[JerryBurks]

I have been playing with different joinery techniques since I built my CNC machine. I was intrigued by Gerry's thread but unfortunately my machine does not allow for vertical clamping and the use of fixed size special dovetail bits limits the usability somewhat. Louie's older thread of flat machined dovetails with a v-bit was more feasible for me but the use of 2 different bits and flipping the boards precisely is an issue. Drawing the exact geometry in VCarvepro isn't that trivial, as well.

So I dug out some 30-year old geometry skills (or what was left of it) and wrote a Basic program to machine tapered dovetails with only a 30-degree engraving bit from one side. Works pretty good and I did some preliminary practical tests.

- con(s):
>The method is not entirely flat but close. The board to be machined must be clamped at a shallow 15-degree angle to the CNC table (assuming a 30-degreee engraving bit). However, such a wedge jig is easy enough to make and I believe most CNC machines that are not set up for vertical clamping get close enough to the table edge to machine a dovetail even to very long boards.
> the pin and tail taper is relatively steep. As a result, the joint looks a bit different than classical hand cut dovetails.

- pro(s):
> Only one machining process needed per board end
> Only one bit needed (a 30-degree engraving bit, e.g. $10 for a 1/4" 2-flute carbide bit from Drillman at eBay)
> any desirable board thickness is possible (I guess 1/8" is a practical lower limit due to setup accuracy)
> tail and pin side may have different thickness
> no bit diameter related gaps or crevices. The joint is void free (assuming accurate clamping)
> the generated g-code will cut the dovetails and pins and square the board end as well
> variable pin/tail distance as well as dovetail-free joint ends
> no CAD or CAM software needed, g-code is generated directly from the machine and material parameters
> selectable overcut (to make sure to cut through), roughing/finishing skin and glue gap thickness.
> board length is not limited by table height (like vertical clamping)

So far so good, first results see below in the pictures. Now my question is what to do with that. In the first place I developed this for my own purpose but I am wondering if there may be a market for such software. I never developed software for sale but maybe you guys have an idea what to do with it. Any suggestions are welcome.

V-Bit in start position:


Joint in 3/4" plywood:


G-code visualization:


more pics in the next post.

[JerryBurks]

More pics.......

Making a cube from 10mm (~3/8") plywood:


Pretty good joints, minimal gaps, no voids:


Everything fits together well:

[gmfoster]

Jerry

That is some impressive joinery....
Garry

[eigstein]

Well done .

sincerely ,

eigstein

[Devastator]

So far so good, first results see below in the pictures. Now my question is what to do with that. In the first place I developed this for my own purpose but I am wondering if there may be a market for such software. I never developed software for sale but maybe you guys have an idea what to do with it. Any suggestions are welcome.

Strong as hell speaker boxes comes to mind. I've made some small reference monitors that could have benefited from this approach. It looks so nice that a clear finish would be in order. I just got some of those bits from drillman in the mail today. Let me know if you need a beta tester.

Way to think outside the box, pun intended.

[pirate323i]

Love it!

[SixOfOne]

I kinda like this idea also, very neat. I'm in the process of designing my first machine and vertical clamping was something I really want to have but this 15deg clamping I assume give you the ability to dovetail boards longer then the 30"-40" boards that vertical clamping would allow.. I wish I could have just dropped 5k on a built machine I wanna cut stuff sooo bad but I'm a long way away from that..

[JerryBurks]

Right now the software is functional but very primitive. No user interface whatsoever, all machine and material parameters are hard coded and compiled. Perfectly easy to use for me but not a very practical way for a distributable software.

If I really wanted to sell this software I would need to create a GUI, write a manual (if a very short one) and I was just wondering if that is worth it. Maybe I should create a voting thread to gage interest?

In the meantime I did another experiment joining boards of different thickness (1.5" and 5/8") and I will post a few more pics tonight. It worked also well but I noticed if the boards don't have crispy edges to zero in with the bit or if the bit tip is broken off a fraction of a millimeter (which still may allow the bit to be used), it can not be properly zero'd. As a result the accuracy and fit really suffer. I am thinking of an improved clamping wedge jig with a bit-specific tool sensor surface.

[robedney]

This is exactly what I've been looking for -- and I think you have a market. I'd buy it today with a GUI. Really, I want it! I've been looking for a corner joint, and I -- like many -- don't want to clamp to the end of the machine or modify my bed. The 15 degree jig ought to be easy build and calibrate. I'd be more than willing to beta test, even paying for it. I could also help in trade if you want to set up a website to market it.

One word of advice -- keep the price reasonable. This is the sort of thing that lots of garage hobbyists are likely to buy if the price is right, so I'd look at making money on volume -- and I think you'll get that with this with a nice, clean GUI and an easy-to-buy-it-on site of some sort that's nicely done. Forum-talk and word-of-mouth are probably the only marketing you need.

So, yeah, you're onto something!

[JerryBurks]

...........So, yeah, you're onto something!

Thanks, I sent you a PM.

Here some more pics and a video clip of my newest experiment. This is for unequal joints in construction redwood 2x6 (actually 1.5x5.5) and 1x6 (actually 5/8 x 5.5)
It worked well but the remaining gaps are caused by inaccurate zeroing due to rounded board edges and a chipped bit tip. From start to finish this "project" took me about 90 minutes. About 90% of that was actual machine run time. O.K., after that was a bit of gluing and sanding so maybe 2 hours.

Dovetails2x6 - YouTube


These are pretty massive boards and don't allow for many pins/tails per joint

[SixOfOne]

Very cool!

Are the joints good and tight? looks like there is a little air in there but could be an optical illusion. Either way they look better then any of the ones I have done on an expensive manual jig. Now you can make giant dovetailed dice!

[JerryBurks]

....Are the joints good and tight?.......

Well, not really. There are pretty big gaps, maybe 20-30/1000, which are unacceptable for a real project. As mentioned, tight joints require very precise positioning of the bit tip onto the edge of the board. If the board edges are rounded (like this ConCom redwoood) and the bit tip is chipped off this is impossible and such gaps will result. For this reason I am thinking of a special tool sensor.

[ger21]

I didn't look that closely when you first posted this, but is one side of the joint just resting in the other side? So with no glue, it will just fall apart? Or does it somehow lock together?

[JerryBurks]

.....but is one side of the joint just resting in the other side?.......

That is correct for one load direction since v-carving both sides of the joint will always lead to tapered pins. That means you can pull the pin board out of the tail board but the tails do have an undercut and you can not easily pull the tail board out of the pin board. That is similar to a classic straight pin dovetail joint although the latter may give you some additional friction to hold them together "dry".

I my experiments so far this did not appear to be a big problem. Actually there is some advantage to the taper that the glue is not scraping off when plugging together, resulting in a better fill of the glue gap. If the glue ever comes loose I guess any dovetail joint will eventually rattle apart. Besides the aesthetic aspect I see it in the first place as a method of multiplying the glue surface and to convert a peel or pull load on the joint into a shear load which the glue can handle much better.

One thing I want to try is making a double dovetail like in this thread.

[ecurb5]

Jerry, what was the outcome of the software? Did you ever pursue the gui?

-Bruce

[JerryBurks]

Jerry, what was the outcome of the software? Did you ever pursue the gui? -Bruce

Oh, sorry that I did not update this thread. I believe I created another one a while ago.

Yes, I finished this software and set up a web site HERE

You can download an evaluation version there and buy a full license if it works for your needs. But even the free eval version can already do the usual drawer boxes and the like.