A conundrum

So, I have started having some issues that appeared recently.

Prints fine for quite some time and then, at some point, stops extruding.
Extruder temps are fine, never fluctuate really as far as I can see.

I have tried a different spool to make sure it was not filament specific and played around with higher extruder temps to make sure it was absolutely in the melt zone.

When this happens the hobb will eat out a section of the filament trying to extrude it and then stop extruding since it is now no longer grabbing the filament..
I even swapped between the stock hobb and two other differnet style ones that I have, no joy.

I am leery now of doing a longer/more complex print as I have tossed an awful lot away

Smaller prints are non-issues. Since this occurs a hour or two in it is difficult to troubleshoot...

 

OK, well... Solved I think. Sort of.
I printed this today and to simply get this to even print a lot of tweaking beyond what we had done already.

http://www.thingiverse.com/thing:243994

For this to work your temperature must be set really well, your level must be bang on and all the settings require adjustment (speed is critical). Took 5 tries to get it dialed in (most aborted well before the 50% mark).

 

I apparently have not 100% solved this either.
I have had prints that run for 4-5 hours, pull the head apart and there is no evidence of grinding with the hobb. No scrap in it, nothing...clean. Others after only a couple of hours will start grinding away and cease extruding.

My candidates for this so far:

1) something is causing the hot end temp to fall/drop and suddenly the extruder is falling down.
2) some imperfection in the filament is making the extruder struggle to push the filament through and eventually fail
3) Hobbed bolt issues
4) ????

The temp graphs in Repetier do not in these cases show any variations in nozzle temp so I think #1 is out the window. Different spools have done this so #2 is also less likely. Just not yet sure what exactly is the issue. I have had 14+ hour prints work fine and then a 4 hour print die. I have tried three different hobbed bolts (factory supplied one and two different styles bought elsewhere) so I suspect that #3 is not the case either. Making it print hotter does not make this better which is another reason I think that hot end temperature is not the issue. If I print way hotter than needed then the temperature would have to fluctuate a LOT more (which is not seen) and yet, sometimes it occurs.

The print that worked (above) ended up taking too little time I suspect because the 3rd one of them I printed after assuming it was fixed failed at the last dozen layers.

 

Aggravating, this is.

If it would consistently fail on longer prints that would be a symptom, but no.
If it would consistently work (or fail) with certain filaments or models, that would be something (but--no).

grrrh. I will figure it out eventually I guess.