Leigh D4 dovetail jig driving me nuts

[Mark & Juanita]

OK, just posted a better sketch of what I'm trying to illustrate on abpww.

[J. Clarke]

Eyeballing? One can measure you know.

[J. Clarke]

Only if the user screws up, like its competitor.

[DJ Delorie]

Assuming we're talking about the Incra, which is the only dovetail jig I know about that requires eyeballing a center...

By design, you can't make solid half-blind dovetails with the Incra. There will always be more wood removed than needed, because the jig just can't make exactly the right cuts, so there are always small air pockets hidden within the joint (even if there are no visible gaps).

The Incra can make nice round-bottomed pin sockets for half-blind dovetails. However, to make the tails, it uses two cuts perpendicular to each other - resulting in a square "peg". The socket is deeper than the peg, so the pieces fit together, but there's an air pocket where the square peg and round socket meet.

In the case of through-dovetails, the Incra can't make the spaces between the pins properly. It actually cuts *tails* first, then removes most of the extra wood using a dovetail bit with the wood flat to the table. Even then you have to hand-carve some remaining wood. But the initial tail cut leaves an air gap when you assemble the pieces.

[charlieb]

snip

Sure, you can measure, find and mark the center of the board - that's not the problem. It's the aligning that center of of the board mark with the center of the router bit that's the problem. Few, if any, dovetail router bits have a convenient center mark to align to.

If you looked at the photo of the sample DT with the gap problem. you'd notice that either the pins part or the tails part is off by a noticeable amount so that the pins part's top edge sits proud of the top of the tails part - the two parts being of identical width. THAT indicates the pins or tails or both are not centered on the parts width.

My Oh Bullshit alarm goes off whenever I see ANY machine dovetail making tool claim "fool proof" and/or "no test cuts necessary". It's getting the damn bit heights right that raises all the hell - and takes so much time to get right. And when you have to use TWO bits to make the DT joint - you'e going to lose one set up to get the other - unless you're using TWO routers - which you can do with a JIG but not with a router table and fence system.

[J. Clarke]

In that situation the Jointech has the same limitation. It's basically just an Incra with some tweaks.

As for making solid half blind dovetails, you can over cut or you can take a chisel and round the pins, depending on how picky you want to be.

[J. Clarke]

Which is why the Incra has a centering scale.

[Morris Dovey]

Hmm - then I think you may have an obsolete (pre-CNC) trigger circuit on your alarm. :)

Nah - I just /tell/ my router how deep I want it to cut.

Any more, I just use one small straight (spiral) bit...

...and I /tell/ my router how tight to make the joint. :->

[DJ Delorie]

The offset you care about is the center of the bit relative to the center of the bushing. The offset from the finger to the bit is irrelevent here - the jig is designed to accomodate *that* offset.

When pondering your diagrams, think to yourself "the bit is to the left of where it should be".

[Mark & Juanita]

OK, I got this -- I figured out what I was missing. Although my sketch is correct as far as the comparison in offsets between pins and tails, that's only 1/2 the story, when the router comes around to the other side of the pin or tail, the other side of the bushing contacts the jig, thus, just as you said the pin or tail is offset to the left or right by the amount of the difference you stated, from center of bushing to center of bit (the other difference to the bushing edges cancels out). As you stated, if I rotate the router 180 degrees, I'm shifting the fit such that the pin and tail board will be offset. Thinking about this, all other things being equal, rotating the router 180 degrees should be a way of measuring how far offset the router bit is from being centered in the bushing. It should magnify the offset by 2 if I'm looking at this correctly. That might be useful for measuring that offset in a straightforward way -- cut the pins and tails with the router rotated

180 degrees and measure the test piece misalignment. Doesn't help any as far as being able to make the adjustment to fix it any easier since one is essentially rotating a base anchored by 4 fasteners.

Thanks for setting me straight -- what can I say, an EE doing ME stuff.

[Mark & Juanita]

Thanks for all the responses thus far. Based upon all of the input, I've determined the following is my best path forward:

1. Ditch the Colt and go back to the Bosch 1613. Having used the Colt for test cuts was leading me to that conclusion anyway, it just isn't built for that kind of extended hard use. My purpose for using it was because I did have a way to fix the router base offset on it while not having a similar capability for the 1613. The small router also has a tendency to chatter and depth adjustment has been less than simple. However, I'm seeing the same kinds of gaps with the Colt as I was seeing with the 1613 when I used it on previous projects, so this just returns me to my starting point.

1. Because of the centering issue and because I would prefer to use a PC style screw-in bushing: order one of Pat Warner's clear router bases for the 1613 and probably the centering kit to aid in getting the alignment correct. This will alleviate the centering problem and might address some of the gapping by making sure that the base is uniformly flat -- I was seeing gaps before, even when presenting the same side of the router to the template (I didn't try the 180 degree rotation until this time) so flatness may have been an issue with the 1613.

2. Reset all of the finger guides, starting from the left and using a machinist's square to ensure the finger guides are square to the guide bar.

1. Verify that the guide bar is uniformly distanced from the work piece along the entire work piece width.

2. Pay particular attention to technique to ensure that I'm not rocking the router or tipping the router.

1. Take a very serious look at the Akeda jig. After looking at it on-line, I see some really significant benefits to this vs. the Leigh, particularly from the router support standpoint. The only downside I see is having to change out fingers each time one changes from pins to tails. I'm good with the 1/8" discrete steps and can live without infinitely variable spacing if the discrete setup guarantees that the fingers will be positioned reliably. The only uncertainty then is that the fingers are precisely machined for repeatability.

[Martin H. Eastburn]

[Mark & Juanita]

Thanks. I'll do that in the future

[Leon]

One more thing to strongly consider.

BIT CHATTER. If you are using 1/4" shank bits or a light weight router the bit could be spinning outside of it normal area of travel. In other words the bit could be bending while spinning and cause too much material to be removed. This could be exaggerated at higher bit speeds. Leigh recommends

8mm and 1/2" bits to help fight chatter.

I typically don't have a problem but I do use the larger shank bits.

[Leon]

Chatter along with a smaller based and light weight router could do that.

[Upscale]

Hmmm. You could be right, but that's an awful lot of bit chatter for the amount of dead space shown in the pictures he presented. I'd freak out if any router I had caused that much wobble.

[Larry Blanchard]

If by that you mean cutting first one way and then again with the 180 rotation, you're right.

Some routers have the ability to center the base plate - does yours? Of course that still doesn't help if the bushing itself is not perfectly centered or if the router body changes its position when raised or lowered.

Given all the variables, one solution is to adjust everthing as best you can, cut the tails a little tight, and pare with a chisel to get the "perfect" fit.

[samson]

That sort of defeats the purpose of a $300.00 dovetail jig.

s

[CW]

[Tom Watson]

I've used what may be an older version of this jig for quite a few years, in its 24" version and have had no problems with it.

Will admit that I set up the left side for my stock height short drawers and the right side for the taller ones. Used the middling version of Loctite on the screws.

Used a full width backer and fillers, so that the clamping comes out even.

Router set up is a 3 hp Porter Cable and the cutters are CMT's.

Have not had fit problems but that may be because I mic my pieces coming out of the planer.

Remember having some blowout problems when I started with the jig but the waste board cured that.

Have had good results on multiple runs using this method but stock prep for flatness and thickness has been critical.

Only problems that I've had were from not setting the router dead flat throughout the cut.

Regards,

Tom Watson