Published 4Oct2015, updated 6Oct2015
This cabinet’s design is inspired from The Royal Crescent in Bath . The couple I was making it for (Randi Hasle and Anders Jynge) is in love with Bath, and they had mentioned that building several times. From the curved shape of that building evolved a curved back. Making it hang on the wall, with no part of the cabinet (except its four “feet” or “arms”) touching the wall was an extension of that idea: why hide the rear side of the back in the dark? Why not let the light and the shadow equally touch the wall behind the cabinet by making it “float”?
It’s about 65 cm x 160 cm x 25 cm. Two ratios are close to the golden ratio (1.62). The height-of-lower-part / width-of-lower-part (1.0 / 0.65) and total-height / height-of-lower-part (1.6 / 1.0). It ended up being as the prospective owners wanted it, quite sleek.
The basic design resolution of having the cabinet float could not be finished before I had a solid bracket design in the cabinet for the screws that were needed in the wall. The cabinet shall hang in two screws for at least a hundred years was my spec. I decided to design the bracket myself, where the basic idea was that the cabinet should be built around the bracket. Or the bracket built into the cabinet with lists around it and through it. I was lucky that Profi mekatronikkindustri was in the same building, and has some shared history with Autronica, my employer. They made two pairs of brackets on their Amada LC-3015 XI NT laser cutting machine (so I have a pair for another cabinet?) When I went to fetch them only the right angle bend was remaining. I waited, and they were hotter than fresh bread when I carried them in a cloth to my office. (Disclaimer: this is no ad for them, I only mean to say: such a company may be closer than you think!) You are free to use my design provided you tell so. (Disclaimer: I am not aware of any other similar design)
Here (above) you see how the brackets were built into the structure of the cabinet. All the lists are 6×44 mm pine, and the structure then become a frame of glued, laminated wood. You will see below (fig.5) that I have put a list through this structure to hold the frames together, and minimise future cracking by differential temperature coefficient (between steel and wood).
The left picture above shows the two side frames when glued, but not yet finished. The center picture shows the compass made from a metal wire and a pencil. My main table was large enough, as the radius became 107.5 cm. The circle sector should stretch the width but only be so deep that the rear side of the curved back had about 1.5 cm of free air to let the light pass. The right picture shows the horisontal center frame structure of the curved back. The frame needed to be very sturdy, so I didn’t want to make the curved back with lists in one height. So I needed to make something almost invisible in the finished cabinet. Observe the square 12×12 mm list that passes through the structure (already mentioned above). It passes through the frames on both sides. This design is used for the three horisontal parts in front and in the back. This, in addition to the lamination are the main design elements for the stiffness of the cabinet.
My small bandsaw (above, left) does the job. It normally sits on a shelf. My workshop is only 6×3 m so I can’t afford big tools or permanently positioned tools. It’s good to have the old living room table that I made in 1973 that doesn’t argue when I bore a hole or two in it. And then I know that the tools sit where they shall. The center picture shows one of the 12 three-way joints during a clamping. Then, all of a sudden I could bring the frame up to enjoy for a day.
With the precision I want (the thickness of a credit card is typically much better than IKEA’s, and the hinges I use are typically much better than IKEA’s in my opinion) I need to build the doors into the frame opening. I can’t afford a millimetre extra but then I err up to a mm tolerance on some parts. The doors were also laminated, typically from three 4×12 mm lists. I made the slots for the glass by extra depth in the top and extra depth on one of the sides. This way the glass can be positioned into the frame from the inside and then pushed to the side. The top is not a groove as such but only covers the glass as seen from the front (figure 13, center). The reason is that it’s not possible to push up and let down and push into left and push right at the same time.
The inside top and bottom needed special attention. For the inside bottom we decided to use matte glass. It then became part of the set of other glass shelves. But glass in the top? No, that looked unnatural, we didn’t think glass “up there” would be as nice as the alternative. (All the time I discussed many of these details with the couple for whom I made the cabinet). The alternative was a wooden board, that I glued from my standard lists (left, above). The challenge is to make such that it won’t crack after ten years. So I screwed the glued list board to three small crossbars. Then I slid it under a horisontal list in front and screwed it in the back only, so it’s free to move when moist expands it and dryness shrimps it. Also, a small hole to let air in and out of the confined space. But the important design-matter was to position the board higher in front and some 1.5 cm lower in the back. I thought that would amplify the felt depth of the cabinet, especially when seen together with the curved back. In the finished cabinet it looked nice. I painted (and swiped off) paint before I closed the space, so that moisture should have about the same effect on each side.
In the center picture above you may see the rear lists. Observe that they are about the length of half the height, since the earlier described horisontal crossbar stops them. But from the front they are the same lists. I was accurate to use the same boards going through (afterwards cutting them in two) and then I numbered them. With the intention of an optical effect of a deeper curvature they get wider by some 2 mm per board from each side towards the center, with the widest on the middle of the back. Except the center list, which is actually a little less wide, an effect of not concentrating (even if I try to switch off the radio when I need to concentrate). But it turned out quite ok.
The right picture (above) shows the knee or corner in the back, bottom. You can see the lists for side glass and bottom glass mounting (also fig.13, right). Observe the earlier mentioned horisontal 2x12mm list that gets into the right frame, which is visible once you see it. Preceding this is the center picture below. The smaller frame is glued into the side frames.
The rear lists I made into panel lists (left, above). They are 8 mm thick, so I decided not to make then have groove and tongue, which better might have prevented them from twisting. However, they are so short that I seem to have avoided it. Right (above) is my not so old DAB radio where I listen to NRK P2 (mostly), and if I go down just after dinner then there’s my favourite program Salongen. It may steal my concentration some times. They are switching off both FM and DAB in favour of DAB+ only in Norway in 2017, so even that radio will have to be replaced by then.
The bottom glass needs to be removed without destroying it, so I made a push-up button that may be pushed from underneath the cabinet (above, left). There’s a space there, too, where I stored some spare screws for the future and also the glass measures and the paint colour code.
We decided that there should be a central handle for all the doors. The upper and the lower right doors are released when the handle is rotated. I made it from 12 pieces of 2 mm lists. The drawing (left, above) was the basis, but as always, when one make a single unit it’s really a prototype and then the object in the end is more or less equal to the drawing.
To help with boring the hole for the handle exactly 90 deg in all directions I placed the drill table mount on top of the cabinet while it was laying on my table (left, above). With a prototype, doing this before mounting is only a dream. I made two washers around the wooden axle of the handle, one to avoid it scratching the paint and one to let the stopping nail on the inside/back also not touch the wood (center, above). I hope the plastic I chose will last a hundred years. If not, the nail may be taken out and the washers be replaced. Also, should the axle and hole bulge, a 600 grade sand paper will help. The handle was cut in such a way that the those parts were used for the handles on the two doors (right, above). I placed brass nails in bored holes straight through them, and sanded the heads down – since they, too, must never be compromised.
I glued and screwed each rear list at the ends. It was important to glue only at the ends since the lists must be able to expand and retract in width not to invite cracks. But I glued a wooden list as a “hoop” at the back of both sections (also see the last picture below). This is when I needed something heavy to press it down and something underneath to hold it. This made the back rather sturdy. During the construction of the cabinet I bought a Proxxon FET 27070 table saw (above left top), which I regret I haven’t done before. My small Minicraft I burnt in the process. (Disclaimer: no hidden ad here!)
We decided that we should not let the glass shelves lay in wooden frames, as that would detract from the slim appearance of the cabinet by adding more horisontal lines than we wanted. I wanted the rests of the glass shelves to appear minimal, but sturdy. So I laminated small pieces that I screwed and glued into some sloped spaces that I cut. I needed some templates to do that accurately enough (above, left). When the shelves came up (above, right) I saw that it must be possible to swipe the glass shelves clean, also from the underside, without having them fall out and get broken. So I had to add some stoppers that stopped them coming out and being pushed up. It ended up being quite ok, but I forgot to shoot that detail.
The two brackets needed to be hidden. I realised that this could be the weakest part if I made it a proper part of the cabinet. During transport and replacement such details might be destroyed. So I made the two as loose parts. Should they be destroyed somebody might or might not make new. But the cabinet would survive without a personal scratch. It least that’s the idea. The top of the inside of the door (above, center) has glass as it is when it’s in position. The loose nail holds the glass such that it gets most of its weight pushed close to the hinges (down diagonally), to avoid the door becoming sloped over the years. You can also see the back of the earlier mentioned corner nail. There is one in every corner. The glass in the side frames are fastened with lists that may be screwed out (above, right). I used Märklin screws for fastening model track: 1.6×13 mm part 74990. They are so small that the small black dots they make are quite ok.
The doors need to stop and they need to be held. The right doors stop the left doors with a vertical list on the left doors. Both doors are stopped on a piece of duraluminium that I took from a processor heatsink. It’s terrifically solid (above, left). The right doors also needed something to keep then closed and thus to avoid them becoming twisted over the years. I used neodym magnets (above, center) and a piece of iron. I then screwed the two metal pieces beside each other in the top and the bottom. Where all four doors meet on the middle the handle holds them.
By the way, in all the corners (8) and middle joints (4) I added rather thick brass nails crossing through the joints. They can’t normally be seen since I did it from above, below and inside. The purpose is to avoid them cracking up in sixty years time(..)
The four arches (above, left) make the three arches in the back, already covered. The middle arch was glued together from two of them. Press the picture above and watch the back of it. (If you haven’t discovered it yet pressing the pictures takes you to larger resolution.) You may see how I screwed (and glued) at the panel ends and the two “hoops”.
Finally, the painting I do with oil paint that I dry off with a cloth while it’s still wet. Then I always seem to discover some glue that I haven’t been accurate enough to remove, and have to sand it down and apply the paint again. With this technique I can do this many times and it’s not visible. At a final touch I saw that the paint was too flat (matte), so I added an extra layer with paint thinned with White Spirit. This made it the surface more saturated and not that flat.
There are 14 pieces of glass. The shelves are 6 mm, bottom and center 5 mm and the doors 4 mm thick. The middle and bottom glasses are matte.
I have estimated the time spent on this cabinet to some 200 hours. And of course, it’s thoroughly signed with cut letters, seen underneath. Maybe that signature, in addition to loving care by the owners, will help the cabinet survive, when they see my name, place and “2015” in it? Some time this will be a long time ago.
©2015 – Original design and woodcraft by Øyvind Teig, Trondheim, Norway
- Please help me with the English terms here, I’m not a woodwork-English specialist!
- Search words in Norwegian: Hjemmesnekret vitrineskap som henger flytende på veggen, dvs. lyset vil kunne skinne bak hele skapet. Buet bakplate. Oppheng med spesialdesignet brakett. Listverk og panel. Messinghengsler. Hengeskap
- Om du likte det, se forrige skap der.
- Glassmaker’n i Trondheim har kuttet alt glass for meg, alltid. NB! Dette er ikke noen annonse. Det er bare slik.
- The Royal Crescent, Bath, see https://en.wikipedia.org/wiki/Royal_Crescent.