, other than the Incra Miter Express. The short lengths were done very easily in two passes, and all gang-cut at once. With the Shop Stop set, the first slot was cut, and then the whole bunch rotated and the second cut. Took no time at all. I had made some trial cuts to ensure the blade height was just right.

The two long lengths obviously took a little longer, and the V groove track on the Incra fence was invaluable, allowing me to move the stop exactly 22mm between cuts (20mm for the gap, and 2mm for the kerf)

These V groove racks that ensure accurate positioning of the Shop Stop are invaluable.

So the whole jewellery box was coming together. Next, we will look at the lid, and then final assembly.

Hope you are enjoying the process!

Opening Pandora’s Box

I always seem to leave my run very late, but seeminly managed to kick off a project this weekend just in time. My wife tends not to read this blog, so if you happen to do so, and also know her, best to not mention this post at all!

With her 40th approaching (in a couple of weeks), I’m wanting to make a bit of a jewellery box that is specifically designed for her collection of Pandora bracelets, necklace and spare charms.

It is getting designed as I build it, so I don’t know what the final item will look like yet- each step reveals a little more detail.

To start, I’ve taken a length of Silky Oak off the woodrack. It was really twisted, so rather waste a huge amount of the timber trying to get it flat, it was easier to rip it down the middle.This worked out ok anyway, as the resulting boards were about the height of the sides I envisaged I’d need. These were run through the jointer/planer to get one flat side, then using that side as a reference, run through some more passes to get an edge at 90o to that side.

From there to the thicknesser, running both boards through to ensure a uniform thickness, then finally back to the tablesaw to set the final box side height, then docking the boards to length.

20120708-014255.jpg

Next, it was over to the router table, where once again I utilised the Gifkins dovetail jig to create the joints I wanted. As always, it didn’t take long, and the resulting dovetails were perfect.

20120708-014450.jpg

Changing over to a small slot cutting router bit (one with a bearing, limiting the depth of cut to 5mm), and temporarily assembling the sides with clamps to hold the box together, a groove was cut all around the inside of the box, about 5mm up from the bottom.

Instead of wasting another piece of Silky Oak, (especially considering I am planning on covering both sides with felt), I took a piece of crapiata (Pine), and cut a base for the box, 10mm larger in each dimension than the inside dimensions of the box.

Back to the router table, and a groove was cut all round (also 5mm deep) to create the lip that will fit the groove in the box sides. Over to the thicknesser, and the extra thickness was removed leaving a board around 10mm thick, so that when it is in the slot of the box sides, it is flush with the bottom.

20120708-015217.jpg

Later on in the build, I will cover this with some felt.

Turning the box upside down and pulling off one side, you can see how the base engages with the box side. The sides are then glued together, leaving the base unglued but captive.

20120708-015509.jpg

My final job for the day was to experiment a bit with the dividers that will be used to separate each of the bracelets, and create cells for each of the loose charms. I will create two trays that fit inside the box, each with a different arrangement of dividers. To get the dividers, I took a block of jarrah, and cut it into thin strips, about 2.5mm thick on the bandsaw.

20120708-015855.jpg

It was only a small block to start with, but that is the benefit of the bandsaw- resawing, and a very thin blade kerf wasting a minimum of timber.

20120708-020023.jpg

Using my recently-created thin stock carrier for the drum sander, the strips were passed through to get them to a required finish and uniform thickness. I wanted to cut slots using a thin-kerf tablesaw blade, so sized the strips down to match that.

20120708-025019.jpg

Ripped once more to the final desired height, that one small piece of jarrah yielded a fair collection of dividers!

20120708-025140.jpg

There wasn’t enough time to do much more, other than a quick test of the dividers.

20120708-025238.jpg

Next time, I will create a couple of trays, fit the dividers in, and sort out the lid and hinges.

At least I have made a start!

Timber of-the-Month: Silky Oak

Timber of-the-Month: Silky Oak – Grevillea Robusta (or possibly the oft-mistaken Cardwellia sublimis)

This month’s timber of-the-month from Brad’s Burls is Silky Oak.  It’s not actually an Oak, and is an ornamental tree (still growing up to about 40m), particularly in northern Australia (primarily Queensland).  Also known elsewhere in the world as Lacewood.

Raw As-Cut Silky Oak

Raw As-Cut Silky Oak

Even in an as-cut state, you can see the very distinctive patterns and textures in the timber.

It is quite physically light, and has a tendency to split.  The boards I have though seem relatively stable, and have minor checking at the ends.  However the ends have a more significant problem, and it was simply an unfortunate event that the boards had to be used to nail shut the pallets of wood that Brad’s Burls were bringing to the Melbourne Show.  Unfortunate for them, very fortunate for me!  And the timber around the nail holes will not be wasted either – there are plenty of pen blanks in the remaining portions between the holes.  Of course, the bulk of the boards are in a pristine state, and will soon become boxes as likely as anything.

Oiled Finish

Oiled Finish

Again, a light oil and sand gently accentuates the contrasting timber colours.

Closer Look at Texture

Closer Look at Texture

Closer

Closer

Closer

Closer

Here you can clearly see the interesting contrasting textures in the timber – linear structures with diagonal highlights.  However, a rather uniform pattern of medullary rays and a high lustre and finish occasionally breaks down into fascinating, detailed swirls that could make for a real detail piece in a box lid or something.

Abnormal Pattern Swirl

Abnormal Pattern Swirl

Given the average colour of the timber is pale, I think this would look great with a dark highlight colour for a (thin) frame, and perhaps inside a double dovetail. Just thinking out loud.

Brad's Burls

%d bloggers like this: