Official Luthiers Forum!

There is something I’m missing and I can’t figure out cause*; Truss Rod slot cut too wide (measurements below).

I’ve been trying to accurately route the truss rod channel (TRC) on some singe piece neck blanks and cannot get it right. I’m using my Craftsman Professional benchtop router table (haha, Bull puckie, oxymoron I know), with a new, sharp 2-flute .250”, 1/2” shank bit. Pics show the tools

Once I figured out how to cut dead center, now the slot is too wide for a trussrod. Measurements every 5” of slot: .264” wide, .265, .260. .265. My sample test cuts were constant .247” wide. Depth is consistent.

I’ve spent many hours squaring, straightening, flattening, measuring to no avail.
1) the work piece square straight and flat

2) the router bit runout (<.001)

3) The Fence square, straight and flat (.005 flat across the 24” fence)

4) the table square, straight and flat

5) the router plate square, straight and flat

6) Work piece, to fence, to table: square.

7) Hardware tightness

7) Cross checked that the squares are flat, and square (Starret scales and square, Groz Engineering Squares, 1-2-3 blocks

8) Fence deflection (.003” with a lot of beef on with work piece in position.

There is something I’m missing and I cant figure it out.

* I’m now left with wrong type of bit, wrong method, wrong technique. Is a 2-flute the wrong bit? THe router table seems the best way to do it, and I wrong? Your help and advise are appreciated. This is my first neck from a billet

_________________
Measure Twice,

Karl Borum

Are you using a quarter inch bit? How deep of a cut are you taking on each pass. 1/8” deep is the normal depth of cut per pass. You might want to test your cut on a 2x4 until you get it right.

As Barry says, if you’re trying to cut too much at once, things will chatter and move around. I have had router bits that came with significant runout. Figure out how to measure the actual runout at the cutting edge when the bit is installed in the router.

If you can’t find the issue, you can always work up to the 1/4” using multiple passes with a smaller bit.

I usually cut my slots using a medium size handheld router with a long fence attached to the base.

That’s an incredibly thick neck blank. Is it dead flat on the fence side? I use two featherboards and take about three passes to complete.

IMG_7401 by Terence Kennedy, on Flickr

IMG_7401 by Terence Kennedy, on Flickr

_________________
It's not what you don't know that hurts you, it's what you do know that's wrong.

The setup I use is pretty much the same as yours. I have what looks like the same Bosch router in a router table. I use a 1/4" two-flute bit that can cut 7/8" deep. I cut the channel depth of 0.35" in a single pass and it always works so it should be possible for your setup to do it too.

If the channel is cut consistently wider than 1/4" all along its length, it seems that there may be lateral movement in the bit or the whole router for some reason. I know you checked hardware tightness, but I would still be looking for some tiny amount of looseness or flex in the mounting of the router to the table, the plate it's attached to in the table, or whether the height adjustment mechanism lock is truly locked down to rule those out for certain.

What did you do the test cuts in that gave a 0.247" width?

_________________
Once in a while you get shown the light in the strangest of places if you look at it right - Robert Hunter

Thanks for the questions everyone.

Yes the bit is 1/4” bit.

The sample cuts were shallow in a test board and one test cut on the first pass of the neck blank. A professional would have validated the method before cutting good material (note to self).

I’m cutting a .425” deep channel in four passes: 3x .125 deep, with the last pass shallow, to depth.

That is a big billet I agree. I already cut one neck blank out of it. I left it that size for mass, until I cut the channel and tenon.

I’ve checked every piece of hardware (unless I missed something )

I’ve measure deflection of both the fence and router plate using the blank and 2X the force I would normally use. There was negligible flex movement (under .002” both fence and router plate). To get the router plate to move I had to lever the motor.

Seeming most likely now, I well could have left the router unlocked during a pass, I have to unlock and lock it back 4 times, a good case for a router lift (I’m too old to justify the cost of a lift). I should have measured the width after EVERY pass for this new-to-me process. I should rout on a squared up 2x4 and measure the result before cutting another channel.

It’s good to know that others are using a table with good results, there is hope.

_________________
Measure Twice,

Karl Borum

The reason I don't think it's due to things like deflection in the fence or things not being square is that those problems should give a routed channel that's 1/4" wide, just not straight or square. They shouldn't result in a channel that's wider than 1/4". Forgetting to lock the router down could result in lateral movement of the router while it's running which could give a wider channel.

_________________
Once in a while you get shown the light in the strangest of places if you look at it right - Robert Hunter

Yes, it’s dead flat on the fence side. I went through that arduous exercise on the first neck, as my jointer, and subsequently my billet, was not perfectly square. I’ve all but reached my limit of accuracy with my home-owner machines. I squared it with a hand plane (thank you Paul Sellers for teaching us how).
I will add a second feather board!

_________________
Measure Twice,

Karl Borum

While checking that hardware/ flex, I did make a small discovery BEFORE I started routing the second neck’s channel: my plate was not very level, and the screw hole edges had displaced material upwards, from compression of the screw into the plate material (phenolic/ PET plastic maybe). So the table was not aligned well (level) with the table and the screw edge material made it not flat either. I found that with a (dial indicator and stand). When I initially leveled the plate (and what I believed was best practice) I left the out feed edge of the router plate over flush “a little” so the work piece would glide off the router plate onto the table without catching the edge

I was able to sand those “sprews” down flat using a small hard sanding block. In the process, I used the surface gauge to level the plate at the four corners. I also removed the plate and all the hardware and lacquered the threads of the top leveling screws (that screw into the bottom inserts that are threaded to accept the top screws). You can’t loctite those attachment screws or they will unscrew the leveling inserts from the table when you remove them.

The whole reason I went through this exercise is that I could not, for a million dollars, properly nail the dead center line of the neck with the channel. I worked around that issue (finding the center of a crazy geometry router bit to measure to the fence) by using a 1/2” S.S. Centering pin and eyeballing it from the end of the neck. I found center within .005” eyeballing the centering pin.

Sorry for the droning long posts….I kinda geek fest on applying this stuff to guitar making, because every little error is amplified and magnified by the very building process (thinking how a board .003” out of square at the bottom of the fence gets magnified over a 4” height on a square fence).

_________________
Measure Twice,

Karl Borum

It's a Crapsman, what do you expect?

If your neck blank is a little wider than the widest part of the fretboard and headstock which it should be, nailing the absolute perfect center is not that important. You have a little wiggle room. The truss rod slot will define the center for all subsequent operations. I try to get it as close as I can though.

I use the Blanchard double action rod and previously the LMI design. If the slot is a hair loose it’s easy to shim the ends where they need to be snug.

_________________
It's not what you don't know that hurts you, it's what you do know that's wrong.

Did I read of hand planes, if so I suggest googling for this video which you can find fairly easily,
here I've posted the exact title foru
David Charlesworth. hand tool techniques part 2: hand planing.
That older video is still the very best hand planing video ever made.

Some other things worth mentioning might be the collet, should you remove it and test fit the cutter
a recent enough thread on a UK forum I can't find, had a very similar issue.

Not being very experienced with routers, but I don't see hold downs,
though worth mentioning for a milling machine, the usual thing to do when wanting a better tolerance/cleaner wall
is to use a smaller bit.
Not sure if that applies here.

All the best
Tom

I found a centering bit on Amazon:

https://www.amazon.com/gp/product/B003N ... UTF8&psc=1

It made it a lot easier to set the router fence.

Another possibility for setting the router table such that the bit is on the center of the neck blank:

How about a 30 degree V carving bit, the kind that has a flat side and a round side? You could set the bit low enough in the table where the flat side can bump flush up against the end of the neck blank. Set the fence such that the pointy end of the V bit sits right on the neck blank center line, and Bob’s your uncle.

The problem that I see with centering bits is that they are completely round, so the pointy end is still a small distance from the end of the neck blank until you actually start a cut. I think this introduces error. Maybe the flat face of a V carving bit can help with that part of the problem. Once you get it dead center, replace the V bit with the bit you want to use for cutting the slot.

Another way of solving the problem is to go ahead and use a round centering bit, but set it so low in the table that it barely sticks up, and only the last 1/16” of the tip bumps up against the center line of the neck blank. But I like the flat sided V bit idea better, if you have access to one.

Good luck with this in the future!

For thru cuts like yours, I'd use my dado stack in the table saw.

_________________
Hutch

Get the heck off the couch and go build a guitar!!!!

This sounds like a tolerance stack issue, with some slight irregularities in fence, router collet runout, guided on one side only, etc. When I was doing conversion of Artcore archtops to shorter-scale 8 string necks at Greenridge, the design called for an accurately routed 1/4" channel which would align with the hidden truss rod access channel in the underside of the neck extension (Martin Size 5 adjustable truss rod in a from scratch custom maple neck/ebony fretboard, etc).

Although I had cut through-channel truss rod slots with the table saw (worked fine on a cabinet saw with accurate fence) as well as router table (a nice phenolic-top/adjustable fence Jessem, with centering jig, Milwaukee 1-3/4 hp, etc.), we ended up doing a Baltic Birch jig which used a good grade of double-stick tape to stay on the neck blank. We cut the slot with a 1-1/4hp DeWalt trim router with an aftermarket base and 3/8" OD bushing. Provided the jig is closed at both ends and the Fastex double-stick tape extends the length of the jig footprint on the neck, this is rock solid. For accurate alignment of jig with centerline of the neck, a thin 3/8" wide, .095" thick Lexan insert with centerline scribed on underside and highlighted with Sharpie ink makes matching up with the pencil or (better) marking knife line easy.

_________________
A constellation only takes shape when one maps the whole.
- Beth Brower

Something is loose. Have you checked that the router is securely connected to the table insert?

Just a couple thoughts/questions.

I'm pretty confident you didn't flip the blank end for end between passes even though that is one way to "center" a cut. But if that is done with a full slot dimension bit, it will magnify variations in slot width caused by irregularities in the fence, etc.

Making the slot with multiple passes can cause additive variations with each pass just based on how firmly the blank is held against the fence/table while pushing or repositioning hands. Do you see any light "steps" on the sides in the slot?

I'm also curious about your bit. For precision I like solid carbide without brazed on carbides. Don't know how yours was made. But you might try a solid carbide 1/4 shank 1/4 cut bit (common at big box stores) and avoid the cheap ones. 1/2 inch shank is beneficial with large bits but probably provides little if any for bits 1/4 dia or less.

You might also try making the first cuts to depth with a 3/16 dia bit and make one final cut to depth and width with 1/4.

Good luck

_________________
Brian R, Wood Mechanic
N8ZED

This! It is how I do saddle slots and you can be very precise. If you are taking a sizable cut, the bit can pull to the side and widen the slot. A skimming cut will be less prone to that.

See if using a feather board on both the infeed and outfeed sides helps. .015" isn't a lot, and it might be introduced by the inherent movement of feeding the piece into the bit. I like my Bow Products feather boards FWIW.

I confess I’ve only skimmed this thread, but a couple of things caught my eye:
1. straight bit, rather than spiral — much more prone to generating vibration, especially if used for a deep cut, causing a wider-than-intended cut
2. Regardless of type of bit, it’s best to first hog out most of the material with a smaller diameter bit, near final depth, then clean up with a single pass at final width and depth.


Sent from my iPad using Tapatalk

I have cut truss rod channels with multiple passes on the table saw with a regular blade, 1 pass with a dado setup, and on a router table. Then, I built a jig for my Dewalt laminate trimmer. I would never go back to any of those other options.

It is a frame made of 1"x 2" boards, that the router rides in, with 1/4" Plexiglass floor. The Plexiglass has a slot routed down the center for the bit. Neck is held in place, under the Plexiglass, with clamps or double stick tape. Works like a charm.

If you stay with the router table method, hogging out most of the slot with a smaller bit, which several folks recommended, is the best way to go. The final pass just does a light trim and will produce a much more accurate channel.

It's been mentioned a couple times that the best approach is to remove most of the wood with a smaller bit and then finish with a bit that has the target size. Maybe this is true as a rule of thumb for routing in general, but my experience has been that it's not necessary for a truss rod channel. I use one bit and one pass to cut a 1/4" wide channel that is 0.35" deep and I'm done. Fast and efficient. Works every time, at least for me.

_________________
Once in a while you get shown the light in the strangest of places if you look at it right - Robert Hunter

Stack up is also a reasonable explanation I didn't wholly consider; optimism bias me thinks.... thank you.
Maybe I will plug the channel (yet again) and go for another run to be DD sure I've clamped the router.

I've experienced some violent events (twice) using dado blades which makes me hesitant to use one for a "long" piece, where I would have to hold down the work past the blade. The panel that got away (on two separate occasions) was ripped up pretty bad, as is was drawn across the blade. I underestimated the amount of weight required, behind the dado blade, to keep the panel from vibrating during the blind-channel cut. The panel actually dislodged the downward forcing feather board off the fence.

_________________
Measure Twice,

Karl Borum

My truss rod channels, cut with this method, are minimally wider than the truss rods. Once placed, there is a very small amount of motion.

Should some type of shimming or fitting be done, or are the rods held in place, without hopefully any rattling, once fretboard is installed and rod is gently tightened?

Thx. Jon