Underextrusion on UM2, new PTFE coupler not sitting flush with hot end isolator

[Lizziegill 0]

I've been using a ultimaker 2 at work for 18 months now, it's been modified to also be able perform pneumatic extrusion in addition to conventional FFF which has worked well. Recently however, the FFF feeder motor at the back keeps ticking/skipping and eventually grinding its filament down, which has meant I have had to take apart the feeder assembly a few times to remove trapped filament.

-Potentially the feeder tension could be wrong having been reassembled, although the tracks on the filament from the feeder look okay and having looked at various posts about this problem I have ensured that the dial on the side of the feeder assembly is pointing to the middle bar (I believe the printer was bought after the critical time when this change was made).

-The bowden tube seems in reasonable condition, I haven't felt friction when I've had to remove filament manually from it before!

-I think the nozzle is also okay, in my various trials before I have thoroughly checked that it isn't clogged.

-I just replaced the PTFE coupler and I have a couple of small obervations: the new PTFE coupler seems to fit inside the hot end isolator more loosely than the old one and I can't get the new coupler to lie flush with the hot end isolator. I am using an ollson block and if I fully tighten the ollson block on the hot end isolator the coupler is displaced and compresses the spring- see picture.

The gap between the isolator and coupler wasn't there before but I am wondering if this is a problem? I figure the main purpose of the coupler is to protect the bowden tube from heat and guide the filament to the hot end?

I tried reloading filament with the new coupler and it was fine initially but shortly after filament started being extruded I heard clicking from the material feeder again. After a quick test print I have noticed that the extrusion is fine for lower extrusion rates but feeder skipping happens for faster sections of a print and infill. This makes me think there is something restricting the filament feeding at the head still rather than the feeder being ineffective?

I've been using Simplify3D rather than cura to generate my gcode but the print settings I'm using are not wildly dissimilar to what I have used successfully before. Running out of ideas here, any other advice would be greatly appreciated!

Apologises for the lengthy post!!