MFT setup

[dframe]

I am new to woodworking in general, Festool in particular, and of course, to the FOG.  I see quite a few threads about setting up an MFT to be square.  I am having difficulties.  I have looked at the recommended videos, used Qwas dogs and etc, and have made something like 30 tries at the 5 cut method.  Not only am I not square, the results do not seem to be reproducible.

So I decided to repeat the same cut, same edge to fence x 6 and see if I was getting reproducible results.  Short answer, I am not.

6 cuts yield: 

+15 (all in thousands)
+15
+35
-15
+50
-10

All of these measurements were repeated with identical results...so at least I can measure.  My question is, do these results suggest a deficiency in technique?

The rail and the fence both appear stable to me.  I would appreciate any ideas about technique, or thoughts about the capability of the MFT and TS75.  The TS seems to be correctly adjusted for no wobble on the rail...perhaps this is what can be expected?  My guess is that there is something about technique I am missing. 

The driving question for me is:  Can this setup be used with comparable accuracy to a table saw (which I do not have, and do not have room for) with the intention of building furniture.  How accurate does one have to be to build a 6' long desk?

 

[Svar]

What was the length of cut? You report deviation in thousands, but over what length?

 

[Gregor]

I found that my results are not reproducible when I forget to support the rail over the whole length (using TS55) because of bending, or with small (or bigger, when the table is covered in dust) pieced when forgetting to fix them in place because of sliding. With a TS75 (heavier) I would expect at least the first issue to get bigger.

 

[grbmds]

[member=62169]dframe[/member] I'm sure you'll get a ton of answers over time. . . First, I will assume that you are using the MFT with the supplied fence and miter head from Festool. (If not, then some of my answer is irrelevant.)

All of my crosscuts on the MFT3 are more accurate than the table saw I used to own (and possibly than the tablesaw I currently own although I rarely crosscut with the tablesaw since the MFT with TS55 and track is so accurate.)

I have never used the 5-cut method so can't speak to how well that works. What I would say about it is that it does magnify how off-square the setup is since; I would think 4 times the actual difference since you are making 4 cuts (after the first one to create a straight edge) Thus, it is not out of square nearly as much as you think.

Shortly after I bought my MFT3 I also bought the Slop Stop which is installed in the underside of the track. It essentially eliminates any play between the MFT's rail support and the track underneath. Slight "play" seems to be a given with the MFT and rail setup. That alone might get rid of most of your problem. There are other things you could do instead of buying the Slop Stop and, a search on You Tube or even the FOG should turn up at least a couple of ideas.

The reason I don't use the 5-cut method is because I have a square which I am certain is essentially a perfect 90 degrees. I square my rail to the MFT fence with that before I make any final crosscuts on a project and I can truthfully say the result is essentially a perfect 90 degrees when I assemble the final product (or so close I can't tell). There are many videos online about squaring up the MFT and I'd suggest searching for them and picking the one that makes the most sense to you and/or gets the best result for you. As a starting point, see if you can find the Festool video online which talks about squaring. It is by Brian Sedgeley who teaches for Festool and, for me, makes the most sense, is the easiest, and works every time.

Woodpecker's makes great accurate squares, like the 1281. There is a plywood square by Anderson which I understand is accurate also. Find a square in which you have complete confidence and square your fence up with that.

As for technique, it could be that part of the way through the cut you are applying side pressure to the saw and, thus, moving the blade sideways just enough to cause the problem. I stand behind the saw and push forward or at least try to keep my are directly behind. When you push from the side, you do exert sideways pressure on the saw and blade.

Sorry you're having trouble but hopefully something I said will help.

I'm sure there are many others who can provide more suggestions at least as good or better than mine.

 

[mkasdin]

Try ripping a piece of sheet goods, do two parallel cuts have any deviation across the length?
Do the same with a cross cut.  Is the scrap equal across the length?
You can also take a pencil line that you've made with a known good framers square or a WP 1281 i.e. Be see if the saw (you) cuts evenly.  If you can make a square cut then that's half the battle. I think the slopatop is probably a good idea as the above poster mentioned also try making two cross cuts or rips turning the work piece over. Are the diagonals even. Check them with a square.

 

[Birdhunter]

My process does not involve using the MFT protractor. I use a Woodpeckers framing square with one edge against two Qwas dogs and the other against the guide rail. I do have the SlopStop installed. If the guide rail is not square, I loosen the two bolts at the head of the guide rail and line it up as above and re tighten the two bolts. Unless I have bumped into the rail in its upright position, the guide rail is always square.

When I make cuts, the top edge of the wood is butted against several Qwas dogs and the rail is locked down over the top of the wood. There isn't any slop in the rail and the wood is secured. I try to push the TS55 from directly behind so as to not put and lateral pressure on it. Lateral pressure on the saw can throw off the cut slightly.

If I am making high precision cuts, I use my big Sawstop. Other than its original set-up, I've never had to re square it's. It's a big hunk of steer and cast iron and it is bullet proof.

 

[Michael Kellough]

The MFT/track saw can be utilized to produce extremely accurate and repeatable results.

Before anything else you need to develop consistent technique.
If all else stays the same and yet the results vary then you're not consistent.

Do some more basic trials without changing anything, just focusing on consistency and check the results.

Before going on to the 5 cut test make sure you can do a single cut repeatably. Think through the whole process; checking that the edge of the wood you place against the fence is; solid (no compressible debris), straight (it doesn't rock against the fence), etc. Everything counts.

If you're sure you are consistent, then you need to look closer at your setup to find out what is changing.

 

[dframe]

Thanks to all!

ERROR CORRECTION:  I reported the deviation of my  test cuts not in 1/1000, but as 1/10,000's.  So just now, my technique got 10X better!  Across 6 cuts, I am off less than .01 of an inch.  Apologies...

Length of cuts: Cuts measured in my first post was around 8 inches.  Not where I can look at it now

Support the rail over the full length...will add that to technique. 

Festool video...think I've seen it, will make sure.

Purchased a PEC square set that arrived yesterday, will start working with that tomorrow.

Side pressure:  Great suggestion!

Slopstop:  Should be here in a few days

Crosscuts:  Were what I described originally.  I'll work on rips when I master the 8" crosscut. 

Consistency:  Pretty sure that's my problem.  I have been somewhat attentive to sawdust, perhaps not enough.  Totally unconscious as to angle on the saw.  I will do another set of test cuts with better technique and report back soon.  I had this naive idea that high quality tools rendered technique less important.  Nope, it's just like guitars.  Dammit! 

Thanks again all!

 

[Svar]

Thanks to all!

ERROR CORRECTION:  I reported the deviation of my  test cuts not in 1/1000, but as 1/10,000's.  So just now, my technique got 10X better!  Across 6 cuts, I am off less than .01 of an inch.  Apologies...

Length of cuts: Cuts measured in my first post was around 8 inches.  Not where I can look at it now

You worry to much. Even if your greatest deviation accumulated over 4 cuts is 0.005" (0.13mm) over 8" (20cm) it is still accurate for any furniture project. A speck of saw dust between your piece and the fence could cause that. This is a portable woodworking table not a Bridgeport mill.

 

[Holmz]

...
You worry to much. Even if your greatest deviation accumulated over 4 cuts is 0.005" (0.13mm) over 8" (20cm) it is still accurate for any furniture project. A speck of saw dust between your piece and the fence could cause that. This is a portable woodworking table not a Bridgeport mill.

Even an old Bridgeport usually requires some scraping of the ways to get the X and Y axis square.
But I get the point.

Another advantage of metal is that it is not hydroscopic, so who knows what the wood will measure on a rainy day.
And them even the acropolis was designed more to look good than with "perfect dimensions", so perfection is also in the design.

But we do expect that the precision and skill to achieve the levels required need to exist as a basis upon which to layer design, craft and artistry on top of.

 

[Gregor]

I had this naive idea that high quality tools rendered technique less important.  Nope, it's just like guitars.  Dammit!

High quality tools dos't replace the need for technique completely, but make it less important in the way that you only have to deal with your shortcomings and not additionally with the ones of the tool you use...

 

[grbmds]

I would agree with [member=15585]Svar[/member]. 1/128th of an inch is .007 inches; not is on .003 off from your cuts already. While I suspect that using a square like Woodpecker's 1281 would get you a tiny bit squarer, the proof really is in how the final product goes together. There are many times I must do some slight adjustments with a block plane, sandpaper, or try to shave off material to make the pieces fit anyway. Maybe you can find a small project to make as a test to see how square the cuts really are. While mine are not absolutely perfect, all my work has improved since I started doing final crosscuts with a squared up MFT and fence. Just keep at it. Using the track saw does take a little getting used to since it is a little different than using a tablesaw or bandsaw.

 

[dframe]

High quality tools dos't replace the need for technique completely, but make it less important in the way that you only have to deal with your shortcomings and not additionally with the ones of the tool you use...

Just EXACTLY like guitars...except for the dust.

 

[Shane Holland]

I'm always surprised by how many folks work in thousands, or in this case ten thousandths, of an inch.  [blink]

My tapes are good to 1/16th or 1/32nd and that's never produced poor results for me. Maybe 1/2 mm on my metric tapes. Always dead-on square.  [big grin]

[attachimg=1]

 

[Svar]

I'm always surprised by how many folks work in thousands, or in this case ten thousandths, of an inch.  [blink]

My tapes are good to 1/16th or 1/32nd and that's never produced poor results for me. Maybe 1/2 mm on my metric tapes. Always dead-on square.  [big grin]

My point exactly. Some fret about their guide rail not being straight. I say, just chill out.
[attachimg=1]

 

[dframe]

There is progress to report.  After a little adjustment, the error was 0.002 over 40 inches cut.  Pretty sure the next trial would be a little different (and I will try it, and report back if there is anything interesting).  But for now, i am going to declare victory.   

The factor that seemed to make the most difference was carefully dusting and de-burring if necessary, the edge going against the fence and the table under the workpiece.  I also agree with the poster that pointed out that everything matters. 

Wondering about making a micro-adjustable stop for the rail mounts. 

Thanks again all!

 

[Svar]

After a little adjustment, the error was 0.002 over 40 inches cut. 

Congratulations. This must be the most accurate cut ever produced on an MFT.
(0.05 mm over 101 cm, just thinking in metric). How did you fit it between rail supports? Put them on short sides of the table?

 

[dframe]

That was the result of a 5 cut process, with sides of 11 and 9 at the end.  I seriously doubt the next time will be so good, but it's probably close enough for anything I will make.  We'll see.

 

[Svar]

That was the result of a 5 cut process, with sides of 11 and 9 at the end.  I seriously doubt the next time will be so good, but it's probably close enough for anything I will make.  We'll see.

Got it (cumulative length 11+9+11+9). Nice way to express deviation of a 5 cut test.
Think of it this way:
Woodpecker boasts their precision aluminum squares to be under 0.001" over 12". You've got 0.0006" over 12".

 

[Michael Kellough]

Save that big piece and protect it to use as a quick jig to reset 90* in the future.
You can't buy anything as good.