intermittent extrusion by hotend

My latest iteration of a hotend with heated block has a problem of intermittent extrusion (PLA) at any temperature (185 to 200C). I badly need some hints on how to resolve the issue.
I noticed that it extrudes excellently for some time, then stops extruding, then starts up after some more time, and so on. The end result is no prints.

Here are some additional details on how I built my hotend.
I drilled a 3mm dia hole in a 5/16"dia (1.5" long) brass bolt (brass toilet bolt kit from Home Depot). The bolt is 2.25" which I cut short with a dremel.
I angled the nozzle end of the bolt, and drilled 0.35mm hole of length 4mm from the nozzle tip.
Heated block (uses Aluminum encased resistor for heating) sits close to the nozzle, followed by the wood structural support, followed by a structural bolt. Then follows a few copper sheet metal squares for cooling the brass, followed by a 1" long PTFE that screws into the brass bolt.
I am using a CPU fan to cool the coper sheets to avoid plastic ooze from the wrong end. I enclosed the heater block with several layers of kapton tape to avoid cooling.
Image of hotend [forums.reprap.org]

Any suggestions to debug this issue without physically dismantling the hotend ?

Edited 2 time(s). Last edit at 09/11/2012 07:15PM by peddiparth.

just a quick idea, is the hot end cooling off while extruding?

The hot end is holding the temperature steady at the set temperature with only 1C variation. I have to figure out a way to measure the nozzle temperature. (probably with a thermocouple)?

I also noticed that I can not retract the whole filament out. This indicates there is a melt zone between brass and PTFE.

you'll find there is probably a gap between the brass and the ptfe, an easy fix for this is to very slightly over tighten the brass barrel, then redrill the ptfe part where it will get squished with the barrel, this way the transition between the two surfaces isn't met with a gap,

I am guessing that this intermittent extrusion issue is related to a blockage in the brass nozzle that stops pla from melting & pressure build-up for continuous extrusion, or the presence of a melt zone (solidified?) between brass and PTFE. I will be testing these two conflicting theories shortly by opening up the hotend.

The issue is caused by the viscosity of melted PLA sticking to the 2" long brass that I was using for the nozzle. I cut it down to 1" and it works like magic.
One other design change I made is to use a stainless steel coupler between PTFE and brass. I also added high temperature PTFE tape on the brass threads to avoid leaks in the coupler.