@Shauki this one’s for you.
I think if I’d been aware of this aspect of ‘rolling your own printer’, I might have stuck with a more developed kit…still, it sure has been educational!
I had been thinking of this since my Misumi arrived but with a v-roller bearing.
Boy do I know that feeling - but the parts are all reamed - fit and put together neatly now - all I need to do is sort out Repetier - or Marlin - or just make something work…
@Doug_Rector I’m in the same boat as you right now. Have everything set up and wired - need to start playing with marlin to get some movement and then calibrate.
Could you use the slots in the extrusion as a guide for bearings?
I’m pretty sure you could…these were just bearings I had laying about and not the right ones for the Job, @Shauki also surrounds the bearings in PTFE tubing to reduce the potential for wear.
Probably 
I’d assumed the bearings were the only things supposed to do the touching…some PTFE sheet in there could be desirable to take up any misalignment and reduce friction.
BTW, the Z axis drive was weak…swapping and adjusting got that taken care of. I still have a creeping long-term issue that presents itself 4-5 hours into a print (yet another failed celtic skull, I’ll post another thread on that.)
@Shauki I’m not sure that the self lubing qualities of the igus filament are present in raw filament.
perhaps you could press the igus filament between the printer frame and a hot bar of aluminium / steel and somehow have it set as a thin film over the surface, possibly some trial and error with adhesives to get it to stick but, could be great for what you guys are trying to do.
If we’re using ptfe as the sliding surface, would we need bearings at all? ( some deltas get by with just sliding smooth PTFE like lubricated surfaces - I forget which…there are just too many out there to remember.)
Well, there’s theory and practice, I wonder just how bad it would be in practice? (There world being a sloppy, analog place, and quadrap attempting to be a found-object, cheaper is better, other designs are hopelessly over-designed kind of printer…) I also thought of this as a source: http://en.m.wikipedia.org/wiki/Thread_seal_tape
@Mike_Miller There is always a trade off between theory and practice but my father (a P. Eng. and a farm boy) told me once ‘Good enough is the enemy of best !’ I think we should do the best we can with what we have. @Shauki You often mention your ‘water stepper’ … what is that project ?
@Shauki - in some CNC machines where it is desirable to drive from the center of gravity, sometimes I have seen designs being symmetrically driven from outside of the center of gravity (like the ultimaker style gantry, x and y are driven at BOTH ends, rather than the center)
Could you incorporate this into the cross bar design? Add another bar in the center? Adds weight, but gives the opportunity to have equal forces either side of the center of gravity?
If that doesn’t make sense I can probably knock up a sketch 
Bearings are expensive or relatively hard to come by…thinking about the offset moment issue, I wouldn’t be opposed to adding weight to the top of the printhead assembly. A couple of nuts at the top of the peek tube, with everything balanced about the axis of travel might slow things down slightly, but if it cuts down on parts list complexity it might be worth it.
One of the things that the kits do, and frustrated the crap out of me was: I’d find I’d need one specific part to continue the build (3’ of PTFE tubing, or a hobbed bolt, or Universal Joints, or 8mm rods, or, or ,or) I always ended up with 8 times the amount of what I really needed, and they were things you couldn’t just walk down to the hardware store to get.
While I see the use of all-thread was a major design decision in the first repraps, I’d think avoding that for the Z-axis at all costs would improve the printer…potentially using fishing filament and a counterweight to offset the built platform…then it’s the microstepping of the servo that dictates precision, not a coarsely cut thread.
The counter weight should achieve a little less than the build table so it keeps tension on the support strapping and drive gear/pulley