Underbench cabinet

[derekcohen]

I decided to add a cabinet to my work bench. There are just too many tools on the wall, and many would be better off stored in drawers where I can reach for them when needed.

Please feel free to post your underbench cabinets here. I do not recall a thread on this topic.

The cabinet will span as wide and high as it can go without being impeded by either hold downs or the sliding deadman.

The cabinet is deep - too deep for drawers. The plan is that the drawers will not be full length deep internally, but have full length sides through to the rear to create a full extension when siding out. Internally, it has been my plan to use siding trays within the drawers ... fewer drawers externally, but more drawer space internally.

This tool cabinet is inspired by the North Bennet Street School version (a well-known woodworking school in Boston). The tool cabinet is one of their training pieces. One became an article by Tommy MacDonald in PW magazine.

Tommy's tool cabinet ..

Mine will be a little larger, more drawers, and a more complex construction involving mitred through dovetails ...

Dimensions: 660mm x 400mm x 400mm (26" x 15 3/4" x 15 3/4").

Small drawers: 205mm x 70mm (8" x 2 3/4")
Large drawers: 305mm x 95" (12" x 3 3/4")

Merbau is definitely going to add some weight to the bench! The case being dovetailed.

Well, this is about the fourth mitred through dovetail case I've built in about 18 months. I must be getting the hang of it now, since this was straight off the saw. No tweaking needed. Merbau is hard and has no give at all ...

The Merbau comes as a panel from Bunnings, a local hardware store. This is shop furniture and I make no excuses for taking a shortcut. The 18mm thick panels are flat and ready to go. Literally all I have done is cut them to size.

The penalty is that the wood is bloody hard!

The drawer fronts will be Jarrah. I intend staining the Merbau case to match the drawer fronts.

Just to prove to myself that it was no fluke, it happened again ...

I continued on until all four sides were dovetailed. Through dovetails with mitres at each corner. And every one went together off the saw ... well, almost - one mitre required a smidgeon of a mm pared away to close tightly. All tight and square. I am quite chuffed.

The difficult part is to get the mitres to close along with the sockets ...

Inside the bench ...

Regards from Perth

Derek

 

[rmhinden]

Derek,

Very nice, as always.

First time I had seen the word "chuffed", had to look it up.  I like it!

Bob

 

[derekcohen]

Time to use the slider. THIS is what the parallel guide on the slider can do. It is like a Fritz & Franz jig on steroids ...

With the case done, the next step is to prepare the boards for the drawer blades/frames. I have found a chunk of Jarrah, about 50mm thick and 180mm wide and about 950mm long. This needs to be sliced up into 50mm wide boards (which will be further reduced to 12mm thick drawer blades.

Place the board against the parallel guide ...

... and rip one side to 50mm ...

Now rip the second length ...

.. and the third. How safe is a slider? This is where one stands - well away from any possible kickback (which does not occur on a slider, anyway. And the hands are no where near the blade ...

How good are the saw cuts? Good enough to joint with, and not require a jointer for the edges.

Here is the board ...

Close up ...

But ...

Regards from Perth

Derek

 

[derekcohen]

The Rebate

A rebate can be made with a handheld router, router table, table saw, a handsaw and chisel, and a hand plane such as a moving fillister. My preference is the latter.

What can be more simple than a fillister plane along an edge? Well, the plane needs to be set up, especially when planing interlocked grain, as we have here. And before this can take place, the case needs to be prepared if the desired result is an accurate - flush and square - rebate.

The first step is to level and square the front and back edges of the case. My plane of choice here is a small bevel up plane with a high cutting angle. It is low like a block plane for easy handling, which is helpful when the case is high on the bench ...

The case is 18mm (3/4") thick. The rebate will be 7mm deep x 12mm wide. This will allow for a 6mm thick rear panel.

The cabinet will have four rows of drawers, with the lowermost row running on the bottom of the case. It is important that this surface is perfectly flat in order that the drawers run smoothly.  The boards making up the sides were flat out of the packaging. Certainly flat enough for a case, but not quite flat enough for drawers to run on with the level of precision desired here. They need further work ...

The case is pulled apart, and the lower panel is traversed. Note that the surface is first covered in pencil scribble to monitor where the high- and low points are ...

A straight edge and a longer plane are used here ...

The blade here is slightly cambered to avoid leaving track lines. A very light surfacing is completed with a smoother, more to remove any fuzz than to level ...

The moving fillister of choice is the Veritas Skew Rabbet Plane ...

Those familiar with this plane will note that the front knob has been removed. My preference is to rest my thumb on that spot and apply downforce, while the palm applies force against the side of the plane. Here is an example from another build ...

This fillister has a deeper subfence. The depth stop knob has also been slotted for ensure that it has been tightened securely ...

The plane is generally only set up to slice with the knicker ahead of the blade when planing across the grain. However, the Merbau used here has especially interlocked grain, and the nicker it employed to prevent spelching on the shoulders.

Here, the nicker is a smidgeon outside the body of the plane. The skewed blade lies in-line with the nicker. This has another purpose, which is to cut into the lower corner of the rebate and keep it clean and square. Otherwise it would allow waste to build there, and the inside would create a slope.

In addition to the line created by the nicker, a cutting gauge is run along the rebate boundary. This may be used after or during the rebate is cut to clean out the inside corner.

One last item of preparation is, following marking out the rebate (again with a cutting gauge), the lower boundary line is highlighted with blue painter's tape. This is simply to aid in monitoring the plane as it gets close to the line.

This is what the shavings from fairly straight-grained wood looks like ...

This is the result when the grain is significantly interlocked ...

The case is dovetailed with mitres at each corner. There are two benefits for this: the first is aesthetic; the second is that it permits the panels to be rebated through the full length (otherwise stopped rebates are needed) ...

Here is a better glimpse of the grain direction ...

The case back is done ...

The finish we were looking for ...

Regards from Perth

Derek

 

[derekcohen]

The case is a dry fit. Above the case is the outlay of the drawer blades for the drawers. Inside the case is the template (story stick) made to mark out the dados. The dados will be 12mm wide and 6mm deep ...

The template is alternated for each side, which ensures that they are marked at exactly the same position ..

The knife lines are deepened and undercut with a chisel to create a wall for a saw ..

A straight edge is clamped along the knife line, and a kerf is created with an azebiki saw ...

The waste can then be removed with a router plane ...

Why do it this way, and not use a power router or tablesaw? I believe that I can be more precise with hand tools. This includes the positioning and fine tuning of the dados.

The router plane's depth stop is set to 5mm, and this is reached incrementally ...

The final 1mm cut (to a depth of 6mm) is made by a smaller router plane ..

The reason for this is that, as with a smoother, which follows the undulations of a panel and removes the least about of material, so this small router plane will create an even depth.

The depth is checked ...

Any waste in the corners is removed with a side rebate plane (this is one from Veritas). be careful not to remove waste from the upper edge as this will change the position of the dado. The side rebate plane is the only plane which can plane along the inside edge of a dado or groove. It is used to increase the width of the dado - but if doing this, only remove waste from the upper edge side of the dado.

Test the fit as you go ...

Once done ...

... the surfaces are sanded to 240 grit. This is an original (!) Festool sander, when it was still "Festo"! I have had this about 25 years. Heavy, but works well for this task.

A final test for the accuracy is to align the sides ...

... and then run a drawer blade across both dados ...

Time to glue up

Regards from Perth

Derek

 

[nclemmons]

Great progress photos, Derek.  Thank you. 

 

[derekcohen]

I don't know about you, but I face glueing up with mixed feelings. On the positive side, it is great to have reached a milestone. But then the fears creep in .. will it come together like the dry fit ... what if I get something upside down ... yeah, you know I have done this!

I get everything ready ... glue (Titemark Liquid Hide Glue) and spatula ... clamps ... mallet ... wet rag ...

The bench is wiped down and covered in old newspapers ...

Both pin boards receive a generous amount of glue at the same time (all surfaces) ...

No glue is added to the tail board, with the exception of the mitres.

The two pin boards are inserted into the mutual tail board, and then the exposed pins receive their glue ...

Lastly, the remaining tail board is attached, and all corners are hammered down ...

Any glue spills and runs are immediately removed with a wet rag. I have not had a problem with finish doing it this way. I am more concerned that dry glue will act as a barrier to stain or finish, and that removing it will damage the surface.

The case is now clamped. Happily, all is square and no adjustments are needed.

Once dry, the case is checked for square once again. It is necessary to hold one's breath at this point.

All is square ...

... and in all directions ...

Continue breathing.

Time to flatten the outside of the case. The choice of smoother is an HNT Gordon with a 60 degree cutting angle. This low plane will make it easier to plane with the case high on the bench, and it can be pulled for extra leverage ..

Perhaps sacrilegiously for some, the case is now sanded (80/120/240 grits) as the plan is to stain the wood to match the Jarrah drawer fronts.

And then we are done ...

... and ready for the drawer blades.

Regards from Perth

Derek

 

[AlexThePalex]

Grmbl, I'm trying to read here, but you keep putting dead trees on top.  [tongue] [smile]

 

[derekcohen]

We left off with the dados for the drawer blades made and the case glued up ...

The plan was to make the drawer blades, partially fit them, add the drawer dividers, and complete the fitting. Then Christmas came along ...

Measuring the drawer blades had been done. First, a pinch stick obtains the width from inside the dado, and then a template is made with scrap ...

Set the template on the slider ...

... and cut to size the front- and rear rails ...

That was just before Christmas ...

Returning today, I thought that it wouldn't hurt to stain the Merbau case to match the Jarrah rails and drawer fronts. By the way, Merbau is also known as Kwila.

As mentioned at the start, the reason for choosing Merbau for this cabinet was simply that it was cheap and already available as a panel. This came with a cost, in that it is not the nicest wood to work with - interlocked and coarse grained. Plus, of course, it is light in colour. The result needs to blend with the Jarrah bench.

The first step was to use a grain filler on the outside and inside of the case. The surfaces had already been sanded to 240 grit.

The first two coats of the stain were initially mixed with a little methylated spirit, and the concentration increased for two further coats ...

With a few loose rails ...

Regards from Perth

Derek

 

[sancho57]

Very nice work. Its gonna be beautiful when yer done

 

[derekcohen]

We left off with the case complete, dados ready for the drawer blades, and the parts semi-prepared ...

With 10 drawers in 4 rows, there are a goodly number of joints to make for the drawer blades. Typically, these are made with mortice-and-tenon joinery, which has been my method to date. Today I decided to do something different .. use a Festool Domino. I purchased this four years ago to aid in building a multitude of frame-and-panel doors for our kitchen. It did a great job, and then it was retired to a shelf.

The rails for the drawer blades are all 45mm wide x 12mm thick. I used a 6mm x 40mm domino for each join. This is not the dimensions I would have used with M&T, where one tends to follow the 1/3 Rule. A centred 6mm domino leaves 3mm on each side ....

I was concerned whether this would create a stiff and rigid join, and made a couple of test pieces. No problem at all.

Not having any dedicated Domino hold downs, my bench did a sterling job ...

Flush the joins (not that there was much to flush) ...

Done x 3  ...

Time to fit these. Sliding them in to the ends of the dados ...

These is a large gap to the front of the case ...

My design calls for a 6mm set back for the drawer fronts. To ensure that this is even around the circumference, this is marked off with the help of blue tape (I really need to take out shares in this product) ...

The last step is to rebate the drawer fronts to move them forward in the dados.

The base is scored with a knife (note that the frames are a tight fit in the dados) ...

A cutting gauge is set to the line ...

Blue tape helps outline the rebate for old eyes. Note that the short side is sawn first. This is to prevent the long sawcut slitting off as the offcut is end grain and weak.

The cautious will sawn away from the line, and finish by paring with a chisel. I really do not fancy much paring in this really hard Jarrah, and decided to just saw to the line. The saw gods were smiling on me today ...



Now the drawer blades can be moved forward into their final position ..





The next step - for next time - is to begin the sliding dovetail drawer dividers.

Regards from Perth

Derek

 

[Svar]

This is an interesting development where we discover that Derek not only uses power sander  [eek], but also Domino [scared]!
What do your planes, scrapers, and chisels say about it?  [poke]

 

[HarveyWildes]

This is an interesting development where we discover that Derek not only uses power sander  [eek] , but also Domino [scared] !
What do your planes, scrapers, and chisels say about it?  [poke]

Hah! They laugh because they know that this is just shop furniture .

Seriously, I just found this thread, and it's very nice.  Once again I'm inspired by your work!

 

[derekcohen]

This is an interesting development where we discover that Derek not only uses power sander  [eek], but also Domino [scared]!
What do your planes, scrapers, and chisels say about it?  [poke]

The chisels are not speaking to me at this time!  [big grin] I received a new chisel in the mail yesterday, a 6mm Kiyohisa. These are no longer made, and I seek them out whenever I can find one at a reasonable price (rarely available under $300 each). I used the 6mm with the sliding dovetails, which have begun, and now the Domino is sulking. You cannot win!  [unsure]

Some are surprised when I pull out power tools, but I have been a power tool, and machinery, user for decades. I have some pretty nice power tools ... including a Mirka Ceros ROS which is my usual go-to. It is just that hand tools are more satisfying to use, and in certain cases they are more precise in their results.

Regards from Perth

Derek

 

[derekcohen]

After making dados, and drawer blades, and then fitting them, comes the vertical drawer dividers. I consider that these are the make-and-break of a chest of drawers. It is tough to get a case square. It is tough to get the the dados positioned correctly. The drawer blades are simple. But then comes the dividers ...

Drawer dividers are needed when there are more than a single drawer in each row. In this cabinet, there are two rows of three drawers and two rows of two drawers. Care in aligning the dividers at the front and rear is necessary to ensure the drawer case is parallel and square if the drawers are to run smoothly.

It is easy enough to attach the dividers with dados. However, this is a less-strong design than attaching them with sliding dovetails. Building sliding dovetails is more complex, but the advantage here is that they tie the drawer blades to the case, making the case more rigid. That is a good thing, especially for a tool cabinet.

This article is about the preparations for the dividers, and the next article will be making them - there are too many photos for a single article.

Issue: drawer blades flex and bow. Accurate marking out needs a stable foundation.

The first step is to make spacers to keep the height even. Start with the ends ...

.. and then move them into the centre ...

The second task is to make templates for spacing out the drawer dividers, similar to the heights for the dados ...

Later, I made these thicker so that they could also aid in ensuring that the dividers were plumb when marking out.

Lastly, for now, the drawer dividers were cut ...

The dovetails will be 3mm deep at each end ...

Onward to the dovetailing ....

Regards from Perth

Derek

 

[derekcohen]

Part 2

This is the front elevation of the cabinet ...

I decided to work on the rear of the chest first .. so I could get in some practice and make my mistakes where they will not be seen!

Much preparation has gone into accurately positioning and marking the sliding dovetail parts for each divider. This was described in the previous post ...

Sliding Dovetail Tails

It is a good place to start. All the dividers are cut to size. These are each 12mm wide x 45mm deep. The heights vary per row, but there are essentially two drawer heights.

The tails will be 3mm deep as the drawer blades are 12mm thick, and two sockets take up 6mm of this.

This is the work area and tools ...

There is a bench hook for a dozuki, a Stanley #79 side rebate plane converted into a plane for dovetailing, a cutting gauge, a knife and a wide chisel.

The aim is to make dividers like these ...

Begin by marking the shoulders of the dovetails ...

Ensure these lines are deep. Score them with a knife.

Pencil in the cheeks ...

The pencil is going to act as a "depth gauge" when dovetailing.

The Stanley #79 was originally designed to take a slither off the inside of a groove or a dado. I have modified this one by adding an angle (6:1 ratio) to the depth stop ...

By running the #79 along the edge, the angled blade will now slice away the cheek at a 6:1 ratio ...

Look carefully at the beginning of the cheek for a sliver of pencil. This shows what has not been cut to the edge ...

Just work that area. One slice is enough. Done ....

The drawer blade is 45mm wide, of which 20mm contains a tenon from the mortice-and-tenon joint (created with a Domino). The dovetail socket must remain clear of this, and therefore the maximum length of the Tail is 25mm.

The 20mm waste is removed with the dozuki ...

Result ..

Sockets

Positioning and marking out the sockets is a aided by a template and blue tape (it would otherwise be impossible to see anything this small in dark wood).

Time for just one example. I have chosen the more difficult dovetailing into the case. From the rear it is possible to excavate into the case, itself. (This is not possible at the front, and blind sockets will be made there).

The divider is positioned with all drawer blades inserted - the drawer blades will be removed once the marking out is complete to permit room to work ... there is a lot of removing and replacing throughout!

The tails are marked ...

The boundary lines are scored ...

It is important that the lines are scored from the inside of the square, otherwise they will be in danger of moving to the wrong side.

Similarly, when positioning a sawing guide (also 6:1 angle), ensure that the saw is inside the cut line ...

One added advantage of using the saw guide is that accuracy of marking the tails is not critical. The tails are cut at 6:1, and the saw guide just needs the apex of the tails to be marked accurately.

Sawing with the dozuki leaves a triangle of waste at the far end. The kerf can be extended using a razor saw ...

With the sidewall cut, zip out most of the waste with a paring chisel ...

Finish to the desired depth with either a router plane ...

... or the wheel gauge ....

Here is the rear of the case with all the dividers completed ...

The result is a very solid lattice of drawer blades and dividers. Rock solid!

Work has now begun on the front drawer dividers ...

Regards from Perth

Derek

 

[Tinker]

Derek,
whenever I see your name attached to a thread,I know I am going to get a great read and an education. You aways come thru to expectations.
Thank you
Tinker

 

[Svar]

Why only two narrow dividers to position and guide the drawer? If the drawer skews sideways when closing it'll hit the back divider.

 

[derekcohen]

Hi Svar

There will be a rail connecting the two dividers. These are installed once the dividers are done.

Regards from Perth

Derek

 

[derekcohen]

A blind sliding dovetail

This is the front of the cabinet, with all the vertical dividers to install.

The single lower- and the two dividers need to be fitted into a sliding dovetail ... however, unlike the dividers at the rear, this must be from inside the case. In other words, a blind sliding dovetail.

This post will show the steps taken for the lowermost, central divider.

This could not be done without the aid of blue tape. This is used to mark the apex of the tail ...

This is how the tail will be positioned. Below is where the pin socket has to be cut ...

A line is scribed at the two apex points, and the dovetailing guide is placed against it ...

Very carefully, using the dozuki, saw about 5mm from the boundary line. Be careful not to get closer than this. Saw marks must not show outside the dividers.

Do the other side ...

Swing the case around and work from the other side. Use the razor saw to cut up to the boundary line ...

Now chop out (shallowly) the trench/dovetail in the same way one would do a hinge mortice ...

Finish and level the surface to a depth of 3mm (the height of the tails) with a router plane ...

The result ...

What is difficult to see here is that the trench is fractionally wider in the rear half to allow the tail section to enter, and then slide along.

Unlike the internal tail sections, those used in the blind dovetails require a small beauty rebate ...

The divider slides along into position ...

Why do we go to all this trouble to use sliding dovetails? Well, this is how strong they are ... one hand attempting to lift the case ... neither are budging.

Incidentally, I was asked "how does the drawer run between two dividers?". Well, of course it does not - all will have a rail to guide the drawer ...

One the lowermost divider is in, the one above it can be marked ...

Progress to date ...

Regards from Perth

Derek