TS55-REQ and MFT-3 Not Cutting Square

[Stone Message]

When I saw the pictures that Crookedcutter posted above my first thought is to ask if the top and bottom edges of the piece of wood are straight and parallel?

The I saw Ron's post about the unreliability of squares. I bought a brand new woodworkers' square from a well established company and used it for several weeks before I realised that it was not truly square. I then checked every single square that I owned and found only two out of six were up to the standard that I now expect to achieve with Festool kit.

I now use an engineering square which is made to a much higher standard than the run of the mill woodworkers' squares.

Just yesterday I was cutting walnut drawer fronts, with my Kapex, for a pedestal desk that I am making and checked my cuts with my engineering square. They were out by too great a margin for this sort of project. So I spent the latter half of the day checking everything...

I had recently checked the alignment of the Kapex but only for cuts with the stock on the left of the blade. The out of square measurements were noted from the stock from the right side of the cut. My Kapex left and right rear fences were not in line. I verified this using a Veritas straight edge. I removed the two sliding fence parts and checked the fixed base part (which the left and right fence slide on) and this was in line as it should be. I then removed my zero clearance sacrificial parts from the left and right sliding fences and refitted the fences to the Kapex. Using the straight edge I found that the sliding fences, when fixed, are not in a straight line. The error occurs due to play in the sliding mechanism. I have managed to clamp the sliding parts in place in a straight line by holding them in the same plane as they are being tightened with the clamps.

I love my MFT3 and use it all of the time but when I am doing cabinet grade work I cannot rely on the accuracy of the fence/rail parts as designed for square cuts - hence the Parf Dogs. But for 95% of my work the MFT3 is perfectly good and I would not be at all concerned if I were doing normal work. The OP has a different problem to solve and it will be interesting to see exactly what is causing the issue.

Peter

 

[crookedcutter]

That means that the saw and your technique is not introducing the error. Just wanted to narrow down the potential causes.

What this tells me is that the rail and fence are simply not square.

For kicks, can you check your square? Do the same thing as before. On your plywood, strike a line with your square. Flip the square and strike another line. They should be right on top of each other.

EDIT: You would only need to be off by less than 3/32" at 12" to have the margin of error shown over 22".

Shane

Hi Shane,

Below is a picture of the square and the lines you requested.  Everything seems to be fine with the square.  The square is brand-new, as I bought it a few days ago when I realized I was having problems.

Thanks...

[attachimg=#]

 

[Administrator_JSVN]

I meant referencing the square from the edge of your plywood, strike a line. Then flip and repeat. It looks like you drew a line to reference from, which isn't very accurate since you have to eyeball the alignment.

Can you try again? Thanks.

 

[crookedcutter]

I meant referencing the square from the edge of your plywood, strike a line. Then flip and repeat. It looks like you drew a line to reference from, which isn't very accurate since you have to eyeball the alignment.

Can you try again? Thanks.

I can do that, but that would not be checking the square's  two outside edges, which are what I use to square up the table.  That method would check, for example, the outside long edge against the inside short edge.  But, I'll do it.  Thanks...

 

[Stone Message]

I meant referencing the square from the edge of your plywood, strike a line. Then flip and repeat. It looks like you drew a line to reference from, which isn't very accurate since you have to eyeball the alignment.

Can you try again? Thanks.

I can do that, but that would not be checking the square's  two outside edges, which are what I use to square up the table.  That method would check, for example, the outside long edge against the inside short edge.  But, I'll do it.  Thanks...

What Shane is asking you to do will check whether your square is out.

Put it against the edge of your board with the thick part to the left (or top). Draw a line along the square. Then, flip it so the thick part is on the right (or bottom) and draw a second line.

The two line should be on top of each other if your square is spot on.

But...

If the edge that you are using as a reference is not dead straight then the test is meaningless. Have you checked that the edge you are referencing is dead straight? If so, what did you use to do the check?

Peter

 

[grbmds]

For what it's worth, I have found that the only type of square which has an extremely high chance of actually being square is a machinist's square and then only when they are high quality and a little bit more expensive. Having said that, however, I have also found other squares that are really square, but the surprising thing to me was that, even with high end squares, you are not 100% assured of squareness if you are looking for complete perfection. That was definitely a surprise. Many are very close though. It depends on what you require and what satisfies your needs I think. I'm kind of a perfectionist (unfortunately for me).

 

[Canoescapes]

I too had problems with square cuts on the MFT after setting the table up repeatedly with a Starrett framing square.  I then acquired Peter's parf dogs and some qwas dogs and used them to check the framing square.  It was not accurate -- a surprise for a company with Starrett's reputation.  I then used the parf and qwas dogs to set up the table and the problem has been solved.  As suggested by bkharmon, check out the videos by Peter and Paul Marcel on setting up the MFT using parf or other dogs.  They are relatively inexpensive and have multiple valuable uses, not the least of which is ensuring the MFT cuts square.

 

[bkharman]

I too had problems with square cuts on the MFT after setting the table up repeatedly with a Starrett framing square.  I then acquired Peter's parf dogs and some qwas dogs and used them to check the framing square.  It was not accurate -- a surprise for a company with Starrett's reputation.  I then used the parf and qwas dogs to set up the table and the problem has been solved.  As suggested by bkharmon, check out the videos by Peter and Paul Marcel on setting up the MFT using parf or other dogs.  They are relatively inexpensive and have multiple valuable uses, not the least of which is ensuring the MFT cuts square.

One other thing here...  You can get very repeatable results using dogs, but be careful using rail dogs. If you have some that fit loose in the bottom of a rail, it can cause the rail to not be square with the holes it is using. Most of the newer versions have a small lip that settles into the rail nicely.

If you need/want a fence, invest in some fence dogs (I don't think that is the name, but it is in the aftermarket section).  When I absolutely need a fence, I use those. I will post a pic after work (damn 9-5 job!!)

Cheers. Bryan.

 

[tjbnwi]

Without doing a 5 cut test you have no way of knowing whether it is the square, the fence or the rail. You have not established a true reference edge to work off of. Once you do a 5 cut and get the MFT dialed in you'll have a piece to check your square with.

Tom

 

[Carl Prentiss]

Isn't that like building a house from the roof peak down? ???

 

[rst]

Don't bother with a square... make gauge blocks, one for the rail and one for the fence.  The MFT hole pattern is precise.  Using dogs and a precise measurement (preferably a digital caliper) to make a gauge block for both guides.  When ever you need to re-calibrate, use the blocks to to set your distances.

 

[tjbnwi]

Isn't that like building a house from the roof peak down? ???

???

Tom

 

[Carl Prentiss]

Maybe not a good an analogy Tom. Just saying a known accurate square to calibrate the table vs a known squared table to validate a square. And with that I'll stop creating background noise so the OP can get his MFT squared away.  [oops]

 

[crookedcutter]

For what it's worth, I have found that the only type of square which has an extremely high chance of actually being square is a machinist's square and then only when they are high quality and a little bit more expensive. Having said that, however, I have also found other squares that are really square, but the surprising thing to me was that, even with high end squares, you are not 100% assured of squareness if you are looking for complete perfection. That was definitely a surprise. Many are very close though. It depends on what you require and what satisfies your needs I think. I'm kind of a perfectionist (unfortunately for me).

That's crazy.  I'm with you.  You you charge a ton, and call it square, it should be square...

 

[crookedcutter]

Isn't that like building a house from the roof peak down? ???

I'm sorry, I don't understand.  Are you saying there's an issue with the order or process of my cuts?

 

[crookedcutter]

Without doing a 5 cut test you have no way of knowing whether it is the square, the fence or the rail. You have not established a true reference edge to work off of. Once you do a 5 cut and get the MFT dialed in you'll have a piece to check your square with.

Tom

Tom,
I'll be the first to admit that I'm certainly no pro, and there are a lot of things I don't understand.  I could very well be wrong, but, wouldn't the order of cuts that I made cause there to be LESS of an error than the 5-point method would?  If I'm getting this kind of error cutting in the order I am, wouldn't a 5-point cut just magnify the error?  Thanks for your insight...

 

[crookedcutter]

I meant referencing the square from the edge of your plywood, strike a line. Then flip and repeat. It looks like you drew a line to reference from, which isn't very accurate since you have to eyeball the alignment.

Can you try again? Thanks.

Hi Shane.  Please see below.  The lines lined up perfectly.  Thanks,

[attachimg=#]

 

[bkharman]

Without doing a 5 cut test you have no way of knowing whether it is the square, the fence or the rail. You have not established a true reference edge to work off of. Once you do a 5 cut and get the MFT dialed in you'll have a piece to check your square with.

Tom

Tom,
I'll be the first to admit that I'm certainly no pro, and there are a lot of things I don't understand.  I could very well be wrong, but, wouldn't the order of cuts that I made cause there to be LESS of an error than the 5-point method would?  If I'm getting this kind of error cutting in the order I am, wouldn't a 5-point cut just magnify the error?  Thanks for you insight...

You re correct... The 4 cut (or 5 cut) method will amplify the error, generally it would be used to amplify small errors (1mm or less) but in your case, I would say it would be to verify your "offness".

I say you do it anyways.  If you aren't sure how to do it correctly, you can refer to this timely thread...  http://festoolownersgroup.com/ask-festool/5-cut-calibration-method/msg310894/?topicseen#msg310894

 

[Carl Prentiss]

Isn't that like building a house from the roof peak down? ???

I'm sorry, I don't understand.  Are you saying there's an issue with the order or process of my cuts?

No, not at all. Sorry for the confusion.

 

[Administrator_JSVN]

Hi Shane.  Please see below.  The lines lined up perfectly.  Thanks,

Ok, good.

So, it looks like your square isn't the issue.

And, we ruled out the saw being the issue.

I'd like to rule out the guide rail next. Can you repeat scribing a line on your rail on side of the rail OPPOSITE the side you checked before, flip and repeat. See if those align. If they do, your rail appears to be the problem because it aligns on one side and not on the other. If they do not, then your rail is not the issue.

And, for those wondering, I come from an IT background. So, my methodology is to troubleshoot by systemically removing potential causes.

Shane