Trouble with 90 degree plywood corners

[Bob R., Ottawa]

I have a King KC-10CCX table saw with the Tru-Rip fence and have had no issues since I've owned it. My projects to date didn't really require me to make perfect 90 degree corners, I've been busy with making frames, and smaller projects with regular cuts. I haven't really had the need to make near perfect 90 degree corners until now. I have to apply 45 degree mitred edge banding on 3/4 inch thick plywood.

I cut my (1.5x2.5 foot) piece of plywood top but when I measured the corners they came out:
89.75, 90.15, 89.85, 90.05
I know if I let this go I know when I put the 45 degree mitred edge band pieces on they will have gaps. I'd like to graduate from glue and wood filler.

I spent the morning making adjustments:

I measured my mitre slots to be parallel to my saw blade. I used a ruler resting on my mitre gauge and took measurements from one marked tooth edge, then turned it 180 degrees back and measured the same tooth at the other end. It looks pretty good to my naked eye anyway. I don't own any dial indicators.

I then measured my rip fence to be parallel to my mitre slots. I basically pull the fence next to a mitre slot and put a straight edge down the side of my fence and down in the mitre slot and look through to see if they're flush. I even made sure my fence was parallel to my blade by doing a similar routine.

I have an Incra mitre gauge which I adjusted to 90 degrees from the blade using my digital angle gauge.

After all this I get the cut measurements above. I'm making sure I use a featherboard to keep it up tight against the fence to help eliminate human error. How close to 90 degrees are you guys and when do you recommend?

[ypinet]

Seems to me that your plywood sides are not parallel to each other when you start, and that your miter gauge is not perfectly at 90 deg. This could also be caused if you have dirt, splinters etc, stuck between the board and the miter gauge. Have you considered using a sled to cut your plywood, ensuring of course that you have first have true parallel sides to start with.

[Bob R., Ottawa]

I had a sled once but I found it wasn't accurate. IT was a plywood jig with that white cutting board plastic stuff for the runners and they were a little loose fitting. The fence didn't stay 90 degree and I didn't rig it so you could adjust it. I'd prefer something made of metal like the incra sled or something.

When you back the sled away from the blade before you make you cut, won't the whole sled be backing off the front of the table? I suppose you'd have to have long runners to deal with that.

I figured everyone here has been cutting nice square panels for years, maybe without a panel sled. I figured if you had everything set up right it should be easy. What do you suggest about my setup? How can I check to see if there's any error? Get a dial indicator rigged from a mitre guage to go against my blade tooth? Is that what most people do? It seems pretty good to the naked eye.

Are those degrees fairly severe or within normal tolerances?

Originally Posted by ypinet

Seems to me that your plywood sides are not parallel to each other when you start, and that your miter gauge is not perfectly at 90 deg. This could also be caused if you have dirt, splinters etc, stuck between the board and the miter gauge. Have you considered using a sled to cut your plywood, ensuring of course that you have first have true parallel sides to start with.

[Maxwell]

The only way to machine a perfect mitred edge on a panel is with a shaper, and a power feeder.
There are to many ways to fudge up an edge on the table saw, the least of which is your saw set up. (Humans are fallible)
Consistent feed speeds, and constant pressure, through the hole travel of the cut is what creates a perfect mitred edge.
The last cabinet shop I worked at, would only run mitres on the shaper, and they would run the two halves, at about 44.9 Deg to allow for the glue in the joint.
Believe it or not the wet glue in a mitred joint offers enough resistance to cause your final assembly to be not a perfect 90 Deg.
So IMO a 45 Deg cut that is slightly under is better than over, and I find your numbers perfectly acceptable.
Good Luck.

[stevem]

rob, mitering large panels is very tricky even with the best tablesaw especially at 45 degrees for a 90 corner, 22.5 for a 45 angle seems to be much more forgiving

first:forget the miter guage, period, even the best miter guage depends on the miter guage slot and a ton of operator skill

if you are doing a miter using the fence, use your best blade, precut a little over sized and trim back a tiny bit at a time, say a 32nd make absolutely sure that the panel sits flat on the saw all the way through the cut, even th best saw/blade will deviate a tiny bit when the load on the blade changes

if you are doing cross cuts, dont depend ona miter guage, use a sled, fine tune it carefully, and make the final cut as before ie creep up on the final dimension

remember, load changes is what throws any powertool off, so a final cut of just enough to cleanup the previous cut is the goal

[Bob R., Ottawa]

Here's what I'm concerned about. Crude drawing:
If I can't get my plywood top with all 4 x 90 degree corners when I cut my 45 degree edge banding to glue around the edge they won't meet up so I'll have to go back and doctor the 45 degree ends up to make them fit which isn't ideal.

[willr]

Originally Posted by Bob R., Ottawa

Here's what I'm concerned about. Crude drawing:
If I can't get my plywood top with all 4 x 90 degree corners when I cut my 45 degree edge banding to glue around the edge they won't meet up so I'll have to go back and doctor the 45 degree ends up to make them fit which isn't ideal.

Picture framing techniques...
http://www.leevalley.com/wood/page.a...22&cat=1,42884

Or follow Maxwells advice -- cut at 44.5 to 44.9 deg -- a touch longer than required -- close as you can get -- then trim with sander or plane.


Blades flex, and fences flex and miter gauges need to be a large size to allow accurate settings...

I suspect the heavy European Table Saws with sliding tables might do that OK -- but not a "light duty" table saw.

One other issue, deciding where the blade will strike the wood can be tricky. Mark with a razor knife on the leading edge so you can see where the saw will strike and line up the "outside edge" of the tooth that is angled towards the razor line. Hope that's clear. You learn a few tricks when you want nice corners on a jewell box.

This operation ain't easy!

[OttawaP]

A dullish blade can easily pull a cut out of square also.

[Maxwell]

Sorry Bob, I misunderstood, I thought you were cutting a 45 on the panel, and a 45 on the edging so the top would look like a thicker piece of solid.
The picture is much more helpful, Thankyou.
For a picture frame type edge, I use my mitre saw, and cut a samples until I get the angle dialled in to 45+45=90, with those two sample pieces.
Glue on three sides at once, so two mitred corners, with lots of glue, it should drip on the floor and leave a nice bead of squeeze out on the show face. Then for the third piece, cut it oversize, cut one side, and then sneak up on the other joint buy trimming small amounts until it fits nice and tight.
If you don't have a mitre saw, then as others have said a mitre sled jig is the most accurate and stable way to cut mitres on a table saw.
It will take at most an hour to build the mitre sled.
I use my mitre gauge on my table saw, maybe on 2% of the cuts I make.
Good Luck.

[KenL]

Originally Posted by Bob R., Ottawa

Here's what I'm concerned about. Crude drawing:
If I can't get my plywood top with all 4 x 90 degree corners when I cut my 45 degree edge banding to glue around the edge they won't meet up so I'll have to go back and doctor the 45 degree ends up to make them fit which isn't ideal.

Bob,

Something is a bit cockeyed as the angles don't add up to 360 degrees like they should. Probably a bit of a curve in one edge or another as a guess. That aside, all of the angles you have are near enough to 90 degrees to fall within the range of accuracy required for the task you describe, save the 90.5 degree one. That could cause a bit of a gap at the outside edge of your mitre; it would only miss by a hair on 3/4" wide banding but it gets worse as the banding widens. I suspect that you will find one short side has a very small curve and, once that is corrected, you'll be fine.

Just my guess.

Ken

[Bob R., Ottawa]

Originally Posted by KenL

Bob, Something is a bit cockeyed as the angles don't add up to 360 degrees like they should. Probably a bit of a curve in one edge or another as a guess.
Ken

Yeah, I know, I measured them a few times making sure my angle gauge was zero'd and calibrated hoping they would all add up to 360.
What do you mean a curve? You mean my 3/4 thick plywood has a bit of a bow on one side?

Originally Posted by Maxwell

For a picture frame type edge, I use my mitre saw, and cut a samples until I get the angle dialled in to 45+45=90, with those two sample pieces.

45+45=90 would be nice if my plywood was actually cut that well
My main issue is getting that plywood cut to 90+90+90+90 not what I ended up with.
Maybe I'm missing a procedure when cutting my piece? As long as I have one straight edge I should be able to turn it sideways, put that flat edge on my mitre guarge or sled and push it through to get a perpendicular side to make 90. Then I use those edges with my fence to make the opposite sides parallel. With the measurements and aducstment I made with my blade to mitre guage and blade to rip fence I would think I'd come very close. I'm wondering if what I ended up with is ok or not to your guys' standards.

I was hoping for more of a vote system like "looks ok by me" or "sounds about what I would get" or "those angles suck, you gotta fix your table setup"

I just want to know what's normal. Should I be worried about my table saw not being set up properly, or is it within specs? I'm not in the habit of squaring up larger tops like this.

[dan]

i use my plane and a shuteing board(never could remember the spelling on this) to make the trim fit, i find my power tools only get close where as a hand plane will get down to the thousands

[stevem]

i misunderstood what you were trying to do too!!

the best way to do this is cut one piece mitered on both ends a little long, then trim it i tiny bit at a time until it the right length, glue, clamp, and go to the next one etc etc

the last piece will have tobe trimmed to fit between two miters

you're not alone with this scenerio, so dont fret about your equipment or techniques

[aussie]

Rob before I got hold of all the digital devices I have now I used a bit of trial and error to get my Incra mitre gauge set to exactly 90deg. I took a piece of scrap hardwood and ran it through the blade for a straight 90deg cut. I took the two pieces and turned one upside down but with the cut faces still facing each other. I placed a straight edge along the two pieces. I kept adjusting the mire gauge till I got the straightedge to lay perfectly flat down the sides of the wood. Same thing for making sure the blade was at 90 to the top. Now I'm lazy and just use digital angle finders and an Align-It.

[KenL]

Originally Posted by Bob R., Ottawa

What do you mean a curve? You mean my 3/4 thick plywood has a bit of a bow on one side?

Bob,

That is quite likely the case, that your plywood has one side that is not quite straight but is probably within 1/64" on an 18" side. Straighten that and you have it knocked! Your technique seems fine to me and the results are very good as I indicated. The math on your angles is such that they are square within normal woodworking practice except for the 90.5 degrees.

Cross cutting plywood without a sled is a b---th to keep precisely square, and, in my experience, plywood rarely comes from the factory precisely square either which adds to the misery.

I hope this helps

Ken

[willr]

If you feel that it is the plywood not square, then you can do the following:

Make one straight edge -- lay a ruler along it to check it after you cut!.
Using a T-Square (Drywall) mark the first 90 degree corner. Hopefully the T-Square will give a 90 degree angle.

Check the corner using a 3-4-5 Triangle method
(This is easiest with a metric ruler)

Layout the 3-4-5 Triangle with a measuring tape or ruler as appropriate for the size, and then check the hypotenuse with the same deice.

When you measure the Hypotenuse with a ruler or tape, check the variance in the expected length. It's easiest to calculate the percentage error with the metric measurements.

Starting from the same side as the first corner -- layout the second corner -- and check the markings as above.

Cut the board and check the variance introduced by the saw.


This seems pretty involved -- but it only adds a minute per side. It allows you to quantify the error and decide on a strategy.

3-4-5 Triangles below -- use inches centimeters, millimeters yards meters or feet as appropriate. The longest measurement within any row below is the hypotenuse of course.


3 4 5
6 8 10
12 16 20
15 20 25
29 36 45


Once you have square corners, lay out the opposite side. It is easy to check that the opposite side remains parallel to the original size. Using the T-Square -- check the the opposite side remains a constant distance to the first side.

When you cut the piece, cut a bit oversize -- so you can see how the saw tracks to the lines you created.

Adjust miters and fences appropriately to get the initial square corner.

When you cut the second 90 degree corner -- you typically flip or rotate the board -- this means any error is likely in the opposite direction. Adjust as appropriate.


When you make the last cut -- the parallel cut -- again slightly oversize -- you can now get an exact (within measurement capability) check of the how the rip fence tracks. Again trim as necessary or adjust fence.

Record any offsets as you go -- an offset being the difference between the reading and what you actually cut.

This is a lot easier with a track saw -- but I do the same with my table saw when I need "exact" cuts.

You should be able to get accuracy within the width of a ruler line using the 3-4-5 checks.

If you know geometry this all seems pretty obvious -- but you have to do this occasionally to check the accuracy of your marking and cutting. This method will check the system results. You then use those results to find the particular error.

Use a razor knife to mark the measurements -- pencils don't cut it -- too much error.


Another note: To those who use calculators to check the sine and cosine or tangent to check the angle -- near the "boundaries" the algorithms in cheap calculators often fail. They return answers that might be good only to a couple of decimal places -- maybe threee places. The better calculators have "look-up" tables and the answers are "pretty good". Some calculators have different calculation algorithms near the "boundaries) -- 90 degrees, 0 degrees 180 degrees etc. and the accuracy is differen in these regions. This is why I suggest the 3-4-5 method. Be aware that some angle measurement devices may display the same problems at these "critical" readings. With 3-4-5 you can see if the hypotenuse is off by a millimeter or a fraction thereof.

3-4-5 was good enough for the Egyptians and the Mayas to lay out their temples...

After you have done this you can check your LV square with the accuracy of 1/1000 of an inch against your corner.

The measuring device you use must have fine markings, it must be only one device -- and hopefully it is accurate or consistent -- either will do!.

[niki]

Hi Bob

In a perfect world, if the blade is aligned to the miter slot and the rip fence is aligned to the miter slot and, the miter gauge is set to 90º to the miter slot (or miter gauge bar), everything should come square at 90°.....

But, usually, there is no "perfect world"...there are some very small misalignments that when summed together, can lead to a big (or not so big) accumulated mistake....

The rip fence, should be aligned to the blade. Just imagine that the blade is set - say - 5° to the miter slot and the rip fence is also set to 5° parallel to the blade...you will get a parallel cut with consistent width all over the workpiece.
If the rip fence is set at 3° to the blade, you'll get, pulling or pushing of the workpiece, most probably with cutting and burn marks and the cut will not be at a consistent width...

When I'm aligning my TS, I first align the blade to the miter slot - at the same method that you are using....I call it "Static check"...After that, I use the "Dynamic check" to fine tune the blade / miter slot alignment....

Why "Dynamic" ? because it's done with the blade spinning at normal speed and all the misalignments - that you couldn't see in the "Static check" - will show-up...

Please look at the pics below for the method that I do the "Dynamic test"...use MDF as a test material...

01.jpg


02.jpg

On this pic (and the next one), I'm showing the blade stationary which is not the case.
You should push the test piece past the blade while the blade is running at normal speed and listen to the sounds...
03.jpg


04.jpg

After you align the blade / miter slot, you can use any method to align the rip fence to the blade...I do it like that...


05.jpg


06.jpg


07.jpg





To make the "Dynamic check" I use the same method as before...








I don't know your table saw but just remember that the alignment might change when the blade is tilted to 45º (or any other angle) so it will be a good Idea to check the "Dynamic test" also with the blade tilted to 45°

Regards
niki

[niki]

08.jpg

17.jpg

19.jpg


And the last pic....

20.jpg

niki

[Bob R., Ottawa]

Above and beyond Niki! thanks for the pics and description.
I'll try out the dynamic test soon for the sounds. I did pick up a dial gauge today so I'll try to also follow what I've read by rigging it up to a mitre gauge to measure from the front and back position of the blade (same tooth).

[Erron in Arnprior]

Originally Posted by Bob R., Ottawa

Above and beyond Niki! thanks for the pics and description.
I'll try out the dynamic test soon for the sounds. I did pick up a dial gauge today so I'll try to also follow what I've read by rigging it up to a mitre gauge to measure from the front and back position of the blade (same tooth).

Hey Bob,

Something else that I didn't see mentioned. It actually takes quite a bit of practice to cut a perfectly straight and parallel line on the table saw. The bigger the piece that harder it gets. Try practicing on scrap. You might be surprised how much you're off.

Could be just me though. Went through the same frustration you're having. Incra miter gauge is great. Building an accurate sled is absolutely critical. I put it off for years, and ended up with the same problems your having now.

One other note, if your miter gauge doesn't have a stop, then you'll never get accurate results. You just *can't* hold a piece tight enough, or steady enough to crosscut a perfect 90.

Good Luck,
Erron.