My latest jig build - this one is special

[Ola Carmonius]

Hi all,

Thought there might be some interest in my latest jig build. I try to explain it with pics and short words.

It's a table saw jig for tenons and box joints. It cuts single tenons, double tenons with individual spacing and boxjoints on all mm numbers larger than 4mm. In first video I said only even mm numbers but later found out it could also cut the odd ones, also boxjoints on halv mm numbers are possible.

Accuracy is better then 0.05mm and it's super fast to set up. The workpiece is clamped to a carriage, and the movement is acchieved by a large control driving a ball screw to which the carriage is connected. The pitch of the ballscrew is 4mm, backlash is less than 0.03mm and the control wheel is devided into 80 segments with holes at the ends. Pitch 4mm divided by 80 means there is 0.05mm movement between each segment of the control wheel and this is also the resolution of the system, in other words I can trim the thickness of a tenon or boxjoints in step of 0.05mm.

Coloured pins are used for positions on the control wheel where to start or finish the cut, when the pins matches a reference line on the jig. The control wheel is also equipped with a scale ring that can be zeroed in any position and is used to set the distance between the coloured pins.

The jig has customable templates/notes that keep track of the cutting process, so I don't have to count the turns of the control wheel. Coloured pins gives exact position, templates shows the bigger picture. If I made boxjoints or tenons of a certain type before, I just bring out the template and is ready to cut without test cuts or other preparations.

Set up time once used to the system is extremely short, I cut a tenon or boxjoints better than 0.05mm accuracy with a minute set up time.

Sounds very complicated in words but, this is actually a very easy to use jig.

[Ola Carmonius]

Some pics that shows how it looks in operation. Boxjoint pieces are clamped to the large area of the carriage at both left and right position (only left side clamps shows in pic). Tenon cutting pieces are clamped to the wing of the carriage.

Max width for a boxjoint pieces is 240mm and max for a tenon piece it's 8130mm. Super efficient built in dust extraction tunnel through the jig to the backside where a vacuum hose is attached, tunnel visible in attachment #2.

[Lee Schierer]

It appears that some of the parts were made on a CNC, is that correct?

[Ola Carmonius]

No CNC involved in this, just plain woodworking (and some very basic hand tool metal working) and standard off the shelf components.

Although the jig itself, once finished, performs on par with a CNC in terms of accuracy:)

[Tony Zona]

Good grief.

[al ladd]

This is remarkable! But I have to say CNC entry costs and DIY costs are low enough that if you go this far with precision, you might as well go the rest of the way...

Cutting box joints with a conventional jig works well, but I always found the actual cutting really tedious, and on harder woods, annoyingly loud and gnashy. I occasionally make box joints on my CNC, and am always cutting piles of hand cut-style through dovetails on my CNC , (which was built for joinery, with a spindle that goes horizontal). Though the precision is of course the usual reason to go CNC, for someone cutting lots of joints the fact that you can clamp it and then go do something else for the 30-90 seconds it takes to cut most joints is the real kicker. Also, you don't have to be directly next to the noise of the cutting action.

But the engineering and build quality in this jig is really something! And it makes me realize that a CNC positioner that would then slide a sled back and forth to make the cuts, making a true CNC joinery table saw device--now that would be really cool! And something the right person could perhaps design for mass production at a reasonable price. Box joints are an example of where the blade geometry of a table saw with a dado is a big advantage over a router with a spindle. The big boys have "aggregate cutters" for their massive CNC machines, but an attachment for CNC table saw joinery could easily weigh under 100 pounds.

[Maurice Mcmurry]

Thats excellent! It seems like that should patentable.

[Ola Carmonius]

@al ladd. Thanks and thanks for your thoughts on this jig vs cnc. For what I'm building which is typically one offs this system is faster for some operations. That's based on my cnc experience, a skilled cnc operator might have other opinion. I typically use it for cutting tenons and the set up time I would say is 1-3 minutes depending on if I've done the exact same tenon before and which blade I have in the saw. Imo very short time to acchieve a tenon within 0.05mm of target. If I cut 4 or 8 tenons which is a typical run for me, set up time is crucial. If I would cut 100 tenons same size, run time would be more crucial and there my jig would be beaten by both cnc and some other methods for cutting tenons.

For boxjoints the set up time is pretty similar but the time to make the actual cuts cnc is way faster. The boxjoints in my second post is pretty much max width the jig can handle, 240mm, and I think the complete run time was around 15 minutes for all four parts. I ran those in stacks of two. Compared to many existing boxjoint jigs mine (again) exceels in the set up time and how easy it is to trim the fit in steps of 0.05mm, but actual run time is the same as the rest of these type of manual boxjoint jigs.

Not really up to date on pricing for cnc's, anyway, the material cost for my jig was around 100-140$ (the lower cost corresponds to no digital readout, nice to have but not at all necessary) + wood.

But maybe most important, although I work with CAD every day I like to keep my shop computer free:)

[Dan Cameron]

Ola, I was quite interested to see your jig.489770489771489772

[Tom Dixon]

Very nicely done! What were the rail and ballscrew diameters? I'm not sure I completely understand how the the dial and the colored pins work and would love a more in depth step-by-step explanation with photos of that part. Did you make 3D drawings first or did you just wing it from an idea in your head?

[Chris Parks]

But the engineering and build quality in this jig is really something! And it makes me realize that a CNC positioner that would then slide a sled back and forth to make the cuts, making a true CNC joinery table saw device--now that would be really cool! And something the right person could perhaps design for mass production at a reasonable price. Box joints are an example of where the blade geometry of a table saw with a dado is a big advantage over a router with a spindle. The big boys have "aggregate cutters" for their massive CNC machines, but an attachment for CNC table saw joinery could easily weigh under 100 pounds.

Shark CNC have just released a router table kit with pre-programmed joints coded in.

[Carl Beckett]

It is a work of art! Very much appreciate the thought and engineering that went into it. Awesome!

[Ola Carmonius]

Very nicely done! What were the rail and ballscrew diameters? I'm not sure I completely understand how the the dial and the colored pins work and would love a more in depth step-by-step explanation with photos of that part. Did you make 3D drawings first or did you just wing it from an idea in your head?

Thanks Tom. Both rails and ballscrew are 12mm. This started out many years ago, inspired by Wandels with wooden gear wheels + other screw using jigs. My first version had the same layout as this latest version of my jig with rails and open construction with built in clamping. Never became friend with my first version, didn't like the backlash in the threaded rod and used it to seldom to remember the different gearing options.

Rebuilt it last christmas when I installed the ballscrew and invented the control wheel and the templates. No 3D at that time, this jig idea has evolved in my head over the years.

It's a bit difficult to describe in words how it works but I give it a try.

If you look on my fourth picture in my original post, you see a coloured pin and a black line on the jig. When my coloured pins line up with that line on the jig something should happen. If only two pins green indicates start or cut#1 and red indicates finish or cut#2. Attach a pic in this post of control wheel only.

Lets say I make a 12mm tenon and my blade kerf is 3mm wide. I rotate the control wheel to move the carriage/workpiece to where to cut the first cheek. In this position I zero the control wheel using the rotatable scale ring (the part with figures every 0.05mm on the outside of the control wheel). Here I set my green pin indicating a start cut. I should make the second tenon cut at 12mm + the blade kerf, 15mm total travel from where I made my first cut. The ballscrew pitch is 4mm so 15 mm equals 3 turns of the control wheel and then another 3mm. So, in the 3mm hole on the control wheel I set the red pin, stop cut. Then it's time to cut, I begin at my green pin, hold the control wheel steady and make cut#1. Then I crank 3 full turns + until I reach red pin. Here I make cut#2. Done.

The pitch 4mm is divided into 80 segments, meaning 0.05mm between each hole where I can set my pins. So, if a tenon is loose or oversize, I can move one pin and trim the fit it in steps of 0.05mm, same goes for boxjoints.

Scale ring can also be used for more advanced zeroing operations, for example setting the exact location of a tenon sideways, not only the thickness of it.

Maybe sounds complicated but a day or two and you're really up to speed with the mindset. 4mm per turn of the control wheel is pretty much the only thing to keep in mind.

The cardboard templates that the metal arrow is pointing on is there to keep track of the control wheel turns. So, lines on my template doesn't have to be exact, they are there only so I don't have to count the turns of the control wheel when nothing is supposed to happen, in the case of the tenon the 3 turns freewheeling.

I only make templates once for a tenon, double tenon or a boxjoint type. Next time I do the same joint I bring out the matching template where I also have written the pin positions. That means I have minimum set up time if I have made the same joint before.

That was for a single tenon but hope you understood a bit more. Boxjoints works the same way pretty much but there I make multiple cuts between start colour pin and end colour pin, and then some turns of freewheeling the control wheel until process starts over. Looking at pic#5 in my first post you can see a typical template for boxjoints. I start a cut at the green pin=green line on template, then widen my cut a few times until I reach the read pin/line on template, then freewheeling until next green line and start cutting again. For boxjoints the distance between my coloured pins is boxjoint pin size minus kerf, whereas for tenons it was size + kerf.

Single tenons and boxjoints on every 2mm I run with 2 pins. Custom double tenons and boxjoints on every 1mm I run with 4 pins. Also done boxjoints on 1/2mm and that requires 8 coloured pins, went surprisingly well and wasn't much more complicated than running on 2 pins.

The digital readout I don't use anymore, the pin system is way faster and more accurate.

That was the short explanation but there is more to it...:)

490156⁷

[Dan Cameron]

Ola
Two items:
1. If you google pyrouterjig you will find a free program for printing templates for box joints which you may find useful.
2. I would very much appreciate it if you would comment on my previous reply to your thread.

[Ola Carmonius]

Hi Dan,

1. Ok thanks. I know how to do it myself though in the computer but in this case pens are quicker. It's a one minute job to make the template when I dry run my first row of boxjoints of a new type.

2. Sorry I didn't know what to comment. It looks kind of unique but I think you should create your own thread to describe how it works, otherwise this thread will very soon be very confusing to read.

[John Erickson]

Bye far the nicest tenon jig I've seen, thanks for sharing.

[Ola Carmonius]

Thanks John, glad you like it:)

[Mel Fulks]

Ola that is quite something, and you were wise to not build it until Cotton Mather was dead.

[Ola Carmonius]

Thanks!...but I must admit the joke (?) went straight over my head.

[Jay Houghton]

So how does one go about making one of these for oneself...only one...
Jay

[John Lifer]

Agree with Jay, Plans? CNC is great if you are doing a few hundred pieces. But for only a box or two, this would be pretty nice jig. And I'm guessing you could do odd widths pretty easy using the pins.

[Peter Mich]

I believe Ola includes a link to the plans in the build videos he posted…https://youtu.be/kQHUJFsTeuk

[Ola Carmonius]

You can do quite custom things once you are familiar with the pin system.

First pic shows control wheel set up for half mm divided boxjoints that was shown on page 1, final result here in pic 2. Holes are for a carrying handle.To run even mm number boxjoints it's enough with 2 coloured pins, odd mm requires 4 pins and here I was down on half mm which needed 8 pins. Not much harder than 2 or 4 pins actually.

The third pic shows a table leg attachment where the width of the leg was divided equally into 5 segments which resulted in very uneven dimensions, like 11.7mm each. As long as it's only a few fingers as in this pic you can do almost any dimension. Haven't tried unequally spaced boxjoints but that should be doable if the brain can keep track of it.

491995

491997

491996

[Jay Houghton]

I purchased the plans and have ordered a ball screw and linear bearings. I'm using some Imperial screws like the cross dowels because they are easier to acquire locally. But the screws that hold the linear rail bases and ball screw are tapped already so metric is required. I also scrounged around for 15mm BB but didn't have any scrap and can't find it from my local dealer. But I do have some fabulous 18mm BB with walnut veneer! That will require some math to make sure everything fits correctly.

One of the reasons for my interest in this was to do more complex box joints like graduated fingers that get thicker and thicker across the corner. I am going to have a friend with a CNC do the wheel for me however since the accuracy of that determines the accuracy of the finger width. I've studied the plans and watched the YouTube video's and it's by no means beyond the skill of any competent woodworker.

I'll report along the way in case anyone is interested.

Jay

[Prashun Patel]

Wow, impressive workmanship.

Incra, eat your heart out ;)

[Ola Carmonius]

@Jay it should work fine with 18mm, as you say it's only a matter of modyfying a few dimensions. The control wheel is a fun part to make, if you have a drill press table and circle jig for the router. Quite time consuming though but also manually made will be really accurate following the instructions. 1mm off (which is really bad drilling) on one of the pin holes corresponds to about 0.01mm.

Will be very interesting to follow your build later on.

[Wes Grass]

I got the plans too, and have been modeling it in SW. Can't open the individual STEP files within the assy. Maybe a 'me' problem. My thoughts are to adapt it to a slider, and probably reverse it to be 'right handed'. I've done the same with a couple of MW's plans. I added linear rails and a pivoting angle plate to his tenon jig. Most likely another project that will float in the vacancy of my brain, but it keeps me out of the bars ... which is fine with me because I've got better quality hooch here at home, and with the price of gas ... ;

Only thing I've run into is some missing dimensions here and there. Some are to be located with the components, which is fine. The rest I've pulled by measuring off the STEP assy. It'll all get tweaked anyway. I've also thought about how to put a 'zero clearance' insert under the carriage, which could be swapped out for dovetails, etc. Maybe tack it onto the dust chute? Although by nature this isn't a cross cut, so maybe not an issue to begin with?

[Ola Carmonius]

I've also thought about how to put a 'zero clearance' insert under the carriage, which could be swapped out for dovetails, etc. Maybe tack it onto the dust chute? Although by nature this isn't a cross cut, so maybe not an issue to begin with?

Thanks for info about missing dimensions, I will investigate that.

I had thoughts about a zero clearance insert, especially if this jig is also used on the router table. But, it's all rip cuts and there is no tear out on the end grain side of such a cut. What would motivate an insert is if you use a really wide blade, say a dado blade, and then want to be able to cut small pieces also, like tenons on a narrow workpiece. In that case the groove a wide saw blade have made through the jig might be too wide for small material to have something to land on vertically. Primarily this is a jig for narrower blades but if cutting dovetails or using really wide blades or on the router table, insert is a good idea. Also if for some reason this jig is used for cross cuts it would motivate an insert but I have yet still never used my jig for such a cut.

[Wes Grass]

Be prepared for a bit of head scratching, if you're in the US. The drawings are 'European', 1st angle projection. Had me a bit baffled a few times. Back and forth between the model and drawings to sort it out in my head. Fortunately I have 2 monitors.

[Jay Houghton]

Well, I went and bought a 500mm stainless rule thinking I was going to throw in the towel and measure everything in metric and skip the translation. I semi-grew up with metric but it's till a foreign language to me. If you ask me how long 3 5/8" is I'll get pretty close but ask my how long 175mm is and I'll have to think on that a bit. That said I forwarded the files to a friend that successfully opened up the STEP files and is going to produce a set of templates on his CNC out of 1/8" BB. I should be able to then lay down the templates for dimensioning the cuts and then use a center marker for drilling the holes. He's an Imperial guy so all the holes will be marked with 1/4" holes. Once I center mark them I can drill with the right diameter bit. Once you get your head around the basic design it's quite easy to imagine using longer linear rails, taller capacity (I have a 14" table saw), and left or right miter slot.
Jay

[Ola Carmonius]

@Jay I beleive you could also print the 2d drawings on A2 paper and use directly as template. Most drawings are done on A3 sheet in scale 1:2. I had ideas to make the drawings in scale 1:1 A2 format but not many have possibility to print that big.

Must be tough to have a different unit system than the rest of the world;) You will need to master the mm system anyway when you start using the jig.

[Wes Grass]

If you really want it in inches, ballscrews are available in 4 or 5 TPI. Not likely to find them cheap though. I checked McMaster, the screw isn't *too bad*, but the nut and supports ....

Another option is to drive the metric screw with a 47/37 tooth gear set. 4mm = .20", 5mm = .250". The ratio is accurate to the 5th decimal, significantly better than a rolled screw itself. It'll be 'backwards', and you'll have some bit of backlash in the gears, but there may a little bit in the screw/nut anyway. Just keep going in one direction til you're done.

47/37 is a 'cheap' way of cutting metric threads on a lathe with an inch leadscrew. Exact requires a 127/100 combo. Not easy to mount up on a typical lathe, and harder to source.

[Jay Houghton]

If you really want it in inches, ballscrews are available in 4 or 5 TPI. Not likely to find them cheap though. I checked McMaster, the screw isn't *too bad*, but the nut and supports ....

Another option is to drive the metric screw with a 47/37 tooth gear set. 4mm = .20", 5mm = .250". The ratio is accurate to the 5th decimal, significantly better than a rolled screw itself. It'll be 'backwards', and you'll have some bit of backlash in the gears, but there may a little bit in the screw/nut anyway. Just keep going in one direction til you're done.

47/37 is a 'cheap' way of cutting metric threads on a lathe with an inch leadscrew. Exact requires a 127/100 combo. Not easy to mount up on a typical lathe, and harder to source.

My feeling about joinery in woodworking is that it's not metal working. You can't get machining down to the same levels as metal because the next day that fresh surface you just machined will be bigger or smaller depending on the humidity in the shop and the moisture content of the wood. Also, there has to be room for glue, which the vast majority of is water based. As soon as you start the glue up you're on the clock, not just from the glue drying out but of it expanding the joint to a point where it no longer fits together. A piston fit is a thing of beauty but not practical for woodworking. (IMHO of course).
Jay

[Jay Houghton]

Good morning Ola...working my way through the build. Mainly assembling all the parts I need so I haven't started cutting anything yet!


I saw that your "go to" blade is a 6mm FTG rip blade, and since it's orange I presume it's a CMT brand. That sounds like a great idea. I cannot find that blade in the catalog however, do you have a model number for it? What they do sell is a two-blade set where you reverse them to get a 3/8" width. So you're buying two blades essentially. They really are made for the Incra box joint jig you can buy and people that make jewelry boxes like the small kerf.

Anyway, I'll post progress photos as I have them.
Jay

[Ola Carmonius]

Well, that CMT blade is not my go to blade anymore. Thought it cut a bit sluggish and I didn't really like the cutting feeling. Checked it in detail and the tops were not really flat, they were all leaning in one direction and I guessed that effected the feel. That is manufacturer error and I will have it regrinded and give it a new try. Waiting for that to happen, I have switched to a 4mm flat top rip cut blade. Cuts much better than the CMT one. The new blade is from a local Swedish blade manufacturer so won't be of much help. Photos attached.

492266

492267

[Wes Grass]

My feeling about joinery in woodworking is that it's not metal working. You can't get machining down to the same levels as metal because the next day that fresh surface you just machined will be bigger or smaller depending on the humidity in the shop and the moisture content of the wood. Also, there has to be room for glue, which the vast majority of is water based. As soon as you start the glue up you're on the clock, not just from the glue drying out but of it expanding the joint to a point where it no longer fits together. A piston fit is a thing of beauty but not practical for woodworking. (IMHO of course).
Jay

I totally agree. I've made that mistake a number of times. I made wooden screws for a Moxon, threadmilled them on the lathe with an endmill in a toolpost grinder. They fit great, for awhile ... one will probably require a pipe wrench to get out.

As for inch vs metric, there's always going to be some level of calculating anyway when you consider the blade thickness and glue allowance.

What I like about a screw driven jig is that it's 'absolute', rather than a series of incremental movements off a stop pin where error adds.

[Ola Carmonius]

I agree with you both and I think I mention that in one of the usage videos. The important thing is not if I cut a tenon to exactly 12.00mm, its rather that if I aim for 12.00 thig jig will be consistent and keep it within a few 0.01mm from the target, on all tenons. Second and maybe most important thing is the pin system enables predictable trims in steps of 0.05mm. The actual dimensions of wood is really not that important, it's the fit between the parts that matters, and it's really easy to fine tune that with this jig.

[Jay Houghton]

I'm making progress on my jig build. Only a few adjustments for using 18mm plywood. I've learned a few tricks along the way, especially about getting cross dowels to fit right, which is, don't think you can drill that kind of precision on a drill press. You need a CNC! However I used a slightly larger diameter drill for the screw holes and that allowed enough wiggle room to catch the threads of the cross dowel. All the rest of the dimensions were spot on so just build to the specs and the overall design will come together. I'm in the final stages and decided not to install a DRO, and I'm having a friend machine the three components of the hand wheel on his CNC, including engraving the markings and drilling the 80 holes! Huge savings of time and effort. I'll report back once I get them from him. I'm also going to use T-track on the carriage instead of routing a dovetail way, or T-track channel since I have the T-track already.

My cost on this is in the $150 range partly because on some of the components like the flange coupling, cap screws and the like you have to buy way more than you need in order to get the price down. I also paid FedEx/Kinko's to print about 15 drawings at $2.25 per page so I could see them closer to actual size.

All in all I'm excited about the jig. The precision of 4mm/80 resolution should be a nice step up from the usual way of making finger joints. Not counting the labor, to get this for less than the cost of an I-Box jig is not something to discount.

Jay492695492696

[Ola Carmonius]

I'm making progress on my jig build. Only a few adjustments for using 18mm plywood. I've learned a few tricks along the way, especially about getting cross dowels to fit right, which is, don't think you can drill that kind of precision on a drill press. You need a CNC! However I used a slightly larger diameter drill for the screw holes and that allowed enough wiggle room to catch the threads of the cross dowel. All the rest of the dimensions were spot on so just build to the specs and the overall design will come together.

No CNC needed, the solution is just as you did to drill the screw holes in both pieces slightly bigger than nominal. All cross dowel screw holes are oversized in the plans, but if you swapped metric for imperial screws that might not be so obvious.

[Jay Houghton]

I'm almost done, I was going to skip the DRO but decided it would be useful so I ordered one. I did encounter an unanticipated problem, which is that I though it would be no problem to use a Freud box joint jig (1/4 or 3/8") on a metric device. Well, it isn't! There's just way too much math involved in getting a 3/8" finger to fit in the other side. So my solution, since you can't buy a metric box joint blade, was to put a 0.0190" shim from my dado set between the two blades, and now I have a 10.007mm kerf on the blade. My fist sample was too tight but with the resolution being .05mm you can simply move the carriage a couple holes to adjust the fit without creating any cumulative error since your reference for the next cut is always the Zero point on the part, not the previous cut. The other advantage is that you can cut multiple parts at the same time, like (4)-1/2" pieces at once.
Jay

[Ola Carmonius]

Hi Jay,

As mentioned, if the blade is imperial or metric thickness is of no importance. Boxjoints are cut in metric numbers (very important) and control wheel indexes metric no matter if the blade is metric or imperial. 3/8" blade equals 9.525mm and can be used for 10mm boxjoints and bigger with just as good results as using a 10.00 or 10.07 blade. None of the blades though can be used for cutting imperial boxjoints on this jig. So, think metric and cut metric boxjoints and use any blade you want, imperial or metric.

Example: 12mm boxjoints. With 9.525 (3/8") blade, measure the sawed kerf, let's say it's 9.55. Put the green pin in the zero, red pin in 2.45 (12-9.55).

With 10.07 blade, measured kerf is maybe 10.1. Green pin in the zero, red pin in 1.9 (12-10.1).

Kerf effects red pin position but nothing else. When cutting mating piece, move carriage one finger thickness, 12mm. In this case that's three whole turns on the control wheel so no reason to move the pins. Move the notes to equal the new zero position, cut the mating piece. Also this operation is totally independent of the saw blade thickness.

[Wes Grass]

494406

Eyeball alignment should be good enough, but you still need a decent chuck.

[Ola Carmonius]

Jays talk about varying finger widths inspired/challanged me to try out some patterns and varying finger widths the latest days. If you are into these kind of joints with variations this jig will give so many options, these I show here are just a few examples. Once familiar with the system irregular patterns or not complicated at all. All joints done with my 3.2mm flat top blade. None of these joints are glued or sanded, this is the dry fit accuracy acchieved with this jig.

First picture finger widths only 4mm on the "non visible side" of for instance a drawer. Fingers on visible/front goes in a decreasing-increasing series 20/16/12/8/12/16/20.
495310

Half of that series looks like this, maybe a bit more clean look.
495322

Next pic both fingers goes in a series of 16/14/12/10/8/6/4 and in the opposite direction to each other. From the top mahogny goes to thicker fingers, ash to thinner. They meet halfway. Takes a while to see it maybe.
495311

Final pics just to show how versatile this is also for larger joints like this bridle joint in 59x59 dimension wood. With only 5 fingers no evenly divides numbers are needed, with 6 colour pins finger widths can be whatever they need to. In this case my piece of wood was 59mm and the fingers ended up at 59/5=11.8mm. Control wheel set up for this joint also shown. Here the fingers are planed flush, would be to good looking otherwise. Not glued though.
495312
495313