Routing rail ends

[cbrown]

I am working on a set of Shaker-style cabinet doors in 3/4" maple using a Freud adjustable tongue-and-groove router bit set in a table-mounted Bosch 1617 variable speed router. My first setup for doing the rails was to screw a backer board to my miter gauge, square the fence to the miter gauge and align it with the bearing on the bit, and then hold the rails against the backer and slide the miter gauge by the bit, trying to move the rail about 1/16 closer to the fence on each pass. (with the backer board screwed to the miter gauge, the fence itself had to stay put aligned with the bearing.)

I eventually got about 12 of the rails done cleanly (without ruining any of the pieces), but had some problems with kickback (say on 1/3 of the rails) jamming the rail back against the backer board hard enough to wrench the miter gauge from 90° to 45°. I want to try to solve the problem before doing the remaining shortest rails (5-6") since I don't want to be holding them that close to the bit.

I have the router speed at 16,500, which seems better (so far as kickback) than attempts at somewhat slower speeds. Depending on how one reads the speed recommendations (interpolating across the range of sizes and speeds or not), I could take the speed up to 18,000 (the bit diameter is 1 3/4"). Is this likely to help or create other problems?

The tongue-cutting bit has two wings on each cutter. The cutting tips on the top and bottom cutters are angled in opposite directions from the vertical (Freud bills this as "opposing shear angles of the tongue cutter") and the top and bottom cutters can be aligned either directly above one another, or 90° (or some other angle) offset. Is there some advantage to setting up the bit either aligned or offset? Is it possible to assemble the cutters so that the top and bottom angles are backwards? (Apart from the possiblilty of installing the cutting edge itself facing the wrong direction!)

My own analysis is that part of the problem arises from holding the rail down manually, which encourages applying counterpressure against the tendency of the bit to push the rail away from the fence and so winds up jamming the wood against the "front" of the bit.

I am thinking through the design of a sled with a holddown clamp and backer to carry the rails past the bit. The options (so far as I know) are a sled that rides on a bar in the miter slot or one that rides against the fence. the problem with the fence design is that the Freud adjustable set has extra-thick cutters so that it can cut a tongue in a 1" board and would completely eat up the leading edge of any sled when set to cut a centered tongue on 3/4" stock. Would this design still work if there is enough length of sled on the right side of the fence behind the backer board that never gets fed through? Or is a miter slot design (keeping the base of the sled 1/2" or so away from the fence) the one to work with? It does seem to me that if I did get kickback on a miter-slot-riding shopmade sled of 1/4" material mounted to a bar, the sled would probably be destroyed.

I would be grateful for wisdom and/or advice.

Christopher

[Jeffrey Schronce]

Are you saying you took 1/16" multiple passes?

[cbrown]

Are you saying you took 1/16" multiple passes?

That's right--6-8 passes to get to the full (7/16") cut depth.

[Jeffrey Schronce]

That's right--6-8 passes to get to the full (7/16") cut depth.

Given your set up and shallow cuts I have no idea why you are getting kick back like that. It sounds like you are doing everything right. Are you sure that the outfeed of the fence is parallel to the miter gauge track? If it was tilted in a bit that could cause some problem but doesn't seem probable with that shallow of a bite.

[Ken Massingale]

I am thinking through the design of a sled with a holddown clamp and backer to carry the rails past the bit. The options (so far as I know) are a sled that rides on a bar in the miter slot or one that rides against the fence. the problem with the fence design is that the Freud adjustable set has extra-thick cutters so that it can cut a tongue in a 1" board and would completely eat up the leading edge of any sled when set to cut a centered tongue on 3/4" stock. Would this design still work if there is enough length of sled on the right side of the fence behind the backer board that never gets fed through? Or is a miter slot design (keeping the base of the sled 1/2" or so away from the fence) the one to work with? It does seem to me that if I did get kickback on a miter-slot-riding shopmade sled of 1/4" material mounted to a bar, the sled would probably be destroyed.

I would be grateful for wisdom and/or advice.

Christopher

A coping sled will help, especially for the shorter pieces:


Build it yourself and add a bar on the bottom to ride in the miter slot.
I've had the problem you had and I believe it was due to not being able to hold the stock solidly by hand so that it twisted a bit, just takes a degree or 2, and caught between the bit and the fence.
ken

[JR]

A coping sled will help, especially for the shorter pieces:

The latest ShopNotes (Vol. 15, Issue 87) features a plan for a homebrew sled.

JR

[Holbren]

A coping sled will help, especially for the shorter pieces:

Definitely use a coping sled for those short piece end grain cuts.

I made a mistake of trying to do that with a monster lock miter bit freehand, it wasn't a pretty sight.

[Jeffrey Schronce]

Definitely use a coping sled for those short piece end grain cuts.

I made a mistake of trying to do that with a monster lock miter bit freehand, it wasn't a pretty sight.

Ya, there is no way one should do it freehand. However it still seems something is wrong in the set up if 33% of the passes are kicking back when he is using a miter gauge with length wise support and making 1/16" passes.
The sled is the best answer, but I still think there maybe something else going on there.

[LarryG]

Two points/questions:

1. The statement, " ... encourages applying counterpressure against the tendency of the bit to push the rail away from the fence ..." raises a red flag in my mind. Were you using the proper right-to-left feed direction? If so, the bit's rotation will serve to pull the workpiece into the fence, not push it away from it.

2. On what pass did the kickbacks occur? If they always occured on the final pass, with the end of the workpiece in contact with the fence, it's possible the workpiece was catching just slightly on the outfeed half of the fence, at the bit opening.

EDIT: A third question ... was the backer board on your miter gauge always in contact with the fence, on every pass, even though the workpiece itself wasn't? If it was, that was almost surely the problem, because that means the gauge was being guided by both the miter slot and the fence simultaneously ... which is not what you want.

[cbrown]

Sledding

Well, I made a sled out of 1/4" plywood, some 1x4" poplar, an Incra miter bar, a hold-down clamp, and some screws. It works beautifully, lets me advance the piece reliably into the cut from pass to pass, and carries through the cut without any detectable wayward forces. I finished up the rest of the rails in a few minutes.





I use the backer board held out by the miter bar as a handle. Since the sled is riding in the miter slot and I just have to push and pull (no pressure against the fence) it works fine.

I was worried about being able to apply enough clamping pressure without bending the plywood base, but it seems to work OK (I'm sure I'm not getting the rated 300 pounds of clamping pressure though). I don't think hardboard would be any better, and a thicker base would challenge the height adjustment on the router.

To square things up, I screwed the miter bar to the plywood and then (since the distances on either side weren't critical, as long as the sled came at least 1/2 short of the fence, I ran the base through the BT riding in the miter slots. I suppose it actually doesn't matter for this design whether the plywood is exactly parallel to the bar, but it's a lot easier to square the backer to the base than to the hidden bar. The backer board is mounted with a machine screw that goes through to threads in the miter bar and then squared up and set with a pair of countersunk 3/4" wood screws from the back.

Thanks for the advice and help.

Christopher

[cbrown]

Two points/questions:

1. The statement, " ... encourages applying counterpressure against the tendency of the bit to push the rail away from the fence ..." raises a red flag in my mind. Were you using the proper right-to-left feed direction? If so, the bit's rotation will serve to pull the workpiece into the fence, not push it away from it.

2. On what pass did the kickbacks occur? If they always occured on the final pass, with the end of the workpiece in contact with the fence, it's possible the workpiece was catching just slightly on the outfeed half of the fence, at the bit opening.

EDIT: A third question ... was the backer board on your miter gauge always in contact with the fence, on every pass, even though the workpiece itself wasn't? If it was, that was almost surely the problem, because that means the gauge was being guided by both the miter slot and the fence simultaneously ... which is not what you want.

Even though I have a solution to the problem in the sled, I would like to understand what was going wrong with the first setup--

1. The feed direction was correct. The main force of the bit, of course, is right and back. But when the cutting is taking place at the very front of the bit there seems to be a small centrifugal force so that the tendency was for a 1/16" cut at the left (leading) edge of the piece to become a 1/32" cut by the time the right edge was through the bit. I think that this--and trying to compensate for it by holding the piece more firmly--was the source of trouble because it resulted in deeper-than-intended cuts on the next pass at the right side as well as pressure being applied back against the bit in compensation.

2. Kickback usually occurred on one of the early passes--never with the end of the piece against the fence when it was certain that 1/16" or less was being removed.

3. The backer board was screwed to the miter gauge and came up to the fence (or just, say, 1/64 short) but rode along without catching the whole length of the fence once things were squared up. The backer board on the sled rides right along the fence, again, without catching. As long as there is no binding because things are out of parallel, I don't understand why this should be problematic.

Christopher

[LarryG]

Even though I have a solution to the problem in the sled, I would like to understand what was going wrong with the first setup--

As would I. Between your OP describing the problem, and the further details that came out as the thread developed, I can't think of anything that hasn't been covered. But something, somewhere, must've been fundamentally wrong. Almost had to be. The reason I say that is, any reasonably sharp router bit ought to be able to remove a paltry 1/16" of material in even the densest of woods almost effortlessly ... you should've been able to push the workpiece past the bit almost without even feeling the cut at all. I'm well and truly stumped, and it seems everyone else is, too.