I'm Stumped and need some help

[dennis davidson]

I am breaking down some ½” Baltic birch ply for a shop cabinet and I cannot get a square crosscut.
The sheet is about 30 inches wide and my cut is out of square by about 1/64 inch, its shy on the leading edge of the cut.
I am using a Dewalt circular saw DW364 and a factory made aluminum cutting guide (which I have used successfully for more than 20 years).

Here is what I have tried to rectify the problem.
Checked my square by scribing a line, flipping it around and rescribing. It’s good.
Checked and re-aligned the sole of the saw to insure it is parallel to the blade (it seems to be good to about ½ a thou)
My cut is straight it is a good straight-edge and my cutting guide is straight.
I tried a different blade, same result, about 25 thousandth off on the leading edge.

Next – I made a cutting guide from two pcs of the ½ ply and used my saw to cut the edge of the guide.
When I checked that guide, it too was off from one end to the other, still a gap on the leading edge. I am sure that I am not having the saw at an angle when starting the cut.
There does not seem to be any play in the arbor, and I don’t know what the problem is at this point I have tried everything I can think of.
What am I missing?

[Justin Ludwig]

Sounds to me like you're missing a table saw!

1/64 over 30" is not bad for just a shop cabinet, IMO.

I know my reply is of no help.

[Jay Jolliffe]

To me 1/64 on an inch is not much in 30"....Your not working with sheet steel.

[lowell holmes]

I'm not sure whether you are complaining or bragging. :)

If the 1/64 bothered me, I would take a jack plane to it. I do both power and hand tools, according to which is easiest.

If you don't do hand tools, don't go buy a plane at the borg. You can find a good one at Woodcraft or Rockler.

A rasp could be used as well.

[Bryan Lisowski]

My only thought is that maybe the sheet wasn't square from the factory. 1/64 isn't bad, and would be well within my tolerances.

[dennis davidson]

I am definitely complaining. I guess when I put the square on the edge 1/64 (actually .025) gap just looks huge to me.

[Art Mann]

You need pretty good measuring technique to measure 1/64 inch (0.016). I use a precision straight edge and a feeler gauge. I use a Makita track saw for cross cuts that are too big for my table saw. The results are more accurate than what you are getting but even with a track saw they are partially dependent on technique. If I don't follow through, the tail end of the cut will be off about as much as you are seeing.

[dennis davidson]

I need one that cross cuts 30 inches, if only I had the room

[Tom M King]

What kind of shape is the saw blade in? It could easily be a blade pulling issue.

[Charlie Velasquez]

http://c.searspartsdirect.com/lis_png/PLDM/P0212175-00002.png
http://www.searspartsdirect.com/partsdirect/part-model/craftsman-parts/saw-parts/model-315108420/0247/0744410/p0212175/00002

Look at this diagram for a typical saw set up.
On the end of the shaft there is a bearing inside the motor, then another bearing , then a support which rests in the guard, which has a retaining ring, which has the blade washer. All of these things help support the shaft and the blade and mitigate the force of the wood as it cuts through the wood. Some of these, especially the bearings are at very tight tolerances.

**** My conjecture as I have never tested this***
I am betting the bearings are worn just a tad (or possibly your shaft). That is allowing the shaft to shift imperceptibly. A 0.1 thou at the shaft would translate to quite a bit at the end of the blade. If the shaft cants maybe 0.1 thou as the front of the blade enters the wood it might be making your kerf 1/64 wider. Then as the front leaves the wood that force is gone and it may be straightening a bit and now the kerf is smaller and parallel to the shoe.

If this makes sense, then it probably happens a lot. You would just be the first person I know that actually noticed something this small... ;)

I would be interested in testing this.

Could you rip a 3-5" board and measure the width with calipers?? Where does it start to deviate?

Also, could you cut about 6" into a board and measure the kerf then compare to the saw blade including the tooth set?

Finally, if this is true I would expect the deviation to be greater with thicker wood than thinner. Maybe try some 8/4 stuff? Take a look to see if the variance is in only the plane perpendicular to the face. Actually, after thinking about it, I guess the deviation may not be different. I guess once you reach a critical force the shaft would cant as much as it is going to. It would still be interesting to see if it is only perpendicular to the surface...

[Jim Matthews]

I think Charlie may be onto something.

There is play in one or more of the bearing surfaces.
Can you get a "feeler gauge" between the saw base plate and any part of the guide?

Can you test your blade for run out, with a dial gauge?

The first ten inches of most guides will see more passes than the last ten.

If the blade is pulled toward, or away from the guide as you cut, that will cause
the kerf to wander. Does your blade have relief cuts in the disc?

If a blade is hot, it will expand and this may account for some of the error.

I'm with many here, if I get within a 64th of my gauge line, I'll plane to the exact fit.

Can you get a clean assembly, with a dry fit?

[dennis davidson]

I bought a new one thinking that was the issue. A Freud thin kerf 60 tooth carbide.

[dennis davidson]

I think the new blade helped a lot. Being carefully with my set up I was able to cut some that are only off by about .010"
Again at the leading edge though.

however i I did try what Charlie suggested. I jointed an edge, measured off a set distance from the edge to place the guide and then cut off a strip about 2.5" wide. Then measured the width of the cutoff at 1/2" in from the leading edge and every three inches thereafter. Results are:
2.415; 2.417; 2.417; 2.421; 2.426; 2.426; 2.437; 2.442; 2435; 2.435; 2.436 total length of cut was 31"
Not sure what to make of this.

[Charlie Velasquez]

Mmmm.. I would have thought 2.415 and close till the last 7.5 inches (or whatever your blade's diameter) then 2.43X the last couple of measurements as the blade straightens and makes a more parallel kerf.

What is the comparison of kerf width to blade (offset tooth to offset tooth) width?

Are you using the guide that you made with the saw? If so, you may have made the same variable width cut on your guide.

But I can think of no other way to have the cut drift like that unless either the guide or the blade orientation varies as the cut progresses.

[Bob Hoffmann]

Just some thoughts looking at just the numbers ...

The guide is flexing at you cut -- and you can see that by the larger width in the middle where it is not clamped. I also suspect that you are pushing hard against the guide as you cut -- and that pressure is causing guide creep so that when you get to the end of the cut, it has moved a bit.

To see it that is the case -- use another clamp next to the guide clamp and put a board clamped (like a "T" perpendicular to the clamp in the middle) to prevent it from bowing.

The new blade removed some of the pressure that the saw put on the guide as it cut -- so the cut was better.

Just some alternative ideas ...

[Johnnyy Johnson]

I would bet your ruler is off

[Charlie Velasquez]

I would bet your ruler is off But as long as it is the same caliper, wouldn't it be consistently off??

[dennis davidson]

I made a cut into a board for about 8" using my aluminum guide and then measured the kerf width as suggested.
the teeth measured .074" across two teeth and my kerf width measured .066" (with feeler gauge that felt about right) I didn't find any difference in kerf width front to back (my new blade) guess i should try that again with my old one. I'm leaning to the conclusion that it was probably my older blade and perhaps some poor technique on my part. When I first encountered this I was squaring off one edge of the guide and cutting on the other edge, will also make some measurements on the guide to see if the sides are parallel. Thanks to all for your suggestions.