So, it is not squared...why? - Getting crazy - please help

[kitui]

So I had a piece of wood, cut all 4 sides in sequence with the FS track rail (using clips and long bench dogs) and some positioning dogs. I checked.... the positioning dogs and the saw rail are perpendicular (precision square was used to verify and...the square was "verified" as well last week when I purchased it). After all cuts in sequence, I check the squareness and on the last side is not 90 degree with the 2 adjacent sides. I am getting crazy...what could it been? Could it be the blade of the saw is not straight? The saw is brand new (3 months old). I follow this process 3 times with no luck.  [eek]
For your information...I am doing what Peter P does in:=95s

So ... what it is weird is that only one corner is off...this is mathematically impossible

 

[mwbrewster]

Using Peter's method, how have you verified the first (reference).edge is perfectly straight and square?

you may want to make a 5th cut of the first side.

 

[kitui]

It was a fresh cut...so I would say I cut the piece of wood 5 times.  [sad]

 

[Corwin]

...
So ... what it is weird is that only one corner is off...

Sounds like user error. Something must have moved or you don't have a close relationship with your dogs.  [unsure]

 

[GoingMyWay]

Are your long dogs bolted to the top from the underside?

 

[mwbrewster]

Kitui, just sent you a PM.

 

[kitui]

My dogs are held by knobs underneath . I might try another set of holes and my pushing position. I am not pushing forward but side by side (dogs are resisting my push so they are in the right position). Getting frustrated. I have had great results and made so many fabulous face frames. Not sure what happened in the last 5 days.

 

[sancho57]

heres a long in-depth thread on the 5 cut method
http://festoolownersgroup.com/festool-how-to/five-cut-method-on-mft3-explained/

 

[kitui]

Not sure I totally understand the Kapex and the issue about angles. I followed Peter Parfitt 's video which follows elementary math and few algebra property. So assuming that Side A is straight, assuming that my dogs are perfectly aligned and perpendicular to the rail, here what should happen. I cut side B perpendicular to side A. Turn 90 degree, cut side C perpendicular to side B, same for D perpendicular to C. If everything goes well, I am done because of Math Transitive Law: D is perpendicular to A.  Although, I cut side A again to find if there is any difference between the two ends of the cut. After 2 additional attempts, I think that there were 2 issues. My position while cutting and the standard clamps (no Festool) that might have shifted the board. I am getting closer to better results but not as good as Peter. Using a bad ply. Probably if I try with BB, would be best. Anyway, here my result on the "recut" of side A: the difference between the 2 ends is 0.79mm. 0.79mm divided by 369mm (length of the last cut) and divided by 4 gives me 0.00053777. Which is the margin of error on a single cut (ass per Peter's video). I think it is ok after all. Correct?

 

[Birdhunter]

When you said you were pushing the saw side by side, do you mean applying pressure at an angle to the direction of the cut?

 

[kitui]

Yes, I am not very tall and the cut was more than 24"...so I had no choice. Having the cut done side by side allows me to cut short and long (using the longest side of my MFT). However, it should work anyway because the saw rail is 55"...what about if I have to cut ..let's say 47" length piece? Maybe I just need to be aware of angle pressure.

 

[Birdhunter]

I try to push the track saw in line with the cut. There is less chance of skewing the cut line.

 

[mwbrewster]

I try to push the track saw in line with the cut. There is less chance of skewing the cut line.

thats good advice.  I would also suggest keeping your left hand on the auxiliary handle and applying gentle pressure to make sure the saw sits square/plumb on the rail through the cut.  When I get in a hurry and only use one hand, I tend to stand offset from the rail and the quality of cut is diminished.  Using two hands and keeping your shoulders centered over and in line with the rail through the cut is good practice.  You'll probably notice that the saw works through the material easier if you do. If so, that would suggest that you've been introducing unwanted deflection into the saw/cut, potentially resulting in cuts off square.

 

[ElectricFeet]

So ... what it is weird is that only one corner is off...this is mathematically impossible

This is possible only if one or more of the cuts are not straight. Do you have a good straight-edge to check? (N. B. Not a ruler, which are not guaranteed straight.)

Another thing to check is that you are not skewing the cut as you move along. That’s possible if you’re stretching and your saw is not adjusted properly on the rail. This video shows how to adjust:
https://www.instagram.com/p/Bh62EM8H3Ts/?hl=en&taken-by=festoolsedge

 

[justard]

I have learn something new here today

 

[Pant]

So ... what it is weird is that only one corner is off...this is mathematically impossible

This is possible only if one or more of the cuts are not straight. Do you have a good straight-edge to check? (N. B. Not a ruler, which are not guaranteed straight.)

Another thing to check is that you are not skewing the cut as you move along. That’s possible if you’re stretching and your saw is not adjusted properly on the rail. This video shows how to adjust:
https://www.instagram.com/p/Bh62EM8H3Ts/?hl=en&taken-by=festoolsedge

No, this is what is expected: after the 5 cut method is done the resulting keeper piece will have one corner with the error compounded 4 times while the other corners have the normal error. The fact that he thinks the other corners are ok is why this method is so good, it has revealed that there is still error when the try square is lying that everything is ok.

Joining pieces later with small error will compound to a big error, so having also the compounded error corner dead square to start with is a good headache saver. But in this case he cannot get rid of the error because it is inherent to the holes and dogs.