• Posts

  • Joined

  • Last visited

  • Days Won


Everything posted by derekcohen

  1. The legs are on. I must admit to mixed feelings at this stage. This is not my style of leg, but it is what my niece wants. Perhaps I will feel differently with a finish on the wood. The tenons were kerfed for a wedge ... Installed in the bases ... And glued into the socket. Note that only the first third is glued. The rear is free to move ... The bases have been shaped to reduce their impact ... The legs were evened up .. Side view from underneath (one does not see the base otherwise) ... Regards from Perth Derek
  2. Mark, you recall that this table will be a wedding present? Well, the groom’s father is a woodworker as keen as you and I, and very knowledgeable . No doubt he and I will at some stage be discussing the build in some detail. Regards from Perth Derek
  3. Hi Coop That is a straight edge/guide I made for sawing sliding dovetails. It is made from Jarrah, with one square and one angled (1:6) side ... The underside has non-slip made from 240 grit sandpaper ... These photos were taken before I inserted a series of rare earth magnets along each side, which will pull a saw blade against the side. In addition to chopping at 1:6 (9.5 degrees) .. ... sawing sliding dovetails or dados ... ... it can be used on the square side to plane through dados with a HNT Gordon dado plane ... ... or at an angle to plane the female socket of a sliding dovetail (with a plane I made for this task) ... The two planes here ... It's a useful item. Regards from Perth Derek
  4. Following hot on the heels of the last post, where we created the male or tail section of the tapered sliding dovetail, now comes the female or pin socket to house the base for the legs. These are the bases. This post will focus on the socket for the one closest the camera. The base is positioned exactly 3 1/4" from the side. The tapered side is on the inside, with the outside face square to the front and rear of the case ... This process is essentially the same as transferring marks from the tail- to the pin board with drawers. The base tapers towards the toe, that is, the sliding dovetail will tighten up as the base is pushed into the socket. The first step is to register the far end of the base in such a way that the position is repeatable. This is done by placing a long board along the "square" side. The position for the end of the board is marked ... Now the base can be stood up to mark inside the tail with a scratch awl. You can make out the mark aligning the baseline of the tail ... Look carefully for the dots. This is repeated at the other end. The dots are now joined up ... The plan is to saw the socket sides, as if sawing dovetails in a drawer. The angle ratio is 1:6, as it was with the base. Since the socket is blind or stopped, the saw needs to have space in which to begin the cut. An area at the toe is excavated with a router. The depth of the cut is set using a 7.0mm drill bit. I am aware that the actual depth is 7.5mm, but this will be a second pass. I intend to clear the waste with the router - this Jarrah is bloody hard, and I am not a masochist! Using an angled saw guide, the end is chopped to the line .. Now this is space to register the azebiki saw ... I have roughly marked a depth to aim for ... Both sides have been sawn ... The waste is removed with the router, leaving a few mm close to the sides ... This is chopped away with a chisel in two passes, and then cleaned up with a hand router ... The side rebate #79/dovetail plane is used to clean any rough sections .. The power router drops a 0.5mm to 7.5mm and this is cleaned up ... Amazingly, the base slides in and tightens up about 1/4" from the end. It will need a tap to be fully secure. That's it for now. Regards from Perth Derek
  5. As a reminder, we are building a version of this table ... The plan is to attach the legs, which were made near the start of this project. The attachment method is by inserting the legs into compound angle mortices in a base, which will be fixed to the carcase with a tapered sliding and stopped dovetail. We don't mess about here! It will be necessary to do this over two articles, the first being the base for the legs, which will be dovetailed (tails). The second will be the socket (pins) for the base. Before we begin, I want to mention what I did at the end of the last session. I had replaced the central drawer dividers as the grain ran in the wrong direction. The spacers at the ends also did so, and my response was to cut out half the spacer ... Well, I fretted over the end spacers, and just could not leave them this way. Encouraged by the way the halves had come out cleanly, I removed the remainder and replaced the spacers with correctly grained versions ... OK, onto the leg base ... I spent a while playing with angles for the legs, and finally accepted this (mocked up base) ... I have drilled angled mortices with a brace on a number of occasions. This time I decided to used a drill press and some Japanese Star-M augers, which are specially designed for this type of work (no lead screws). I built a 10 degree ramp for the resultant angle. The auger is 30mm ... [ The tenon is straight, but the mortice will receive a slight reaming, and the tenon will be glued and wedged. This is probably overkill since the weight of the case rests on the legs. These are the bases for the legs. The final prototype is at the rear ... Drilling the bases ... The design requires that the legs do not go over the boundary of the case (to avoid tripping over them) ... This is how they should be ... There was a small dilemma: The base at one side measures 3" from the end ... ... and the other side measures 1/4" further ... I could not work out how this occurred. The angles are the same. In fact, I made another set of bases, and the same error showed up again - exactly the same! So what to do? Actually, the decision was obvious after a little think - make the bases the same. What is more likely to be noticed is if the bases are different distances from the sides. No one will notice a 1/4" difference where the legs hit the ground. So be it. This is one of the bases for dovetailing ... First step is to remove a 2mm taper from one side. The taper will be on the inside of the base, with the outside parallel to the side of the case. Taper line drawn ... Easiest way to do this is with a #604 smoother .. This is the one end of the base ... .. and this is the other end ... mmmm .... 0.39 mm oversize. What to do ...? I'm kidding The dovetails will be 7mm deep. A shoulder was planed with a rebate plane ... The squareness of this rebate is important, so check ... The dovetail is now to be created, and the preparatory step is to colour the outer edge of the rebate with a sharpie. This will warn that the planing does not lower the external edge of the rebate. The dovetail is created with a modified Stanley #79 edge plane ... The fence has a 1:6 ratio wedge ... Details of this dovetail plane here: The result of planing. That is a 1:6 dovetail marker ... So what are the numbers for the taper? This will give an indication of the accuracy of the joint. One end is 44.12mm ... ... and the other is 46.46mm, which is a difference of 2.34mm. This will work - the pin socket is measured from this (in the same way as dovetails for a drawer. The reason for the 7mm depth? The case is 20mm thick. the dovetail should be about 1/3 of this thickness. I decided to take it to the depth of the rebate for the rear panel ... So, here is one of the completed bases ... And this is where it will be fitted ... Regards from Perth Derek
  6. I thank you for this - I was just curious how you found that page since I thought that I had not provided a link to it in an index. It is very out-of-date. Kind regards Derek
  7. It is not possible to make vertical cuts into the space. The sections came out so cleanly that it is likely that I will remove the side spacers and replace them. Regards from Perth Derek That is an old page which is no longer on my website. How did you access it? Regards from Perth Derek
  8. CS, which page (on my website) has this link? Thanks Derek
  9. Thanks for the heads up. The link should be sorted now. Cheers Derek
  10. Thanks Tom. I much appreciate your kind words. Cheers Derek
  11. The case was glued up yesterday, with everything tight and square as one could wish, but I did not sleep well. I was haunted by the thought that there was a problem that would come to a head some time in the future. If you look at the grain direction of the two centre drawer dividers, you notice that the grain is vertical. That is the way it should be. Wood moves, expands and contracts. It does this in reaction to moisture in the air. When it moves, it does so across the grain. That is why solid wood drawer bottoms have grain across the width - allowing the drawer bottom to move towards the back of the drawer, rather than towards the sides (where it will be blocked and then buckle). These drawer dividers will be butted up against the rear of the drawer lips and act as drawer stops. The front third of the divider will be glued in the dado, forcing any expansion towards the rear of the case. All good. The two spacers at the inside ends of the case have the grain running horizontally. I glued this in before I realised that I had cut them this way. I had done the same with the internal dividers, but re-cut them, as shown in the previous article. The end spacers will expand vertically, and to allow for this, I provided a 2mm gap below and above the panels. That is what kept me awake. The end spacers are 6mm thick. The case, to which they are glued, is 20mm thick and about 40mm wider. Initially I was concerned that the spacer would be overwhelmed by the case moving, and buckle. Having thought some more about this, I am no longer concerned that this will occur. Why? Because movement in the case would instead "stretch" the spacer length-wise. I started to breath again. In the end, I decided to reduce the height of the spacers by half. This would allow them plenty of space to expand, when necessary, as well as reducing their impact inside the case. Here is one side ... Taped for visibility and protection ... The saw is a 16" Wenzloff & Sons tenon saw (10 tpi) ... Three kerfs ... Deepened with a Japanese Azebiki ... ... and split out with a firmer chisel ... A Bahco carbide scraper cleans up ... The result ... Final cleanup was aided by the only shoulder plane that fitted inside the space ... Regards from Perth Derek
  12. The other item to factor in is that the case is 20mm thick and the internal side panel is only 6mm thick. Not only is it insulated, and has two expansions grooves, but it is 40mm narrower than the case side. I doubt it can have much impact. To be safe, I with add a 6mm groove through the centre. Regards from Perth Derek
  13. No, I would not try and plane it away. There are expansion grooves at the bottom and top. I am thinking that I might also add a groove through the centre - it will never be seen and it just reduces the surface area some more. What do you think? Regards from Perth Derek
  14. We ended the last session with the drawer dividers installed ... Everything was nice and square, but the more I thought about what I had done, the unhappier I became. Such an elementary oversight. I cannot believe I did it, and also that no one pulled me up for it. What was it? Two items: The first was that the grain for the drawer dividers runs the wrong way. Although the boards are as close to quarter grain as possible, which adds to stability, they will expand vertically. That could cause them to buckle, and then the drawers will not run nicely. The second is that I could have built in a way to close up the drawer dividers against the back of the (to-be-built) side lipped drawer fronts ... this is to be used as a drawer stop ... at this stage it would be necessary to add a filler. Not good. So I re-did the drawer dividers. Here is the rear of the case. The drawers are left long on purpose ... Provision is made for the dividers to be adjustable in length (to close up with the back of the drawer front). They are given rebates to slide further forward ... it will be necessary that they move around 15mm forward (to within 5-6mm of the opening). The rebate is 2mm deep (the depth of the dados), and largely created with a cutting gauge. The blade slices away end grain, and the resulting splitting away makes it easy to chop the remainder. Here are the dividers, further forward than before, and capable of moving a little more still ... The plan was to glue up the case. However, before this is done, it is wise to fit the drawer fronts across the width (the height will be done at a later date). This is the board for the three drawers. Removing one end, the board is set on the case ... It is now apparent that the front of the bevel, where it meets the drawers, is not straight. It is possible to see a small amount of flat ... This is especially noticeable in this corner .. This is fairly easy to remedy ... mark with a pencil, and then plane away the pencil marks ... Perfect now ... The other end needs no more than a smidgeon removed .. The upper side is now treated the same way. Interestingly, this needs no work at all. Time to saw the drawer fronts to size. First step is to mark the middle point of each divider (since the lips will share the divider). The mark can be seen in the rebate ... The drawer board across the front ... Transfer the mark, and then saw the drawer front ... This process is repeated. Here are the three sequential drawer fronts. You can just make out the breaks ... I am happy with this. And so, finally, the case is glued up (Titebond Liquid Hide Glue - reversibility and long open time). Looking like a trussed up fowl .... Regards from Perth Derek
  15. Coop, I simply used the bandsaw freehand close to the line, than planed to the line, and the table saw for the parallel cut. Regards from Perth Derek
  16. The basic case complete ... My niece's expressed wish is to have a table front looking as if it was faced by a single board. The original model for this project has two drawers. I did not see this working here since, as their width would be greater than their depth, two drawers would likely rack. Consequently, I decided to build three drawers of equal width (I considered a narrow drawer in the centre, but decided this would be too busy). In order that the figure of the drawer fronts would not be interrupted by the drawer dividers, the drawers are to have half-blind dovetailed side lips, such as these ... The drawers will each have a side lip of 6mm. This requires a 6mm wide side panel on each side of the case, and two 12mm wide drawer dividers. This will allow three drawers to run adjacent to one another, and the three fronts to be cut from a single board. The drawer fronts will come from this board ... Below are the panels for fitting ... It occurred to me later (of course!) that the 6mm end panels could have been made to run with the grain direction of the case. Being the same Jarrah, this would have counted for any expansion/contraction, and there would not be any danger of movement being intrusive. Too late. It's glued. So I did the next best thing, and planed 2mm off the upper and lower edges. This will permit enough movement, if any (it is a small and thin panel). There will not be any gaps seen as the front edges will later receive edging, which will be used as a depth stop. Frankly, the hardest part of this section of the project was accurate marking out of the two central drawer dividers. These need to be both perfectly parallel, and also aligned vertically (the lower panel with the upper panel). There is a second area that needed to checked, which is important for drawers to work well, and this that the lower panel is flat - that is, does not have any hills. I learned my lesson the hard way about this. All good. The way I go about marking the dados for the dividers is to make templates for their position. These are used on both the lower panel, as below, and then the upper panel ... The process is self-explanatory ... The dados are knifed deeply ... Chisel walls cut ... .. and then the waste is removed with a router plane ... The dados are just 2mm deep. That is deep enough to prevent any movement. This process is quick and relaxing (compared to setting up and using a power router). Once done, the process is repeated on the upper panel ... All ready for a dry fit. The rear of the case ... ... and the front ... Happily, all is square ... Tomorrow I shall glue it up. Regards from Perth Derek
  17. What I need are lipped drawers.The question was whether I make them the easy way, which is by planting (glueing) on fronts. Or, whether I build them out of one piece, which is a lot more work as it requires creating half blind dovetails in a rebate. For those unfamiliar with lipped drawers .. This is the work of Christian Becksvoort ... At this point, I am going to do it the hard way and make half-blind sockets in a rebated front. This is similar to building a secret dovetail. To do this for all the drawers, the insides of the case at each end will require a spacer, essentially a 6mm panel glued to the insides. Each side will be half the thickness of the two middle drawer dividers (each 12mm). The centre dividers will be attached in a dado top and bottom. Regards from Perth Derek
  18. Having completed the dovetailing of the case, the next step is to bevel the front face, and rebate the rear for a back panel. I had been considering a cove in place of a bevel, however when I mocked this up it came across as appearing too busy. So, back to the bevel. The angle for the bevel was finalised at 55 degrees. This enabled a 6mm (1/4") flat edge and a bevel that ran to roughly 4mm of the first dovetail. A 45 degree bevel would run into the dovetail. The lines for the bevel were marked and then roughed out on the table saw ... The table saw is a slider, and the rip fence was used to position spacers, before clamping a panel for cutting the bevel. The bevel was then finished with a hand plane ... This Jarrah is particularly interlocked but planes well with both a high cutting angle (the little HNT Gordon palm smoother) and a close set chipbreaker (the Veritas Custom #4). Once the bevels were completed, the rear rebate was ploughed ... Now the panels could be assembled into a case once again, and the work examined for tuning. Three of the bevels needed tuning. This ranged from a smidgeon here ... ... to a largish amount ... The case was dissembled and the bevelled edged planed down, re-assembled, checked, pulled apart again, planed ... The rebates at the rear turned out to not require any tuning, with the exception of one corner ... ... where I had obviously forgotten to plane! :\ That was easily rectified ( ... but the case had to be dissembled again). Finally, this is the rear of the case and the completed rebates ... This is a rebated corner ... Here are the front bevelled corners ... This illustrates by the mitres on the corners of the dovetailed case needed to be perfect. Any undercutting would show here. Next, the drawer dividers need to be done. I'll mention here - since I would appreciate the thoughts of others - that this area has been my biggest headache. The reason is that my niece would like the drawers to have the appearance of a single board. However, to achieve this, because of the bevels, is quite complicated. First of all, the table cannot have just two drawers. The width of the drawers will be greater than their depth, and this would likely lead to racking. Consequently, I plan to build three drawers, which will be more favourable for the width vs depth ratio.. Secondly, if the drawers have dividers between them, which they need (since I do not do runners), then there will be a gap between the drawer fronts (which will not flow uninterrupted). As I see it, there are two choices: the first is to build the drawers with planted fronts. This is not a method I like (but it may be expedient). The second option is to set the dovetailed drawers sides back (recess them) to account for the internal drawer dividers. Thoughts? Regards from Perth Derek
  19. Thanks CO. The mitres do need to be cut fat at first since I wish to avoid the danger of over-cutting them. It is safer to pare the waste, and if you are going to do this anyway, a smidgeon of fat there in the beginning does not matter. Plus, in the current build it is not possible to undercut/pare the mitres to make them fit. The reason is the inside edge on the model calls for a bevel/mitre all the way around. I plan to do a slight cove. An undercut would open up the mitre from the inside, and any undercutting would show. Regards from Perth Derek
  20. I've been away from the workshop for a month, travelling around a few cities in Austria and Germany, as well as Prague. It was a good trip, but it's great to be home. The current build was on hold. This is the entry hall table my niece asked me to build ... ... and this is where we left off last time - ready to fit the first corner ... Past builds: Part 1: Part 2: Today we shall put the complete case together. What I wish to focus on is the dovetailing. Not just any dovetailing, but mitred through dovetailing in unforgiving hardwood (here, Fiddleback Jarrah). Of all the commonly used dovetails, I consider the through dovetail more difficult than the half-blind dovetail. Why ... because two sides are exposed against the single face of the half-blind. In my opinion, by mitering the ends, the level of complexity is tripled .. at least. Not only are there three faces now, but each needs to be dimensioned perfectly, otherwise each is affected in turn. This is more difficult than a secret mitred dovetail, where mistakes may be hidden. I have posted before on building the mitred though dovetail, and it is not my intention to do this again. Instead, what I wish to show are the tuning tricks to get it right. This is the model of the tail- and pin boards … In a wide case, such as this, it is critical that the parts go together ideally off the saw or, at least, require minimal adjustment. The more adjustments one makes, the more the dovetails will look ragged. Tail boards are straightforward. Let’s consider this done. Once the transfer of tails to pins is completed, the vital area is sawing the vertical lines … well, perfectly vertical. I use blue tape in transferring the marks. The first saw cut is flat against the tape. Note that the harder the wood, the less compression there will be, and so the tail-pin fit needs to be spot on. Where you saw offers an opportunity for ensuring a good fit: if you hug the line (edge of the tape), you get a tight fit. If you encroach a smidgeon over the line, you loosen the fit slightly. Saw diagionally, using the vertical line as your target … Only then level the saw and complete the cut … I do not plan to discuss removing the waste. That was demonstrated in Part 2. So, the next important area is the mitre. These are scribed, and then I use a crosscut saw to remove the waste about 1mm above the line on both the tail- and pin boards … Now we are ready to test-fit the boards … Mmmm …. not a great fit … … even though the mitres at the sides are tight … The problem is that the mitres are fat, and the extra thickness is holding the boards apart … Even sawing to the lines here is likely to leave some fat, which is why it is a waste of energy to try and saw to the line in this instance. It needs to be pared away with a chisel, using a 45-degree fixture. As tempting and logical as it seems to pare straight down the guide … … what I experience is that the chisel will skip over the surface of the hard wood rather than digging in and cutting it away. What is more successful is to pare at an angle, and let the corner of the bevel catch the wood … This is what you are aiming for … Okay, we do this. And this is the result … Not bad. But not good enough. There is a slight gap at each side, quite fine, but evident close up. The source is traced to the mitre not being clean enough. It is like sharpening a blade – look for the light on the edge. If it is there, the blade is not sharp. If there is a slight amount of waste on the mitre, the case will not close up. To clear this, instead of a chisel – which is tricky to use for such a small amount – I choose to use a file. This file has the teeth on the sides ground off to create “safe” sides. Try again. The fit is now very good. I will stop there. So, this is the stage of the project: the case is completed. This is a dry fit … One end … The other … The waterfall can be seen, even without being smoothed and finished … Regards from Perth Derek
  21. Well, access to the workshop is a little difficult as I have been in Europe for the past week, and have another three more before returning home. Currently in Vienna, and next off to Berlin. In the two weekends before leaving, I did a little more dovetailing, but my focus was largely taken by setting up a new drill press - birthday present for January ... Merry Christmas from Vienna Derek
  22. Thanks Mark. One note - I changed the chuck because I wanted keyless, not because there was any issue with the Nova chuck ... I have not used it. Update: Apparently Albrecht guarantees their chucks to have .0015" run out or less. I think I am okay Regards from Perth Derek
  23. Yes, Wegner has some great designs. This is my favourite. Sorry, no templates. I also deliberately did not post dimensions when I wrote up the build as it is still manufactured under license... but you can read the details of the build here (scan down the page): Regards from Perth Derek
  24. I replaced the Nova chuck (which is actually a cheapish Chinese-made keyed chuck), with a 13mm keyless Albrecht clone I have used for several years. I had good results with this in my previous drill press, and was wondering whether to upgrade it or not. Consequently, I completed some run out tests for it. Set up ... I used a 1/4" shank carbide router bit as the test piece. This was a one-time test, so I may have had better results from another router bit, or from re-positioning it. It is what it is ... Results ... This reads 0.045mm run out. That is 0.0017" run out. Is that good or bad? My understanding is: there is run out that may occur with the spindle, then there is run out that will occur at the chuck and quill (which could also be measured separately), and finally there is the run out measured at the bit. The results here are a total of all these together. It was mentioned to me that around 5 thousands of an inch would be acceptable. I have 1 thousand inch. The other item I attended to was to add a Wixey laser guide ... It tucks aware and is quite unobtrusive ... It leaves a nice, clean line ... ... but it is a little wider than expected. The jury is out whether it is just a gimmick, or whether it will prove to be useful. Regards from Perth Derek