Table Router Squaring

[Julie Moriarty]

This isn't making sense. Using two different bits, one bottom bearing and one top bearing, and routing along a pattern from one direction, the cuts from the two bits varied quite a bit.

Bottom bearing bit cut
http://i867.photobucket.com/albums/ab233/jules42651/Stratocaster%20Build/strat_shape_01_zpsaf2d9731.jpg

Top bearing bit cut
http://i867.photobucket.com/albums/ab233/jules42651/Stratocaster%20Build/strat_shape_02_zpsb1ee48db.jpg
You can see where I started the cut, near the "0" mark on the router lift.

The difference
http://i867.photobucket.com/albums/ab233/jules42651/Stratocaster%20Build/strat_shape_03_zpsfd95c245.jpg

The shaft measurements
http://i867.photobucket.com/albums/ab233/jules42651/Woodworking/Tools/Routersquaring_zps2bc97eab.jpg

Pretty much the entire cut was made from the "0" degree point on the table (the wood was on that side of the table.) I checked both bits with a caliper and the bearing on both was about .005" larger than the diameter of the cutter. As far as I can tell, the cutter on both is parallel with the bearing. I don't know how I could measure that accurately so I just laid a straight edge against it. I also checked the diameter of the both bits from top to bottom (on the cutter edges) and all came out the same diameter throughout the length.

I tried two other top bearing bits to see if I could trim that last bit off the edges but they all cut the same as the original top bearing bit. So all top bearing bits are cutting the same.

I'm lost. I can't figure out where the problem lies. Any thoughts?

[Bill Huber]

So is the router bit square to the table?

What is the difference in the cut, I am really not sure I can see it in the pictures.

I have a 1/2" drill rod that I put in the router to check mine to make sure it is square to the table.

[Larry Frank]

Bill is correct and you should use the 1/2" drill rod to check that it is square to the table.

You probably should try putting some pressure on the drill rod and see if it moves.

Let us know what you find out.

[Lee Schierer]

Based on your dial indicator readings I would say your router is tipped toward the 90 degree direction. This tilt would give you uneven cuts as you move the curved piece past the cutter. A taller bit would make the tilt more obvious.

[Ethan Melad]

ignoring the possibility that the bit is not square to the table, is it possible that your bits and bearings are not matched? the bearing on the top-bearing bit could conceivably be too big; maybe the bit has been sharpened a number of times?

edit: oops didn't see that you had already checked the bit/bearing sizes..

[Julie Moriarty]

I'm using a Betterly system for taking the measurements. The shaft in the collet came with the system so the measurements should be reliable.
http://i867.photobucket.com/albums/ab233/jules42651/Woodworking/Tools/router_set_zps87a95937.jpg
The zero setting was established at 0 degrees

I realize that between 90 & 270 it's off by about .050" but I rarely had the piece positioned on those sides. And even when I passed the piece through with the top bearing bit from all angles, I still couldn't nip off that last bit. If the bearing on the bottom bearing bit was smaller than the cutters, all this would make sense but the bearing is actually larger than the cutters.


What is the difference in the cut, I am really not sure I can see it in the pictures.
Bill, if you look at the picture of the guitar body on its side, you'll see on the part opposite the template a change in color in the sapele. That part is slightly raised from what the bottom bearing bit cut.

[glenn bradley]

I believe you want to correct for tilt in the mounting of the motor. Is the lift made for the motor in use or are there pads or adapters in use? Either way you obviously want the motor/arbor/collet dead on perpendicular in any position. A poor collet can cause you to endlessly adjust things in an attempt to align things so try a couple of collets and mount and re-mount the shaft; loosen the collet, move the shaft and re-tighten that is. Once you know you have repeatable results, adjust the motor tilt to achieve perpendicular.

277104
If you can't get there, something is out of spec or needs repair/replacement. Like aligning a tablesaw, this is a sometimes time consuming but, necessary adjustment. The taller the cut, the more the misalignment exhibits itself. Everything works as a team here; motor mounting, arbor / collet runout, plate to table co-planer, the bit and when in use, the fence.

[Stephen Cherry]

One option would be a layer or two of duct tape on the pattern- make the cut from one side, then the other. Finally, remove the duct tape and make a full length cut to the final dimension.

[Lee Schierer]

Another suggestion. Since your guitar body is likely less than 2" thick, use a 1/2" diameter four flute 2" to 3" long cutting edge end mill from your local machine shop supply store and do the cut using a guide bushing mounted in the table. You may need to make your template slightly smaller so you end up with the correct final dimensions for the guitar body. End mills work well as router bits and best of all come in longer overall lengths than router bits for about the same cost. McMaster Carr has one that is 5" overall with a 3" cutting length that is 1/2" dia for about $29.

[Ronald Blue]

As everyone has suggested in one manner or another is the router is not perfectly square to the table. Your method of checking is right on. Because you are out of square almost .045 you are going to experience the problems you have posted with a "guided" router bit. Because of the bearing location you will get different contact but essentially the same problem in the end. I think if you checked the cut with a square you will see the same thing. If there is no built in adjustment you will need to devise some shims to correct this. You are on the right path already. If you can move your indicator out one hole and still be on the lift it will give you even more accurate results. The larger the swing radius the larger the error will appear but this will lead to a much more accurate result when corrected. I hope this makes sense to you. Good luck and keep us posted.

[Art Mann]

The Betterly gadget looks pretty nice, I hadn't seen it before. Judging from the photo, the radius of the circle scribed by the dial indicator is about 3 inches. If that is true then calculating the arctangent of 0.05/3, the tilt of the bit is off about 1 degree in one plane. That probably needs to be shimmed out but doesn't seem like enough of an angle to explain that much of a discrepancy. Maybe the table isn't flat or coplanar with the router plate and the work piece angle changed as the material was being moved through the bit. Do you have a precision straight edge to check that? :confused:

[Pat Barry]

Since your measurements demonstrate the bits are perpendicular I think it proves that the bearing on the top guided bit is too large, larger than the bit itself, thereby leaving a step in the cuts. This could be verified by observation of the cut edge profile. Are the two cuts parallel (vertically)? If so, I think it proves the bearing is larger than the bit. At his point I think you should resort to hand tools to finish the edge detail though.

[Tom Ewell]

Can you flush up by using the shy edge of the guitar stock as the guide?

[Christopher Clark]

Hi Julie,
Couple of things. Are you getting flex on the tall bit? Could one of the bearings be shot?
You could also try wrapping some tape around one of the bearings to cheat it out.

[Julie Moriarty]

Thanks for all the input. I've been using your input to try to figure this out because nothing so far is making any sense to me.

The results I got could be easily explained if the bottom bearing bit cutters were larger than the bearing itself. But, using a caliper and rotating the cutters to find the largest cutting diameter, I found the cutters to be about .005" smaller than the bearing. Same with the top bearing bit. Both of these are brand new and this was the first use of either.

I tested the runout:
Betterley spindle: .003"

Bottom bearing bit at the bottom of the cutters: .002"
Bottom bearing bit at the top of the cutters: .002"

Top bearing bit at the bottom of the cutters: .001"
Top bearing bit at the top of the cutters: .003"

I checked the flatness of the router plate to the table. (The table is an extension wing that came with a Delta Contractor saw that's about 20 years old. I installed the router lift shortly after I bought the tablesaw, so they've both been around the block a few times.) There are some undulations evident when laying a straight edge from 0-180 degrees. It is not perfectly flat.

I set a Starrett 12" rule from the center of the spindle to 0 degrees out, over the gold miter slot. The plate insert I used for the top bearing bit was installed. There was a gap from the insert to the plate of .008". There was a gap at the miter slot of .011". So from the insert plate to the miter slot, there was a drop of .003".

Once again cutting from the 0 degree position, I tool a piece of wood that was square and ran it across the bottom bearing bit. Using an engineer's square and feeler gauges, I found it shaved about .001-.002" just above the bearing and about .004 at top of the wood, over a height of 1.25", the maximum cut that bit can make. However, that piece of wood rested completely on the plate insert. I didn't have a piece as wide as the guitar body and I didn't want to use any good wood for this test. So, despite the readings from the caliper, the cutters on that bit (the one that I used first in shaping the guitar body) are cutting larger than the bearing.

I realize from an engineer's perspective, I've skipped a lot of tests I would need to do to find exactly why this is happening. While there are imperfections in the whole setup, I'm not seeing them adding up to equal what I saw when I shaped the guitar body with the two bits. I'll have to find a scrap piece of wood about the size of the guitar body and just test the cuts.

What I'm really trying to figure out is if either or both of these cutters are flawed. Both of these newly purchased bits are Whiteside. About 20 years ago I bought a Whiteside bit from Woodcraft and wasn't impressed. Since then, I've avoided the brand. But they got good reviews from FWW's latest issue and they were the only brand I could find that made the compression bits I'd need for this project. I just don't want to get stuck with expensive bits that don't cut right. But so far I can't find anything in the bits that can cause such a discrepancy. I guess I'll need to do more tests.

[Richard Coers]

I'm wondering about something flexing as you are feeding the blank through the cutter. That is a lot of length on that bit, and if you really push hard against it in the cut, might it be flexing something? Could be the bit, the router base, the table, anything in the system. I really prefer the shaper for tall work like that. I have a 3" tall cutter, with a bearing below. On the router table, I prefer the Freud 50-126, 1 1/4" pattern bits. You get more tip speed and it doesn't flex like the 1/2" dia. router bit. I've always liked the Whiteside bits, no trouble with them and they are made in the USA.

[glenn bradley]

I'm still missing where we have proven that the motor is mounted perfectly perpendicular in the lift. Even if the router is off by a full degree, runout from any fixed position will read as whatever the runout is. That is; this measurement will not be effected. The feed path of the material as it varies in passing the bit would suffer from this lack of perpendicular setting. With the reference rod in the collet, does a machinists square show 90 degrees between plate and rod from 12, 3, 6 and 9 o'clock positions?

[Julie Moriarty]

I'm still missing where we have proven that the motor is mounted perfectly perpendicular in the lift. Even if the router is off by a full degree, runout from any fixed position will read as whatever the runout is. That is; this measurement will not be effected. The feed path of the material as it varies in passing the bit would suffer from this lack of perpendicular setting. With the reference rod in the collet, does a machinists square show 90 degrees between plate and rod from 12, 3, 6 and 9 o'clock positions?

I don't have a machinist's square. All I have is a couple of engineer's squares I got from Lee Valley and a Starrett combination square. Using the Starrett and a feeler gauge, I first squared up the Betterley spindle the best I could, to try to take runout out of the equation. Laying the 12" rule down on the router plate and checking at 4 points around the spindle, I found overall there was less than .001" gap along the 1-1/2" measurable height of the spindle. At "0" I'd get .0015" while at "180" I'd get .000.

I took a 2x4 that I squared on the table saw and, with the bottom bearing bit, ran it through the router at the "0" degree position. Using the Starrett protractor it read so close to 90 degrees that I'd say it was off no more than 1/4 degree.

The more I look at this and the more I take what's been offered here, the more I'm thinking this was operator's error. I didn't expect the results I got and my only focus from there was to get the cut flush in the workpiece. None of the methods that I expected to work, worked. My primary concern was that I didn't get a bad router bit or two because they weren't cheap. The finish is beautiful. But that discrepancy worried me.

I'll have to cut the miter slot in a bit deeper and flush up the router plate on one end. There's not a lot I can do about the undulations in the table saw wing, other than make a new one. I'll be doing more guitar bodies so I'll need to make sure what I can fix is fixed.

Thank you all for helping out. I may never know for sure how this happened but, as long as it doesn't happen again, I can live without knowing.