Rolltop desk - WIP

[Rocker]

My son-in-law asked me to make him a rolltop desk, which seemed like an interesting challenge; so I have laid in a large stock of 150 x 25 blackwood and 190 x 19 hoop pine for its drawer sides. I bought a set of plans from furnitureplans.com, but decided to modify them slightly, since I wanted to use raised panels rather than plywood panels, and I wanted to make the tambour from interlocking slats rather than using separate slats glued to canvas. I also wanted to install the drawers by the method described in an article in the current Fine Woodworking.

Apart from the tambour and the drawers, the joinery is very simple; the framework of the desk is made up of 10 frame-and-panel panels, which are made by groove and stub-tenon joinery. The grooves are ripped on the table saw, using the two outer blades of a dado set; the shoulders of the stub tenons are cut by using a stop on a crosscut sled (see picture below), and the tenons themselves are cut in a single pass, using a tenoning jig and the two outer blades of the dado set, separated by a spacer.

Rocker

[Carry Pine]

Rocker,
You sound very confident in what you are going to do and the choice of joinery is something special. I learned all about roll top desks when I inherited my grandfather's desk and the removalists damaged it when it shifted in the load in transit.
Best wishes with it and we look forward to seeing WIP.
Carry Pine

[Auld Bassoon]

Interesting project Rocker!

On the drawer sliding mechanism (as featured in the current FWW magazine), I tend to use this method quite a bit.

Making the tambour sounds slightly more challenging. How do you intend to make the interlocks such that the close neatly yet can flex?

Agree with your thoughts on using raised panels rather than the very U.S. approach of veneered ply.

Looking forward to seeing more of this!

[AlexS]

Rocker, I presume the pic is of a slat. could we have some more detail on how they will interlock please - it looks like an interesting concept.

Way down on my to-do list is a tambour using canoe-maker's router bits (concave & convex) and joined by fishing trace running through the slats. I'm interested in seeing how you plan to do yours.

[Rocker]

AlexS,

The picture in my previous post is of cutting the tenon shoulders on one of the rails. The tambour slats will be interlocked as in the pic below. The slots will first be cut 1/4" wide with dado blades and then will be routed using a core-box bit. However your method sounds easier; I might have to fall back on that, if the recalcitrant nature of blackwood makes my proposed method impractical.

Rocker

[Auld Bassoon]

I suspect that you'll need to have some fairly stiff feather boards pressing down on the slats to keep the grooves neatly parallel; nice idea though!

What about the curvature on the face of the slats? Or do you intend to have each one flat-faced?

Most folks just buy LV pre formed tambour panels - talk about doing it the hard way :D

[Wood Butcher]

Way down on my to-do list is a tambour using canoe-maker's router bits (concave & convex) and joined by fishing trace running through the slats. I'm interested in seeing how you plan to do yours.

I've seen that done for pipe organ consoles. There used to be a bloke (Walter Emmerson) in Toowoomba that did that using SS wire that was threaded one end. On the threaded end he put a spring, washer then lock nut after it was passed through the slats. That way it never went slack.

[Lignum]

Rocker your way looks the same in outcome to FWW N0 79 PG 52. Good luck as it looks tricky. Look foward to seeing the progress pics:D

[Rocker]

Lignum,

Yes; I have a reprint of that article in the FW book Traditional Furniture Projects. But the desk in that article seemed to me to be too massive - it is 880 mm deep - so I went for a design that is just 660 mm deep instead.

Rocker

[Jill]

Hi Rocker - I'm looking forward to seeing this. I have at least 4 more desks to build later on (after the house is done - when we'll have a WORKSHOP!!).

But Rocker - what are you doing using a PLAN? I thought you were ze 'Plan Man'!! Great to see you're not sticking to it completely, though, and your changes mean that you'll be using plenty of finely tuned skills!

Happy woodworking,

Jill

[AlexS]

Looks tricky, but will be good when you pull it off. Unlike my idea, it means the slats will be attached to each other for their full length.

[Rocker]

Here is a picture of the stub-tenon being cut on the end of the curved rail that will have the groove for the tambour routed in it later. To ensure that the tenon is centred, I subtract the tenon thickness from the thickness of the stock, and then halve the result to give the thickness of waste that must be cut away from either side of the tenon. I then adjust the tenoning jig so that the distance from the inside of the dado-set blade is that distance from the face of the tenoning jig. The tenon shoulders have already been cut on the crosscut sled, so the tenon can now be cut by the dado blades in a single pass.

Rocker

[Ticky]

Gutsy Project Rocker, but then, so are rocking chairs. I have been considering a Roll top to replace my mess, or is it a desk?

I will be watching this WIP with interest.

I am also very keen to see some more detail, materials & critical measurments on both your Crosscut sled & your Tennoning Jig if you have them available. My Table saw was about the third tool I bought when I decided take up the sport of woodwork (not much art comes out of my shed), & my crosscut sled was about the first thing I made. It's ok, but it could be better.

Cheers

Steve

[Rocker]

I should have mentioned that, coming from the land of the free, the plans I had were in mini-cubits. They specified a width of 3 1/2 mini-cubits (76 mm)for most of the rails and stiles of the panels that make up the carcass of the desk. However, I could only obtain blackwood in 150 mm width, which meant that, to get two rails/stiles from a board, their maximum width, after dressing the timber, was 70 mm. This meant that virtually all the dimensions of the parts of the desk needed to be recalculated. This would have been an error-prone nightmare in MidgeCad Pro, but was easy enough to do in TurboCAD. Google Sketchup could probably have done the job also, but less conveniently, since it lacks a built-in grid.

Steve,
I need your e-mail address to send you the details of the crosscut sled and tenoning jig.

Rocker

[Rocker]

Well, the relatively easy part is done. Here are the dry-assembled panels that will make up the carcass of the desk. Now I have to make the writing surface and top, the pigeonhole unit, the drawers, and the tambour slats.

Rocker

[Lignum]

Rocker skip the boring parts for now and go straight to the tambors. Im hanging to see the WIP on it:D

[Rocker]

Lignum,
Just for you, I have run a test on the female part of the tambour slat joint. I first ripped a 1/2" deep, 1/4" wide groove with the dado-set blades, followed by routing with a 1/2" core-box bit. I was pleasantly surprised that the latter made a nice clean cut with no splintering or burning. I was glad however, that I had an Xtreme collet extension, since the shank of the core-box bit would have been too short to use in the router table without it. The inner shoulders of the groove will be rounded over slightly with a full-bead bit.

The male part of the joint should be fairly easy to make with 3/8" diameter bullnose and roundnose bits.

Rocker

[Groggy]

Rocker, this is interesting the way you are approaching the tambour slats. The one I made years ago just had canvas glued to the back and the edges were rebated slightly to run in the routed guide.

Do you think you may get weather effects having multiple male/female joins like that?

[Auld Bassoon]

Interesting approach Rocker.

How wide do you plane each of the slats to be? Even with slats around 12 ~ 15mm, that'll still be a lot of routing... Also, as Groggy mentions, will timber movement be an issue?

[Rocker]

Steve,

I am following the recommendations of the FWW article referred to by Lignum; each slat will be 19 mm thick and about 38 mm wide overall; but the parallel-sided part of each slat will only be about 21 mm wide. The ends of the slats will be rebated to leave tenons about 10 mm thick to slide in the grooves in the top side panels.

I don't think humidity changes could have any significant effect on the interlocking joints.

Rocker

[Auld Bassoon]

I look forward to seeing some more of this...

[Lignum]

I look forward to seeing some more of this...

Same here:D Using a canvas backing is like comparing machine cut dovetails to hand cut. Rockers way is so much more classy.

[Lignum]

Whilst i dont want to get away from what Rocker is doing, this is my all time favorite Tambor piece. I just love it to death. So exsquisite.

[Rocker]

I have now worked out a procedure for making the interlocking tambour slat joints. I start off with a board 70 mm wide and 20 mm thick. Then,

1. Rout four 3/8" wide 5/16" deep grooves in the faces of the board, centred 3/8" from the edges using a 3/8" roundnose bit.

2. Rout the edges of the boards with a 3/8" bullnose bit to complete the ball parts of the joints. This leaves thin 'wings' on the edges of the ball that are easily pared off with a chisel.

3. Rip the board in half.

4. With dado blades, cut 1/4" wide, 1/2" deep grooves in the newly sawn edges.

5. With a 1/2" diameter core-box bit, rout a semi-circular profile in the grooves.

6. With a full-bead bit, round over the shoulders of the grooves in two passes, rounding over one shoulder with each pass, to complete the female parts of the joints.

The pics show the router bits, the Xtreme Xtension, and the 1/2" to 1/4" reducing adapter used in the routing .

Rocker

[Auld Bassoon]

Hi Rocker,

Very nicely done; a good solution.

I'll bet that you'll be dreaming of tambours for a week :D

I wonder how darksiders did this sort of work years ago - perhaps each slat sliced in half longitudinally, grooved with a router plane, then glued together again?

Waiting with bated breath to see the outcome...

[Rocker]

Today I template-routed the grooves in the side-panels to hold the ends of the tambour slats. Owing to an accumulation of minor errors, I discovered that I had routed the grooves slightly too close to the back panel to give the requisite clearance between the tambour slats and the frame of the back panel. However, I was able to recover from this potentially disastrous error by using an auxiliary high fence on the table saw to saw away about 3.5 mm from the frame members of the back panel, and so provide the necessary clearance. It remains to be seen whether the tambour slats are straight enough not to rub against the back panel - I will keep my fingers crossed.

Rocker

[Auld Bassoon]

Cheering you on David!

Don't you just hate those accumulated errors! - still, you seem to have worked out a solution; I hope it works out...

[Rocker]

Here is a pic of the side panel with the routed groove for the tambour. A tongue on the end of the back panel will fit into the groove at the back of the side panel.

Rocker

[damienhazo]

This is very interesting for me 'cause I've always been interested in this type of 'door'...

Have you tested to joining mechanism under load? What I mean is, how it bends when pulling the weight of the entire roll-top.

I built this bed for the dog and found the wood snaked back and forth when accordianed (is that a word?) together, but lost a lot of flexibility when pulled apart:

24201

Great post. I'll be following with interest...

Damien

[Rocker]

Lucky dog:)

I am still some way off installing the tambour. Maybe another week yet before I can answer your question. But my guess is that it will be flexible enouugh to get round the curve without any problems.

Rocker

[damienhazo]

But my guess is that it will be flexible enouugh to get round the curve without any problems.
Rocker

Hey Rocker,

If you make one long piece, you can run it over your saw at 1 inch lengths. Then you get a bunch of small pieces to slot together for testing purposes. You'll soon see how it reacts when pulled apart (closing the top) and how good it handles the curves.

Damien

[Rocker]

My worries that there would be insufficient clearance for the tambour were misplaced. I had misinterpreted the ambiguous instructions in the FWW article as to which side of the tambour slats to cut the rebate to enable the slat to fit into the slots in the side panels. The article said to cut the rebate on the 'back' of the slats. I originally interpreted this to mean on the top side of the slats. However, careful examination of one of their diagrams showed that they actually meant to cut the rebate on the bottom or inside of the slats.

Damien,

I have already finished the tambour slats, apart from trimming them to final length and cutting the rebates. I think the joints are sufficiently flexible to get around the curves already; but, if I find they are not, I should be able to improve the flexibility by slightly deepening the grooves that form the neck of the ball part of the joint.

Today I will cut escape slots near the half-way point of the tambour grooves to enable the tambour to be fed into the grooves. These escape slots will be covered by the top panel of the desk.

Rocker

[Rocker]

A bit more progress: The side panels have been glued to the back panel (Photo 1), and the escape slots have been cut for installation of the tambour. You can just see them in photo 1, if you look hard enough. The tambour slats have been cut to length, using the crosscut sled, and the shoulders of the rebates on the ends of the slats have been cut. Photo 2 shows the waste being cut away from the rebates, using the tenoning jig.

Rocker

[damienhazo]

Hey Rocker, you keep this up and you're going to have to change your name to Roller. Or Rock'n'Roller if you keep making chairs... Can I just call you Elvis? Awwww go on.....


Damien

[Rocker]

Finally, the tambour is completed and temporarily installed, and, remarkably, it works, although it needs more sanding and waxing of the groove to get it to slide more smoothly. The stick under the side panels is temporarily screwed to their bottom edges as a spacer to ensure the side panels are parallel to one another.

It looks as though my estimate of the amount of blackwood needed to complete the desk was about right; I ordered 95 m of 150 x 25, about 0.35 cubic metres, and will probably have a couple of boards left over. Together with the hoop pine for the drawer sides and the requisite hardware, the total cost of materials is around $1600.

Rocker

[Auld Bassoon]

Another wonderfully executed project David!

The tambour looks terrific; waiting to see it finished!

Quite high on the materials cost (and never mind your labour time), but then that's not the point, is it?

[Rocker]

Steve,

I could have cut the cost considerably if I had used veneered MDF for the pedastal sides and plywood for the back, but I decided to build the whole desk from solid wood. It seemed to me worthwhile to spend a bit more to make a piece that should last for generations.

Rocker

[damienhazo]

Very nice.

Don't forget to sign it so future generations will know exactly when Great Great Great Granddad built it.

Some of these desks have secret compartments built into them. Given any thoughts to tricky hiding places or nifty mechanisms?

Damien


Found this after I saved: I could have cut the cost considerably if I had used veneered MDF for the pedastal sides and plywood for the back, but I decided to build the whole desk from solid wood. It seemed to me worthwhile to spend a bit more to make a piece that should last for generations.

My sentiments exactly!

[Auld Bassoon]

Quite right too David; the money will seem trivial later on, and the piece will last for a very long time.

As Damien says, don't forget the Maker's Mark... :)

[AlexS]

Coming along very necely, looking forward to more WIPs and seeing the finished piece.


Don't forget to sign it so future generations will know exactly when Great Great Great Granddad built it.

Also, suggest to the family that each succesive owner signs it. Makes it a much more historic item in a couple of hundred years, when I'm sure it will still be around.

[Rocker]

I have now milled all the pieces of the pigeonhole unit, mostly from 12 mm stock, but with dividers in 6 mm stock. Pic 1 shows the dadoes being cut with a dado set, using a wide-kerfed crosscut sled. Despite having a print-out of the plan in front of me, I still managed to stuff up with cutting a couple of dadoes, and had to prepare extra stock to replace the the spoiled pieces.

Pic 2 shows the first stage of the glue up. The pine stick on top is a spacer to ensure the dividers remain parallel to each other. The glue up has to be done in four stages


Damien,
I decided to pass on the rather complex drawer-locking mechanism described in the FWW article. It would be feasible to put a second drawer box behind one of the drawers to make a secret compartment, but I am not sure if it is worthwhile - probably too obvious.

Rocker

[Andy Mac]

Hi Rocker,
Its all looking really nice, and we're enjoying the journey! :) Its good to see you kept to solid timber, but I'm surprised at the amount you needed. I suppose all the pigeonholes and drawers will usea bit, and curves always suck up a bit more than expected.
Good luck with it.

[Rocker]

I was a bit optimistic to think the pigeonhole glue-up would only be four stages; more like six. Here are stages 2 and 3. Extreme accuracy in cutting the dadoes and cutting the stock to length is required, in order to ensure that the pigeonhole unit will fit together properly. The first pic shows the shelf clamped to a square to ensure squareness.

Rocker

[Jack E]

Absolutely awesome Rocker:)

I will PM you my details so you can send me the plans for your tamour jig:D

Cheers, Jack

[Auld Bassoon]

I don't know if you work alone or have a helper David, but one of the things that really tries me is the glue-up stage of a complex piece. Even though I go through the dry set-up, including dummy clamping runs, etc, the actual job often has me practising my Arabic, or even Mongolian :o

[Rocker]

I will PM you my details so you can send me the plans for your tamour jig:D

Cheers, Jack

Jack,

I didn't use any jig for making the tambour. All the details for making it are given in post #24 of this thread.

Rocker

[Jack E]

I didn't use any jig for making the tambour. All the details for making it are given in post #24 of this thread.
Sorry mate,

It was a poor attempt at humor:)

Once again, awesome work, I am sure we all plan to build a roll top desk "some day" but it doesn't surprise me that you are the one to do it.:)

Cheers, Jack

[Rocker]

Uh oh! Big stuff-up. I should have marked the positions of the dadoes in the top board of the pigeonhole unit from the glued up shelf/divider boxes, since I wasn't able to make the latter accurately enough to conform precisely with the ideal dimensions of the plan, so I now have to make a new top board and waste about 5 m of blackwood. First, though, I am going to try widening the dadoes to accomodate the errors, and glueing a narrow strip to the front edge of the existing top board to hide the bodgy dadoes. If the bodgy dadoes inside the pigeonholes are not too noticeable, I may stick with this solution, and not waste more wood.

Rocker

[jacko]

Rocker, it does my heart good to hear that I am not the only one that makes left/right back/front up/down stuff ups! Though I was alone in having that sort of senior moment! Seriously though, the desk is looking very impressive. My heart was in my mouth when I saw the description of the tambour method, thought it would bind or have insufficient rolling radius. I am very glad to hear that all went well and according to plan.
Looking forward to more episodes in your journey.
Jacko

[Rocker]

Jacko,

Thanks. I have managed to get the main part of the pigeonhole unit together now, and will add the final pieces in the morning. I went with salvaging the bodgy top board, adding a strip to the front edge. I don't think the bodgy dadoes will be noticeable unless you crane your neck and peer into the pigeonholes. I will post pics of the completed unit tomorrow.

I learned by the dado debacle not to follow the instructions in plans too slavishly. I should have realised that the dadoes in the top boards should not have been cut at an early stage. Instead I should have marked their positions using the glued up shelf/divider boxes.

Rocker

[Rocker]

After many trials and tribulations, the pigeonhole unit is fully assembled. It needs further sanding and finishing with wipe-on poly before installation in the desk.

I am convinced that the plans I bought gave me a bum steer on the pigeonhole unit assembly. By far the easiest way to do it would be to glue up the main boxes in stage one; then slide in the shelves; and finally slide in the pigeonhole dividers. This would eliminate the sort of problems that I had when trying to follow the recommended assembly method in the plans.

Rocker

[Auld Bassoon]

G'day David!

My sympathies with the plan stuff-up. One of those "insert tab 'A' into slot 'B'" sort of sets of instructions, eh? :)

I note that you have used through dados rather than stopped ones for a number of the pigeon hole elements. Any reason for this, or do you plan to face-frame the assembly?

I must say that I often use through dados, but that's because I'm cack-handed, which you, clearly, are not.

BTW, after glue-up, especially with relatively 'fiddly' pieces like this, I find clean up to be a real chore - how do you manage this?

Really looking foreward to see the whole piece together :)

Great stuff mate!

[Rocker]

I note that you have used through dados rather than stopped ones for a number of the pigeon hole elements. Any reason for this, or do you plan to face-frame the assembly?

I must say that I often use through dados, but that's because I'm cack-handed, which you, clearly, are not.

BTW, after glue-up, especially with relatively 'fiddly' pieces like this, I find clean up to be a real chore - how do you manage this?



Steve,

I didn't see any particular advantge in using stopped dadoes. They would just make the piece that much harder to make. I don't think they detract from the look of the pigeonhole assembly.

For sanding, I just hit it with 120-grit paper on a ROS. It is difficult to avoid unwanted rounding over, but doing it by hand would exhaust my patience.

Rocker

[Rocker]

Today I dry-assembled the pedastals and placed them in the correct positions relative to one another, and installed the figure-8 fasteners used to attach the writing surface to their top edges (Photo 1). I then placed the writing surface in position (Photo 2), and marked the positions for drilling pilot holes in its under surface to attach it to the figure-8s. I had already drilled slotted holes in the writing surface for attaching the top unit that contains the tambour. Finally, I used a 3/4" forstner bit to drill 8-shaped recesses in the underside of the writing surface (Photo 3) to accommodate the fasteners.

I needed to drill a hole in the writing surface for computer cables, and puzzled for some time as to how to make it look tidy. I drilled the hole with a 28 mm forstner bit and found that black 25 mm i.d. irririgation pipe is a very snug fit in the hole. When pared off flush with the surface with a chisel, it actually looks quite flash.

Rocker

[Harry72]

Starting to really take shape Rocker, those pigeons are going too like them holes!

[ernknot]

Wonderful work Rocker, a definate heirloom in the making. Really impressed you are using blackwood for all pieces and not veneer stuff etc. Can't wait to see the finished desk.

[damienhazo]

You rock Rocker! Any thought to which crane company you're going to hire to get that thing in the house?:D


Damien

[Wongo]

Great progress David.

[Rocker]

Any thought to which crane company you're going to hire to get that thing in the house?:D


Damien

Damien,

Actually, when the desk is disassembled into its four main components - the two pedastals, the writing surface/tambour/pigeonholes, and the back panel joining the pedastals - any one component is not unmanageably heavy. But I will be getting a removalist to take it to its final destination, Eagle, in the Colorado Rocky Mts. Fortunately, my son-in-law will be paying the freight:) I think it will be safer to send it as four separate components to be assembled on arrival.

As Wongo will know, Eagle's main claim to fame is as the place where Kobe Bryant, the basketball player, was accused, but acquitted, of raping a hotel receptionist.

Rocker

[Wongo]

Kobe taught us a very valuable lesson.

When you are famous, everyone wants a piece of you
When you are rich, you can get away from murder

[Rocker]

Kobe taught us a very valuable lesson.

When you are famous, everyone wants a piece of you
When you are rich, you can get away from murder

Scott,

Surely the lesson that Kobe taught ladies was to listen to their mothers and change their knickers daily. The receptionist's case never recovered from the revelation that she had worn the same pair for several days, and that another man's DNA was found on them.

Rocker

[Rocker]

Sorry about the digression.

The pedastals are now ready for glue-up. I made two web frames for each pedastal. They will be glued and screwed into 6 mm deep dadoes in the top and bottom rails. Since there is no cross-grain involved, glue can be used without worries about wood movement. I decided to install the silver-ash drawer runners before glueing up the pedastals, since drilling the pilot holes for the screws that hold the runners in position would have been difficult after assembly, owing to the narrowness of the pedastals. Installing the runners before making the drawers will require the drawers and the dadoes routed in the drawer sides to be made to accurate dimensions so that they fit correctly.

I also glued, as well as screwed, the drawer runners in position, since the frame-and-panel construction of the pedastal sides should ensure negligible wood movement.

The back panels of the pedastals have tongues that fit into grooves in the rear stiles of the sides.

Rocker

[Rocker]

The pedastals are now glued up with their web frames glued and screwed in their dadoes. The draw boards, which can be pulled out from under the writing surface on either side of the desk, are also completed. I described attaching breadboard ends to their front edges in another thread: http://www.woodworkforums.ubeaut.com.au/showthread.php?t=33390 . A cleat attached to the under side of each draw board acts as a stop, when it hits the back and front rails of the top web frame, preventing the draw board from being pulled right out or being pushed too far in.

I am currently working on the drawers, and will start cutting the dovetails tomorrow. I have cut all the drawer parts to size and ripped and crosscut grooves for the drawer bottoms and for the backs. I might try cutting the tails with the table saw. But it is probably going to be less prone to error if I cut the pins by hand, rather than attempting to cut them on the table saw as well.

I will rout the dadoes for the runners in the drawers sides on the router table. I plan to locate the dadoes precisely in the centre of the sides so that I can always rout the dado from the back of the drawer side to the front, with either the top or the bottom of the side against the fence, depending on whether it is a right-hand or left-hand drawer side. In this way, it will be easier to ensure that the dadoes are all stopped exactly the same distance from the front of the drawer side. This is necessary to ensure that the drawer fronts are stopped level with the pedastal front edges when the drawers are pushed home.

Rocker

Rocker

[Rocker]

I made a high auxiliary fence (Photo 1) to attach to my Kreg mitre gauge for cutting the tails on the table saw. I was pleasantly surprised at how quickly and painlessly it went; it was not necessary to achieve great accuracy in lining up the blade with the marked layout lines, since the fit of the joint will depend on how accurately the pins are cut. I found that the table saw produced a much smoother cut than could possibly be achieved with a hand dovetail saw, and this will make marking out the pins more accurate. The photo shows the tail cuts, with most of the waste cut away on the table saw and bandsaw.

Photo 2 shows the jig I made for chopping away the remaining waste using my 1/2" LN bevel-edged chisel. I designed the tail layout so that the width between the tails was just slightly wider than the width of the chisel. Photo 3 shows the completed tails. Chopping away the waste was quick, easy, and accurate, using the jig.

Rocker

[Rocker]

Much as I would have liked to cut the dovetail pins, like the tails, on the table saw, I was not confident that I could make the cuts accurately enough, especially with my less-than perfect eyesight. So I have ventured towards the outer reaches of the Dark Side, and cut the pins with a LV dovetail saw and guide (Photo 1). The guide takes all the skill out of the process - I like that:). I do not want to have to serve a ten-year apprenticeship before I can cut acceptable dovetails by hand.

I removed most of the waste on the bandsaw; I then used a fretsaw to get the parts that the bandsaw couldn't reach. Now I come to think of it, I should have used a 7° wedge on the bandsaw's table to enable this - I will do that tomorrow. Finally I chopped away the rest of the waste, using the chiselling jig shown in the previous post.

Rather to my surprise, the joint fitted pretty well. Photo 2 shows it dry-assembled before sanding. N.B. it doesn't matter that the groove for the drawer bottom is showing, since it will be hidden by the false drawer front. Like Amanda Vanstone, I wanted to have a bit of wiggle room for the fit of the drawer fronts to the pedastal sides, especially in view of the method I am using to hang the drawers.

Rocker

[Rocker]

At last, after much weeping and gnashing of teeth, the drawers are glued up and fitted in the pedastals. Hanging the drawers by the side-hung method, particularly if the cabinet sides are of frame-and-panel construction, requires great accuracy in their dimensioning, if they are to fit properly and run smoothly. The dadoes in the drawer sides must not be routed deeper than the thickness of the runners; otherwise the backs of the drawer sides will be liable to hit against the stiles of the cabinet's side panels when being pushed shut. It is also of paramount importance that the cabinet and the drawers are square.

Tomorrow I will make and fit the false fronts and the pulls to the drawers. This will complete the woodwork on the desk. I still need to fiddle with the tambour to ensure that it slides smoothly in its grooves, and to fit the tambour lock.

Now that winter is more or less over in the State of paradise, the weather is getting more conducive to applying the finish. Hopefully the project will be finished in the next two or three weeks.

[Auld Bassoon]

Hi Rocker,

I'm intrigued as to why you would use the table saw to cut the tails.

A marking knife and a decent handsaw will see the job done in a couple of minutes.. Easier and simpler!

[Auld Bassoon]

Aha! Rocker old pal!

A cheat there (attached false front instead of half-blind d/t's) if ever I saw one :). I should know, I use it quite often as well.

[Rocker]

The drawers are complete and fitted to the pedastals. To have made the drawer fronts with half-blind dovetails would have been tempting providence, in view of the side-hung method I am using to hang the drawers. I needed the wiggle room that false fronts allow.

Although there are four drawer pulls, there are actually only three drawers in each pedastal. The bottom drawer is double depth.

Steve,

I can't agree that cutting the tails by hand is easier or simpler than doing it on the table saw. Each cut takes about 30 seconds by hand, if you work fast, whereas on the table saw each cut only takes about 5 seconds. Moreover, on the table saw, the cuts are smoother and are guaranteed to be perpendicular to the face of the drawer side and to be correctly angled at 7°.

Rocker

[jmk89]

Rocker

Do you propose to fit out the bottom drawer with hardware to allow A4 files to be stored as suspension files - I think this would be a good feature, if you have the opportunity to incorporate it.

Jeremy

[Rocker]

Jeremy,

The desk is due to be sent to my son-in-law in the US, where they do not normally use A4-size paper. The drawers are too narrow to accommodate files hung crosswise but could take them lengthwise. But I will let the recipient equip it with the requisite hardware, which is more easily available over there anyway.

Rocker

[TassieKiwi]

Good stuff David. The Bandsaw Book has a method of cutting pins with a wedge jig. Why not tilt the table? Mine goes 10° left, 45° right.:confused: Love that °. Ta.

D

[Rocker]

I had forgotten that I still had to make the central drawer to go between the two pedastals; so I did that today, using half-blind dovetails. I will post a pic tomorrow. I cut the tails on the table-saw, with the two drawer sides clamped together, and cut away the waste between the pins using hand tools, except that I drilled away the bulk of it with a forstner bit on the drill press. I tried using the method of hammering in a scraper, but it was not very effective, so I just used chisels to clean up the pins. Drilling out most of the waste with the forstner bit speeded up the process considerably.

Rocker

Rocker

[jacko]

Rocker, I'm glad that someone else tried the "hammer in the scraper" method as in FWW. I was rather sceptical I must say. How many hours to date? Seems like rather a lot.
Jacko

[Groggy]

I tried using the method of hammering in a scraper, but it was not very effective, so I just used chisels to clean up the pins. Drilling out most of the waste with the forstner bit speeded up the process considerably.I tried the same technique last week on a drawer with half blinds all the way around. I found the scraper idea to be a dud, although the scraper was very handy for cleaning it up once the bulk of the waste had been removed.

Since I was doing it mainly at night I did it all by hand. At the end of the day my dovetails were quite good. Unfortunately, the other three corners were done at the start of the day :o .

[Auld Bassoon]

Hi David,

I'm glad that you commented that way on the use of a scraper hammered in to remove waste; I've tried that too, and even though the scraper was quite sharp it wasn't anywhere nearly as effective as a sharp chisel. I may have been doing things incorrectly though, which wouldn't exactly be a first...

[Rocker]

Jacko,

I haven't been keeping track of how many hours this project has taken, but it must be lots. I can see why rolltop desks are not commercially made these days. But I have found it a good learning experience. I am not sure I would want to have one in my house - they seem to be too reminiscent of the 1920's. But I have certainly enjoyed building it.

Rocker

[Rocker]

Here are a couple of pics. Photo 1 shows the writing surface having a coat of wipe-on poly applied. The blotchy appearance of some of the boards is a natural feature of the blackwood, which looks as though it must contain patchy areas which are impregnated with some sort of liquid.

Photo 2 shows the half-blind dovetails I did today. The bottom of the drawer front appears not to be square in the photo; but this must be an optical illusion. It is in fact square.

Rocke

[Groggy]

I stumbled across this the other day:

If you have minor gaps in your dovetails (not big enough to glue a shaving into) you can close them up by doing the following - before planing the surface, insert a small dab of glue into the gap. Lightly tap near the gap with a ball peen hammer (be careful!) until the gap is closed. Let the glue set then plane the surface smooth.

I found this after I'd planed the surface of course, but thought someone may find a use for it.

[Rocker]

Groggy,

I glue my dovetails with techniglue epoxy, which gives plenty of assembly time, and which is gap-filling, unlike PVA glues. It also does not stain the wood.

Rocker

[Rocker]

Here is a pic of the central drawer in place. The two cross-bars above it ensure that the separation between the pedastals is exact, so that the side-hung drawer will run smoothly on its runners.

Rocker

[Lignum]

I stumbled across this the other day:

If you have minor gaps in your dovetails (not big enough to glue a shaving into) you can close them up by doing the following - before planing the surface, insert a small dab of glue into the gap. Lightly tap near the gap with a ball peen hammer (be careful!) until the gap is closed.
.

Good call Groggy, then before the glue dries put a damp cloth (not to damp) on it and put the hot iron on it. The fibre will swell and the glue will set very quick and you will get a near perfect join. :D

[Groggy]

Here is a pic of the central drawer in place. The two cross-bars above it ensure that the separation between the pedastals is exact, so that the side-hung drawer will run smoothly on its runners.

RockerFrom what I can see so far, you've done a great job with grain selection, can't wait to see the whole thing.

BTW, the dovie repair tip wasn't for you, I just thought it might be topical for the thread. Techniglue epoxy huh? Must look that up...

[Wongo]

Oh Rocky, Rocky, Rocky.:D The last picture looks spectacular mate. You done well, very well.

[Rocker]

At last, the top is completed, and I am happy to report that the tambour works well. The finish is 3-5 coats of Minwax gloss wipe-on poly.

All that remains to be done is to apply the finish to the pedastals, which will take another two or three days. So the project will have taken me about two full months, maybe about 250 hours.

Rocker

[Bodgy]

Truly beautiful Rocky, well done!

[AlexS]

That's looking great, Rocker. Can't wait to see the whole piece assembled.

[MajorPanic]

David,

This desk is beginning to look like my dining chairs..............

Will you please get on with it!!! so we can admire a great job. ;) :D

[Auld Bassoon]

Superb as always!

The tambour looks rather spiffy too :)

[dai sensei]

Looking fantastic Rocker.:cool:

[ss_11000]

fantastic work Rocker

[jmk89]

Rocker

This has been a great journey for you and a great thread for me to read!

Some day, when I feel really strong, I am going to have a go at something like that...it would be really great to make and give a piece that will stay in the family for generations

Great work

Jeremy

[Carry Pine]

Great work but I noticed the price of a lock for the roll top in a catalogue- $60. Was a bit of a shock.

Carry Pine

[Carry Pine]

Rocker,

I'm sitting here next to my grandfathers roll top desk (1920ish) and you seem to be making a replica. This one is in Tas Oak or Vic Ash and it 'moved' in ther load when the removalists were taking it up a mountain. I took it apart to repair it and made me wonder after reading your thread:

* Does yours come apart for easy removal?
* Will you incorporate a locking mechanism in the drawers? This one is a simple set up and I would be glad to share.
* Two of the pigeon holes have been adapted to fit 2 small drawers- very handy, use them often.
* This one has a pull out writing panel on top of the drawers- don't remember if you had these. I hardly ever use these but it would generate the breadboard end debate if you included them.
* The handmade matching timber drawer handles are useful but simple enough not to take away from the beauty of the piece.
Willing to post photos if you are interested.
Carry Pine

[Rocker]

Carry Pine,

I got the lock for my desk from Lee Valley for US$19.40. If someone over here is charging $60, it is a rip-off.

The desk is made in three main parts for easy shipment; it will be shipped to my stepson in Colorado.

My desk does have draw boards - hence the discussion in another thread on breadboard ends.

I am considering making some small drawers to go in the pigeonhole unit, but the recipient did not want the main drawers to lock, so I didn't incorporate any locking mechanism for them. The mechanism described in the FW article looked like a major hassle to make and a possible source of problems, if it were to develop a fault.

I would be interested to see pics of your desk; I think they are a pretty traditional design, so it would not be surprising if mine closely resembles yours.

Rocker