JK cabinet with a 'V' door

[Woodbloke]

Noodling through my various books (and online) recently, I came across this cabinet with a shallow 'V' section door:



...which I initially thought was beyond my meagre abilities to knock together. However, with most things JK, it's not quite as difficult as you first imagine, though granted, some of his stuff is insanely complicated. If you have his little tome 'The Impractical Cabinetmaker' on your woody bookshelf, pages 145-151 give some details on how to construct the door, which is the hardest part. As ever, you need to sit down and carefully read his text to understand what's happening and even then, it's difficult. I started off (as normal) by starting to make a mock up in pine and it appears that the 'V' is one piece of wood:



...with the rebates for glass being formed by gluing on 5mm thick bits on the exterior. However, what he doesn't mention though is how to clean up the inside of the 'V' (arrowed):



...as this will eventually be a show surface on the inside of the door. I think this must be virtually impossible, or really difficult and time consuming to do....so then I had a brain wave...... why not veneer it?



I sliced off a piece of knot free pine, stuffed it through the drum sander, shot one half of the veneer joint square:



...and glued it onto an internal face, making sure that it was on the line and square. Once that's set, I'll shoot in the other bit at the appropriate angle and repeat the process, then it'll be split down into top and bottom rails, planed parallel and the 5mm sections then applied to each to form the rebates - Rob

 

[Cabinetman]

Deliberately difficult, it’s as if he is setting himself impossible challenges all the time.
Does it have to be done that way? I’m thinking one piece of wood cut to form the angle with at an angle stopped grooves and a spline? Lots of gluing area, so will be strong.
Ian

 

[Mike G]

Don't you think, Rob, that cutting it that way produces one strong wing, and one very weak wing, with short grain everywhere? Might you not be better off tilting it so that BOTH wings have the same grain orientation?

As is often the case with Krenov (in my view), this looks like something which sets out to be complicated for the sake of being complicated.

Pick your hardware early, I suggest. The photo looks like there is a lock set into the cabinet side, which would mean it has a hook into something on the door. That may not be terribly easy to find.

 

[Woodbloke]

Deliberately difficult, it’s as if he is setting himself impossible challenges all the time.
Does it have to be done that way? I’m thinking one piece of wood cut to form the angle with at an angle stopped grooves and a spline? Lots of gluing area, so will be strong.
Ian

Always difficult Ian, but sometimes more difficult than others

Don't you think, Rob, that cutting it that way produces one strong wing, and one very weak wing, with short grain everywhere? Might you not be better off tilting it so that BOTH wings have the same grain orientation?

As is often the case with Krenov (in my view), this looks like something which sets out to be complicated for the sake of being complicated.

Pick your hardware early, I suggest. The photo looks like there is a lock set into the cabinet side, which would mean it has a hook into something on the door. That may not be terribly easy to find.

I thought so too, but on reading the text JK discounts two separate pieces to form the 'V' as the jointing would be unnecessarily difficult which is why he makes it in one piece. The angle of the 'V' is only 15deg, so the short grain is going to be negligible and there certainly won't be much pressure or force on the door when completed.

Some of his work looks simple and is to make, some of it looks a bit more complicated and is to make (as is this) and some pieces made only once are insanely difficult. The thing I like about this 'V' section door is that it's interesting and unusual, getting away from the normal 'flat' face of a show cabinet.

I already have the hinges from a previous project that didn't work out and there won't be a lock on this project - Rob

 

[Raymedullary]

Whenever I see "JK" I think Harry Potter............... make of that what you will

 

[Mike G]

.......I thought so too, but on reading the text JK discounts two separate pieces to form the 'V' .......

No, no, I meant cut it from the solid, but instead of having one wing following the long edge of the board and the other at a steepish angle, have BOTH wings cut at an angle to the face/ long edge, like this:

 

[Woodbloke]

No, no, I meant cut it from the solid, but instead of having one wing following the long edge of the board and the other at a steepish angle, have BOTH wings cut at an angle to the face/ long edge, like this:

View attachment 39142

I like that Mike a LOT Mike, and agreed, it reduces the 'angle of dangle' (to put it crudely ) 7.5deg rather than 15. I'll incorporate that idea into the real door, remembering that the one I'm making at the moment is a prototype. It also means that the thickness of the initial blank (probably Walnut in the real deal) can be somewhat thinner - Rob

 

[AndyT]

Rob, if you look at the photo at the bottom of page 146, Krenov's offcut is symmetrical, so I think Mike's suggestion is on the money.

The project will certainly need great accuracy and the ability to cut and plane mitres well. Luckily, you're the one tackling it!

 

[Mike G]

I think the very strongest way would be to laminate those two pieces. Make a boomerang-shaped piece, and then shape from that. That would eliminate any short grain. I mean, that's a second-best alternative from cutting it from a grown bend (ie a fork, or a root, or just a bend), but it would be plenty strong enough.

 

[Malc2098]

That angle is quite similar to the angle of the headstock on my instruments. I bandsaw the angle, turn the short piece over, flatten the sawn faces and glue the short length to the long length. To strengthen further, I veneer both sides.

 

[Woodbloke]

Rob, if you look at the photo at the bottom of page 146, Krenov's offcut is symmetrical, so I think Mike's suggestion is on the money.

Looking at that pic on 146, you're absolutely right Andy! Thanks to Mike G for steering me on the correct path. JK does say at the end of that section that quite a lot of folk seem to give up with this door so we shall see what transpires! - Rob

 

[Cabinetman]

Oh yes! Sorry I assumed you were doing that anyway Rob.
Still say a joint with a spline is the easiest way and perfectly strong enough. Is it a case of he didn’t do it that way?

 

[Woodbloke]

Still say a joint with a spline is the easiest way and perfectly strong enough. Is it a case of he didn’t do it that way?

Indeed Ian, made from one bit of wood cut into a shallow 'V' - Rob

 

[Woodbloke]

Overnight I've been pondering this 'V' section door and with the interior veneered with 2mm stuff and the exterior glued with 4mm pine to form the glass rebate, the thickness of one of the prototype door rails is 20mm, which is way too fat:



The core section needs to be thinner so that the rebate wall is a good 5mm thick and the overall thickness finishes at around 18mm, so if my maffs is correct, the core needs to be 11mm or thereabouts.

However, that leaves the core dangerously weak!

As Mike G suggested, the strongest way to do it is to laminate, but that would leave a rounded corner both on the inside and outside of the 'V'...and then how do I machine the rebate on the router table? Difficult if not bloody dangerous.

So last night I was giving it some deep thought. Then I recollected that a box maker from NI (Neat??) who reinforces the corners of his boxes with internal 'L' shaped pieces of 6mm ply which could, in theory, work here. On the piece that I hadn't applied the rebate timber to I machined:



...a coupla 6mm slots by passing the job between my two stops on the router fence. These are pretty ruff n'ready, the underside being a bit of a dogs dinner:



...but you can see the idea, hopefully. I glued in a couple of bits of 6mm multi ply:





...and if I were to do three slots rather than two, it ought to be pretty robust, n'est pas? (gaps at the end easily filled with oddements). Once flushed off and covered with 6mm pine, you won't know they're there - Rob

 

[spb]

Just to take the idea to its logical overkill conclusion, if the core is never going to be seen then you could in fact make it entirely from stacked plywood sections in the same orientation as those splines. Or, for a less extreme version of the same principle, you could make 4-5 sections the same way you have done, but with the direction of the scarf alternating, and stack those together.

 

[Mike G]

.... As Mike G suggested, the strongest way to do it is to laminate, but that would leave a rounded corner both on the inside and outside of the 'V'...and then how do I machine the rebate on the router table?

Make the lamination over-sized, and then cut back to your final size on the bandsaw, inside and out, to leave precisely the same shape you have now but with the grain running parallel to the "wings".

 

[Malc2098]

The two slot repair is exactly how guitar repairers mend a broken neck at the headstock join/bend angle.

 

[AndyT]

Rob, this is all interesting discussion, but if Krenov made a success of that inner corner in solid timber, so could you.

Your original issue was how to get a crisp internal corner where no plane can reach.

One answer is to use a broad paring chisel and a guide block.

@Planeiron showed us this in detail on his diminished stile door. Ok, that's on edge not face, but the same principles would apply.

Diminished (gunstock) stile door

In the process of making a diminshed stile door and learning along the way so this one unlikely to be a quick build. Is being made from oak and will be approx 920 x 2040 mm so will be a beast to lift and not looking forward to moving it into place whenever that will be. So amongst other things...

www.thewoodhaven2.co.uk

Alternatively, you told us how to use a chisel plane, with the same sort of sideways slicing action, last March, building your owl cabinet:

I reckon you could get a nice crisp internal transition by hand planing and paring.

 

[Woodbloke]

Quelle jour! I've abandoned the idea of a single piece 'V' as no matter how much I faff around with ply inserts etc I'm still going to have short grain and I really can't be ar$ed to go down MikeG's route and laminate it, so I'm going to take up Ian's suggestion and use two bits for each rail with a spline to join them. The day has been spent sorting this out making jigs etc and generally experimenting but I think I'm just about there.

Pics and text to follow after tea - Rob

 

[Cabinetman]

Hell I hope it works!!

 

[Woodbloke]

I decided that to do this properly, there needed to be a bit of precision, even though it's only still a prototype. First up were the original 'waste' wedges from the pine block, which were planed as a matched pair and will be needed later:



I then prepared some pieces of pine, 12mm thick; with a little bit of trial and error I found that a 22mm thick blank under a wedge on the shooting board:



...gave me the exact angle as the previous splined, short grain rails:



...as shown above. I decided then that a 4mm wing cutter on the router table would probably give me quite a deep spline, so I spent some considerable time faffing around with a jig. I tried to cobble together something from an existing jig that I'd had for yonks, but it was so worn that I eventually made another:





As you can see, it's simply a board that runs against the router fence and onto the board is a relatively high, right angled fence, tall enough to allow a cramp to be used. My trusty 22mm high blank from the shooting board enabled the angled face to be cramped at the correct angle ready for machining. I found that it's crucial to make just one pass with the cutter as repeat cuts never follow the same path, which leads to a wobbly spline in the groove! It's for this reason alone (the inaccuracy of the cuts due to a crappy earlier jig) it took me almost all afternoon to sort out. Once I'd found a suitable bit of 4mm ply and planed it for a snug fit, the final joints shown dry:



...are pretty good with all edges meeting quite well. Tomorrow's task is to conjure up a gluing jig to remove any of the very slight inaccuracies that can be seen in the above pic. I've already roughed out a blank which will be refined on the 'morrow - Rob

 

[Cabinetman]

Glad you’re able to get it working Rob, I think that will be well strong enough.
But we can’t let you off without a couple of questions!
How will you cover the spline showing? Would it be doable with a circular spline that didn’t cut through the edges of the rails? Obviously you would need a different jig to cut it though - push instead of sliding.
Now we’ve seen you use ply for splines before and whilst ply is of uniform thickness it is only as strong as the plies that go across the joint, personally I’ve always used solid wood, what think you?
Ian

 

[NickM]

Rob, this is all interesting discussion, but if Krenov made a success of that inner corner in solid timber, so could you.

Your original issue was how to get a crisp internal corner where no plane can reach.

One answer is to use a broad paring chisel and a guide block.

@Planeiron showed us this in detail on his diminished stile door. Ok, that's on edge not face, but the same principles would apply.

Diminished (gunstock) stile door

In the process of making a diminshed stile door and learning along the way so this one unlikely to be a quick build. Is being made from oak and will be approx 920 x 2040 mm so will be a beast to lift and not looking forward to moving it into place whenever that will be. So amongst other things...

www.thewoodhaven2.co.uk

Alternatively, you told us how to use a chisel plane, with the same sort of sideways slicing action, last March, building your owl cabinet:

I reckon you could get a nice crisp internal transition by hand planing and paring.

I think I'm with Andy here. Have you tried cleaning up the internal corner on solid timber? I think some cross-grain planing/chiselling and downhill scraping should do the trick.

 

[Woodbloke]

Glad you’re able to get it working Rob, I think that will be well strong enough.
But we can’t let you off without a couple of questions!
How will you cover the spline showing? Would it be doable with a circular spline that didn’t cut through the edges of the rails? Obviously you would need a different jig to cut it though - push instead of sliding.
Now we’ve seen you use ply for splines before and whilst ply is of uniform thickness it is only as strong as the plies that go across the joint, personally I’ve always used solid wood, what think you?
Ian

I could use a solid bit of wood Ian and that would probably be stronger, so good suggestion. As to the spline showing, if I made said splines from a piece of contrasting timber (say an offcut of ebony) that could be left as a feature but it would rarely be seen. Something like this:

I think I'm with Andy here. Have you tried cleaning up the internal corner on solid timber? I think some cross-grain planing/chiselling and downhill scraping should do the trick.

With a bit of luck and a following wind, there ought not to be too much to clean up on the internal 'V' but yep, chiselling and scraping will probably be involved. The timber used for the project will probably be Walnut which lends itself to this sort of cleaning up - Rob

 

[Cabinetman]

I could use a solid bit of wood Ian and that would probably be stronger, so good suggestion. As to the spline showing, if I made said splines from a piece of contrasting timber (say an offcut of ebony) that could be left as a feature but it would rarely be seen. Something like this:

View attachment 39204




With a bit of luck and a following wind, there ought not to be too much to clean up on the internal 'V' but yep, chiselling and scraping will probably be involved. The timber used for the project will probably be Walnut which lends itself to this sort of cleaning up - Rob

Yes, would it be possible to get the groove cut without tear out? Suppose you could recut after with a chisel, or hide the spline and keep em guessing.

 

[Woodbloke]

Yes, would it be possible to get the groove cut without tear out? Suppose you could recut after with a chisel, or hide the spline and keep em guessing.

It can be machined without tear out if the cut is made very slowly as the cutter exits the wood; I do exactly the same thing with splines for picture frames. The other option I suppose is to machine the wood oversize in the width by say, 4mm and then stuff it back through the pt once the slots have been cut and this would then remove any errant break out - Rob

 

[AndyT]

But if you just follow Krenov's example and use solid hardwood with no spline, it will be as strong as his examples, won't it, fit for careful opening and closing by its discerning owners?

 

[Woodbloke]

But if you just follow Krenov's example and use solid hardwood with no spline, it will be as strong as his examples, won't it, fit for careful opening and closing by its discerning owners?

That's true enough Andy, but this is just an alternative way of doing it without having to veneer the inside of surfaces of the 'V'. To do so means that the actual thickness of the short grain timber (albeit not very short grained) is only 10mm which ain't a lot. Doing it this way means that I've got a bit more thickness which isn't short grained and it ought to be strong enough provided I've made a respectable job of the glue joint. This is the fairly Heath Robinson gluing jig:




....with a male and female block in the centre, clamped firmly with a couple of deep throat G cramp and a pair of sash cramps each pulling up against a block screwed to the base board. The first bit came out a while ago:



..and it doesn't look too shabby. Once I'd spent a few minutes cleaning it up:




.. it doesn't look too bad. The shallow 'V' is particularly 'awkward' to clean up especially as there's glue squeeze out to remove from the surface, so that will probably have several coats of polish when it's done for real.

The last pic shows a rail that JK made for the piece from one solid blank but it defeats me how:



...he managed to clean up a bandsawn surface to the one shown in the pic - Rob (answers on a postcard)

Edit - I've just scrutinised the pic above on the 'The Krenov Archive' with the MK 1 eyeball and if you look very, very carefully you can just make out a glue line where the inside surface meets the rebate, so the cunning old sod must have cleaned it up as far as possible and then veneered it! That little detail isn't mentioned anywhere in his books as far as I know. That answers the question, init!

 

[Woodbloke]

As shown in JK's pic above, the outer sections (7mm thick) of the door rails were glued on as shown below (pic taken from TIC):



...except I used a pair of 20mm screws rather than a couple nails. Both the rails were then planed to the right thickness (18mm) taking a few shavings from the exterior surfaces:



Each rail was cut to size and the ends shot in, then I remembered just in time that they needed to be 'handed' so had to adjust the pencil marks accordingly. I then selected a bit of pine from the oddments box, only to find that it had a humungous split all the way down one side so it needed some tlc to repair the thing. This type of 'fix' is really, really hard to repair with normal cramps but:



...if you go to your local bike shop and scrounge half a dozen old cycle inner tubes, slice them up into long strips, you'll find you can make a very effective 'bandage' which exerts a serious amount of pressure when the rubber's stretched. Don't try it on your fingers 'cos they'll turn white...don't ask me how I know! - Rob

 

[TrimTheKing]

...Don't try it on your fingers 'cos they'll turn white...don't ask me how I know! - Rob

My lad, Barney, had ingrowing toenail surgery a few weeks back and the podiatrist who did it used exactly that method to isolate the toe while doing the procedure. Was fascinating to watch as he wrapped it around the whole toe, used some kind of clip to join it, then unravelled from the tip of the toe end, leaving the tourniquet at the base of the toe.

This was all after numbing the toe with local anaesthetic, then the wrapping meant no blood flow to the toe while he chopped and snipped.

As you suggest Rob, whole toe was white for the duration.

 

[Cabinetman]

As shown in JK's pic above, the outer sections (7mm thick) of the door rails were glued on as shown below (pic taken from TIC):

View attachment 39223

...except I used a pair of 20mm screws rather than a couple nails. Both the rails were then planed to the right thickness (18mm) taking a few shavings from the exterior surfaces:

View attachment 39221

Each rail was cut to size and the ends shot in, then I remembered just in time that they needed to be 'handed' so had to adjust the pencil marks accordingly. I then selected a bit of pine from the oddments box, only to find that it had a humungous split all the way down one side so it needed some tlc to repair the thing. This type of 'fix' is really, really hard to repair with normal cramps but:

View attachment 39222

...if you go to your local bike shop and scrounge half a dozen old cycle inner tubes, slice them up into long strips, you'll find you can make a very effective 'bandage' which exerts a serious amount of pressure when the rubber's stretched. Don't try it on your fingers 'cos they'll turn white...don't ask me how I know! - Rob

Haha Rob, I thought I was the only one tight ( and ingenious ) enough to use old bike inner tubes, and car ones too, for bigger jobs, it’s surprising how often the loops get used.
Hadn’t realised that you could hide the spline in the rebate, that’s good !

 

[Woodbloke]

Haha Rob, I thought I was the only one tight ( and ingenious ) enough to use old bike inner tubes, and car ones too, for bigger jobs, it’s surprising how often the loops get used.
Hadn’t realised that you could hide the spline in the rebate, that’s good !

Thee and me Ian. Old cycle inner tubes are incredibly useful for all sorts of cramping jobs where ordinary clamps don't cut the mustard. I've got an old ice cream tub full of strips of rubber - Rob

 

[slimshady]

Looking very good.
Questions if I may . . . .
Would a dovetail spline be better than a straight spline.
Or, does >GaryR< know of any Japanese joinery for same purpose.

Alex.

 

[Woodbloke]

Would a dovetail spline be better than a straight spline.

Alex.

Technically, I suppose that a dovetailed spline would be better but......



...underneath this image I posted earlier, JK (who took the pic) also added a note:

"Photo credit: James Krenov
A rabbet for the glass is created by gluing a face onto stock sawn to shape. The process has the additional benefit of being able to choose a show face. If the show face was simply sawn, the resulting grain pattern will shift in a predictable, but likely disturbing manner. The sawn sub-rail is also stronger than would be the case if the vee was created by joining two pieces with a necessarily short tenon"

The last sentence of his own text is worth noting - Rob

 

[Woodbloke]

A final update on today's progress. Having produced both the 'V' rails, the vertical styles were domino'd in place; they appear narrow but an extra section will be added to make up the width, as can be seen in the original JK pic above in the first post. The door went together quite well but there was a tiny teeny bit of twist:



... which was removed with a few strokes of a try plane. The cast iron surface of the router table is a very handy method of checking that all the twist has been removed and with that done, the limits of the stopped rebates where marked out with a sharp 4H pencil from the rails. The router table was then set up to machine the rebates very carefully, taking one or two mm off:



...and sneaking up to the line. The cutter was stopped a few mm from the corner so the waste will be chiseled out once the door has been glued. The next task is to work out how JK made and fitted the central rail, bearing in mind that it's around 12mm wide...that's if I've read the book correctly!! - Rob

 

[Cabinetman]

Well sorry to say this Rob but shouldn’t the central glazing bar have been morticed in before you glued up?

 

[Woodbloke]

Well sorry to say this Rob but shouldn’t the central glazing bar have been morticed in before you glued up?

Not glued yet Ian, the frame is shown together dry - Rob

Edit - I don't seem to recollect JK mentioning a mortice but he does prattle on about a 'notch', whatever that is...

 

[Woodbloke]

The prototype frame is now in the first stage of the glue up as one side is done at a time, the other being 'dry':



JK shows a stopped chamfer on the edge of the rebate, which I machined on the router table using a small bearing guided cutter:



Once the whole frame is glued and set, I've then got work those rebates into the corner so they join up, but the only practical way I can think of is to use a fine half round file, come in from the side at 45deg and then finish off with a chisel/sandpaper to blend the whole thing in. Open to suggestions though if any bright spark can think of an alternative? - Rob

 

[AndyT]

It does get more challenging when machining blends with hand tool methods.

It now looks like it would have made sense to gauge lines on the backs of the rebates before assembly. You could still do this from the far edges, assuming the widths are spot on. (A fair assumption!)

I'd use a sharp pin gauge, which will mark a stopped line nicely, especially if you press the pin down first, where the intersection of the two lines will be.
Then it's just chiselling...

 

[Woodbloke]

It does get more challenging when machining blends with hand tool methods.

It now looks like it would have made sense to gauge lines on the backs of the rebates before assembly. You could still do this from the far edges, assuming the widths are spot on. (A fair assumption!)

I'd use a sharp pin gauge, which will mark a stopped line nicely, especially if you press the pin down first, where the intersection of the two lines will be.
Then it's just chiselling...

I get your drift Andy, but the chamfers as shown were machined prior to glue up, which is the easy peasy way to do it. My quandary is that I have to blend in those two right angled chamfers into a smooth curve, not a mitre, which is a bit more tricky! Were I to attempt it with a chisel, I'd be chiseling 'uphill' against the grain which would cause all sorts of complications; hence my idea to come in from the side at 45deg on the down plane with a smooth, half round mw file, then finish off with a chisel where needed - Rob