FS 2700/2 Tolerance

[edwarmr]

Hi guys,

Quick question for you. Recently got the FS 2700/2. How tight should I expect the tolerance to be on this rail? When ripping an 8’ sheet it is off about 1/64” which means when I join two 8’ boards together there is approximately a 1/32” gap on each end.

I put my 50” precision straight edge on the part of the extrusion the saw rides on and it is out about 0.008” over that length which brings it to about 1/64” total over 8’. Is this acceptable? I imagine it’s hard to get a 8’ rail dead straight seeing how Starrett charges insane money for just a 6’ precision straight edge. However the FS 2700/2 is also really expensive.

I have seen posts on here saying variation is normal and others saying their rail is dead on. Just looking for some input, thanks!

This picture shows the result.

[attachimg=1]

 

[Lincoln]

Are you sure the material hasn't sprung slightly, as the tension is released when cutting? Can happen when you rip a piece straight off a sheet.

 

[Vtshopdog]

Not really answering your question, but if you need tight fit on your joint you can negate any arc in a long track by clamping and recutting edges of both boards at same time. 

Not sure how straight or wiggly my big track is but I occasionally do below operation:
True up two edges to be joined.
Butt the two edges together and clamp both boards in place.
Place track edge over the butt so saw kerf is split between the two boards and make another cut.

Regardless of how straight (or crooked) your track is the two board edges will now mate up nicely
(note you are locked in to whatever grain orientation you had them in when making the final cut though)

 

[edwarmr]

Are you sure the material hasn't sprung slightly, as the tension is released when cutting? Can happen when you rip a piece straight off a sheet.

Thanks for the input! It’s definitely possible. I did however make multiple cuts and the results were pretty consistent. Also the straight edge does rock ever so slightly against the rail so I know it isn’t perfectly straight.

 

[edwarmr]

Not really answering your question, but if you need tight fit on your joint you can negate any arc in a long track by clamping and recutting edges of both boards at same time. 

Not sure how straight or wiggly my big track is but I occasionally do below operation:
True up two edges to be joined.
Butt the two edges together and clamp both boards in place.
Place track edge over the butt so saw kerf is split between the two boards and make another cut.

Regardless of how straight (or crooked) your track is the two board edges will now mate up nicely
(note you are locked in to whatever grain orientation you had them in when making the final cut though)

That’s a really good idea, thanks! In this instance I am not actually joining these two boards but just trying to get as straight a cut as I can. If I do need to join boards though I will be doing what you said.

 

[makpacman]

Perhaps try some rips of mdf or 3/4 melamine.

When I use the fs2700/2 I get very straight cuts. When I started, I was using mixing different brands of guide rails and the tolerances were just different enough that I often found my saw wandering a bit left to right if it wasn't adjusted to that particular rail. However that's not an issue for me presently.

The rocking you describe does sound like a problem. Fwiw, my rail does not rock on an 8 foot straight edge at all.

 

[edwarmr]

Perhaps try some rips of mdf or 3/4 melamine.

When I use the fs2700/2 I get very straight cuts. When I started, I was using mixing different brands of guide rails and the tolerances were just different enough that I often found my saw wandering a bit left to right if it wasn't adjusted to that particular rail. However that's not an issue for me presently.

The rocking you describe does sound like a problem. Fwiw, my rail does not rock on an 8 foot straight edge at all.

Thanks for the advice! The fact that yours is dead on makes me think it might be worth it to exchange it. On the other hand I’m worried the replacement could be the same or worse

 

[Sparktrician]

What does the rail look like when laid up against a 6' level?  The machined surface of the level might show that the rail is not true. 

 

[Crazyraceguy]

I have had an FS3000 for a few years now and it has always been dead on.
I have found that is does depend on the material being cut. As Lincoln suggested, some of it may be the tension within the board.

 

[TSO_Products]

[member=70011]edwarmr[/member]  – diagnostic suggestion:

Start by determining guide rail straightness before making any cut.
Follow [member=7493]Sparktrician[/member]’s suggestion to use an accurate Level (aka STABILA) along with a post it NOTE paper to check for clearance is an excellent suggestion.

Once you see the variance, you can decide on a course of action for your needs.
• How long is the straightest rip cut your projects requires?
• How much variation would you observe on the longest project part cross-cut from this 8 ft ripped length?
• Decide if the variation actually matters over a 3 ft or 4 ft length of a required part – or even the width of a tall bookshelf side.

Pick your battles and please share with us what you find.
Hans

 

[edwarmr]

Thank you! [member=7493]Sparktrician[/member] [member=58857]Crazyraceguy[/member] [member=61691]TSO_Products[/member]

I appreciate the information and the suggestions Unfortunately I don’t have a 6’ level. I don’t think I would use it for much other than checking this rail. In practical use the rail is probably accurate enough for me but I would like it to be better for the price. I think I’m gonna call Festool tomorrow to see what they say. Then I may make a visit to Acme tools to see if I can exchange it for a new one depending on what Festool says.

 

[mino]

I was posting this on other threads when comparing Makita/Festool/Maffel rails.

The issue/limitation is the anti-slip pads. They are so strong they can "fix-in" a lateral bend of the rail when you place it on stock not carefully enough.

This is one plus for the Makita rails - they are *less* accurately made but their anti-slip pads do not really work - so they cannot "grab" the rail sideways either .. but then one must use clamps most of the time. Nothing is free.

----
1) How this happens:
You place the rail approximately, then move it (against the resistance of the anti-slip pads) on one side to match your mark. Then you go to the other side of the cut and move it to match the other mark. Once done, you check both sides and both marks will match. All good, right ? Wrong! The tension the anti-slip pads create is sufficient to bend a long rail up to about 0,5 mm (0,02") in the middle relative to the ends. You do a cut and the cut is bent - like on the picture.

2) What is wrong with the rail ?
Nothing.
If you put a 2x180 mm aluminum extrusion into an engineering bend calculator, you will get a result that just placing such a long piece of metal vertically, there will be about 1/128" bend just from the rail weight. Is one of the reasons why it is not advised to store the long rails on their side on a wall supported on the ends (instead of at 1/4 and 3/4 length).

3) How to solve this ?
First, place the rail approximately, then do either of these two techniques for final placement:
a)
if you have to work from sides, lift the rail so that at least 2/3 of it is not touching the stock and then place it on your mark. Repeat on both sides until both marks match. NEVER re-adjust the end sideways if you placed it wrong. Not even a bit. That is how the bend is introduced. Re-lift the rail and retry until you hit your marks.

b)
if you can work from the middle, lift the anti-splinter side of the rail and move the whole rail from the middle again until you match both marks. This approach is practical especially with parallel guides.

Always keep in mind a rail is a precise measuring instrument. When manufactured it is about as accurate as a DIN Class 2 straight edge. That is why it costs so much.

It is however WAY more fragile (and elastic) than an engineer's straight edge. For reference, a 3000 mm engineer's straight edge from my local maker weights 60 pounds (and costs $900). And even that is guaranteed accurate only when there is no side-force applied to it..

Hope helps.

 

[edwarmr]

[member=61254]mino[/member]

Thanks for the additional input! I may have read that from you in another post so I was very careful with my cuts to lift the rail and set it down without dragging it. The rail was definitely off.

I called Festool service and while the guy wouldn’t give me a quoted tolerance he told me the rail should be straight. He addressed other common mistakes about not having both cams engaged at the beginning and end of a cut. He advised me to return it if not straight. Interestingly he also told me that the FS 2700/2 is not good for 8 foot plywood sheets even though Festool says is can be used with the TS 55.

So today I returned the rail and picked up the FS 3000/2 instead. I haven’t made a cut yet but I’m hoping this one will be straighter

 

[smorgasbord]

Always keep in mind a rail is a precise measuring instrument. When manufactured it is about as accurate as a DIN Class 2 straight edge. That is why it costs so much.

If my math is correct a 2.5m long DIN 874/2 straight edge has an acceptable tolerance of 0.01mm (seehttps://www.thome-precision.com/DIN874.html for instance, showing (8 + 71/40)um is about 0.01mm.

Typical extruded aluminum tolerance seems to be 0.0125" per foot. Converting to metric and over the entire length gives: about 2.6 mm acceptable deviation. This could be why off-brand rails aren't that great.

What do we know about Festool's extrusion process that gives it such high accuracy?

BTW, while you're correct about best way to support a horizontal rail, I would think that any sagging while stored falls into the elastic range at room temperatures, and so doesn't permanently deform the rail when laid flat on your sheet to cut.

 

[mino]

If my math is correct a 2.5m long DIN 874/2 straight edge has an acceptable tolerance of 0.01mm (seehttps://www.thome-precision.com/DIN874.html for instance, showing (8 + 71/40)um is about 0.01mm.

Got an order of magnitude typo there - the 83um specced for 3000 mm is 0,083 mm which I have rounded to 0,1 mm.

Typical extruded aluminum tolerance seems to be 0.0125" per foot. Converting to metric and over the entire length gives: about 2.6 mm acceptable deviation. This could be why off-brand rails aren't that great.

What do we know about Festool's extrusion process that gives it such high accuracy?
..

TLDR:
Annecdotal, but pretty strong evidence from me only:

----
When I was sorting out straightness of cuts, I have spent lots of time testing and measuring all rails I had on hand.

I ended the research when I got reliably to

 

[smorgasbord]

Thanks for the reply, [member=61254]mino[/member].

I do agree that straightness for rails isn't necessarily one big bend, but could be more localized "waves." That's why the DIN spec allows for choosing any subsection of the item and evaluating straightness in the "per foot" metric.

I also wonder about parallelism. When you think about it, it's the big ridge against which the saw actually rides that matters, not either actual edge of the rail. The cutting edge doesn't really matter at all, as it is there really just to support the splinter strip. The back edge, however, is used by rip guides and right angle guides, so one would want to be sure that edge is parallel to the part of the rail that determines the saw's path. I suspect that's good enough for woodworking.

Anyway, I do agree with you that over long lengths you wouldn't want a rail that was extruded to normal aluminum extrusion specs. It would be interesting if Festool would put out a little video on their rail manufacturing processes.

 

[mino]

...
I also wonder about parallelism. When you think about it, it's the big ridge against which the saw actually rides that matters, not either actual edge of the rail. The cutting edge doesn't really matter at all, as it is there really just to support the splinter strip. The back edge, however, is used by rip guides and right angle guides, so one would want to be sure that edge is parallel to the part of the rail that determines the saw's path. I suspect that's good enough for woodworking.
...

Yeah, I was checking that too ... measured within 0.03 mm, or so, from memory, between the rails I had and within 0.02 (within calipers accuracy basically) mm on a given rail.

It is definitely good-enough for parallel guides and such. The only limitation is the "wavyness" can affect non-calibrated items that place the whole rail, thus possibly multiplying the inaccuracy by an order of magnitude or more - so fixed squares or self-aligning rail connectors are the only things where one shall be conscious of it when working with an unknown rail.

 

[woodferret]

What do we know about Festool's extrusion process that gives it such high accuracy?

They reject it when it's that obvious, although some get through.  Also the 30-day return tends to cull the deviants.  I had a 800/2 that was bowed that 2mm - promptly returned.  It's why I only recommend in-store purchases for rails and/or anything that has a fence.

 

[mino]

[member=61254]mino[/member]

Thanks for the additional input! I may have read that from you in another post so I was very careful with my cuts to lift the rail and set it down without dragging it. The rail was definitely off.

I called Festool service and while the guy wouldn’t give me a quoted tolerance he told me the rail should be straight. He addressed other common mistakes about not having both cams engaged at the beginning and end of a cut. He advised me to return it if not straight. Interestingly he also told me that the FS 2700/2 is not good for 8 foot plywood sheets even though Festool says is can be used with the TS 55.

So today I returned the rail and picked up the FS 3000/2 instead. I haven’t made a cut yet but I’m hoping this one will be straighter

Great to hear it turned out well.

The rep was correct on the length. A 3000 (or semi-permanently joined 1900+1080+MakitaCon being my-preferred these days) is the way to go. I think FT should just drop the 2700. It mostly only confuses people.

IMO the 3000 and 2700 should both be dropped and just a single 2900 SKU replace them. Then there would be 1400-1900-2400-2900. Nicely each 500mm steps.
[cool]

3000 is a round number, but it is 5" unneeded for full sheet stuff and cannot really handle European 2800 chipboard sheets either.

 

[Crazyraceguy]

I would say that it likely comes down to money. $272 to be exact.
The FS 3000 sells for $743 and the FS2700 goes for $471. That's a big difference for 11.8".
It seems that they have gone over the top lately though? I don't remember what I paid for my FS3000

My paper catalog from 2016, says $331 for the FS2700 and $355 for the FS3000  [eek]
The last paper copy is from 2022 and the difference is quite a bit greater, at $355 and $435 respectively.
So, it jumps from $24 to $80  [unsure]
I think I bought mine in '20, probably somewhere between those figures.

While I totally agree that the FS3000 is a better choice for 8' sheet goods, that extra foot may literally not fit in your van. The $272 price difference is harder to justify too.

The 2022 catalog also shows the FS2424-LR32 at $379, currently priced at $429

I'm just glad I got mine when I did, I certainly wouldn't pay over $700....really $800 after tax
Theoretically, mine is a profit monster in hiding.