Michael Kellough said:
I’ve been joining Festool rails for over twenty years. Getting the rails straight requires several steps and use of a straight edge. The Makita connectors are more secure and the TSO connectors make it a bit easier (haven’t tried the latest Festool connectors) but the Mafell/Bosch connector only needs to be tightened (with a simple coin) to make the rails straight, within the limits of the rails themselves. And you don’t have to flip the pair of incompletely joined rails over. That is superior to the other methods.
Also, the Mafell connector is designed to make the spines automatically line up as closely as possible. And since the spine is about a tenth as wide as the Festool spine (and others) the typical slight differences in spine width are negligible on the Mafell rail. You don’t have to put the bigger spine first in the pair and possibly readjust the jibs. I’ve never had to make that adjustment on my Mafell MT55 cordless.
This is a bit OT. But I will bite.
TLDR:
The limitation I noted above comes from how when a rail is manufactured, it has variance/bending along its length but it is
not a single "consistent" bend. Instead, as it is rolled there are multiple bends "back and forth" while the whole rail is made to spec by "oscilating" within a certain distance from an ideal straight line. This directly impacts any accessory - be it a TSO GRS or a connector - which relies only on a short section of a rail for alignment.
-----
Now, lets say a "deflection" of 0.1˚ (for the sake of an argument) is present at the start/end of a rail. As long as this deflection is compensated later on, and does not go beyond tolerances, all is good. Now, imagine a connector is put into this deflected part of the rail (or a TSO rail square, same effect). What happens is now this deflection is
multiplied by the ratio of the connector-interface to the connected-rail length. In practice that is 1:8 or so for the casual 1400/1600 rail lenghts. Same effect happens on the other side of the connection. So, worst case, from two rails which are very much "within spec" we get a deviation that is potentially 16x as much. And, now comes the catch, it is not possible to compensate-out this deviation when using a self-aligning connector - it is fixed/defined by the shape of the two rails we have. The catch is, what if we have two rails whose deflections are either not present, or cancel themselves out? Well then we have a win! So two users can observe a completely different behavior! And unless they have multiple rails and dig into this .. like I did .. may not even realize this!
With a non-self-aligning connector, were are not dependent on the short parts of the rails - we can use a pretty long rail as a reference, even as long as the joined rails in theory - thus getting comparable accuracy/deflection as a single-piece rail gets. Now, this benefit is not free - it requires correct alignment every time the rails are joined. But the alignment is possible. Unlike with a self-aligning connector setup where we just hope on winning the "rails lottery".
I came to be intimately aware of this when my TSO GRS was producing inconsistent results with different rails. I found out that while all my rails were absolutely fine for when used with a PG or with pencil marks, once a rail square - relying on just 8" of the rail back - was used, I got up to 0.01" deflection at the end of a 3' cut. Fine for wood. Resulting in an unusable stock for laminated chipboard as the gap was clearly visible. Hence my earlier posts praising the FS-WA as that allows to calibrate/align-out this limitation.
Recap of sorts
------
As far as self-aligning connectors go, I agree the Mafell option is great and definitely the most convenient. Funtionally the Festool self-aligning connector is indeed inferior (unless used with the second connector from Makita for strength). However, this assessment is valid only within the realm of self-aligning
connectors. Once a whole rail system is concerned, where the (im)precision in the rail manufacturing as well as the rail stiffness come to play, it is not so simple.
In specific, two effects turn the tide to the FS/2 system benefit in some use cases:
- first effect, or a limitation, of the Mafell system: Self-aligning is the
only way to connect the rails. There is no option of defferring to non-self-aligning rail joining, should absolute accuracy be required.
- second effect, or a tradeoff, is that the Mafell rails are narrower and thinner. This makes them more comfortable and easier to work with but at the same time limits their lateral stability at longer configurations - the 2x1600 is one of those configurations. My view is that Mafell went for only self-aligning because when their rails are joined, they are already a bit wobbly laterally. So there is not much point of having a super-precise connection option. For the same reason making a 3000 rail is pointless as well.
The overall conclusion being - there is no
universally superior system. And there could not be. The optimisations Festool chose and Mafell did are mutually incompatible. There is no way to make the Mafell rails as stiff as the FS/2 ones, without losing their convenience. And there is no way to have a quick-connect system as convenient as the Mafell one, without losing the backwards compatibility of the FS/2 system. And that is fine. The fact we have these two systems is great, as one can have both and use for different tasks. Or chose the system better suited to one's workflow.
[smile]