I took the unit apart (i have since redesigned it with thumbscrews and detatchable electrical connectors to ease the pain of needing to constantly disassemble it) and found once again that a solid plug of white plastic was stuck freely spinning on the auger.  I decided that the problem had to be component related as I have tried everything else.

I decided to make a new body from an 8-inch pipe nipple.  It took a little work as I needed to cut out the feed opening with a bandsaw and file down the ridge seam inside the pipe with a rat tail file but it was worth it.

After doing so and getting the auger to fit, I was immediately struck by the change in fit.  The original pipe always fit very loose around the auger to the point that a fair amount of pellets would get ejected out the back of the auger during operation.  I thought this was normal at the time, but the new one allows free rotation but isn't nearly as loose.  It has run all night and produced some nice looking filament (without leaking pellets), so I am hopeful the problem is licked.

Between colorant ratio issues,  undersized nozzles, and constant clogs, it's been a long journey.  But I am hopeful now that after a year of (quite) intermittent tinkering that I have my first good batch of quality, printable filament!

Some closure on this - tealvince just emailed me:

After it was all assembled I tried to match up the spacing as closely as I could remember it was with the old bearing and backing plates.

I would have run all natuual for the best chance of cleaning out anything that was left stuck to the walls. There is a correlation with using certain colorants (or too much colorants) and experiencing what you describe.

After blowtorching, did you soak in acetone? That would be better than blowtorching for removing polymer stuck on the barrel.

When you reassembled, how much of the auger was sticking out of the barrel?

After blowtorching, did you soak in acetone? That would be better than blowtorching for removing polymer stuck on the barrel.

When you reassembled, how much of the auger was sticking out of the barrel?

What colorant? The "most vanilla" setup would be no colorant.

It was the black OSP colorant that I've used numerous times before.  In the past, I've used a much higher concentration, but I backed it off to 48:1 just in case.  True, using no colorant would have been even more conservative, but I wanted to test a configuration that still represents what I normally need the unit to do.

New barrel/auger is certainly an option, you can email me for one.

So what's next?  Should I order a new barrel?  Auger, both?

What was the recent colorant? Just trying to collect data for future users.

Last night, it jammed again, after extruding only about 10 feet or so of filament.  I completely disassembled the unit this morning, and noticed that there seemed to be some residual plastic stuck to the inside walls of the pipe, perhaps providing a smooth surface for the plug and auger to spin inside.  I did recently try using a new brand of ABS colorant.  Maybe some of it baked onto walls, causing problems even after I switched to no-colorant or old OSP colorant.  I burned it off with a propane torch.  Hopefully that will solve the problem.  While I can't say I've made a lot of filament, I've extruded a few spools worth with no problems of this sort whatsoever, so I'm hopeful this was the cause.

Here's what I know on users with similar problems:

Extrusion totally stopped:
elmoret (v1.3) - Problem happened after ~1500 hours of use, 95% MG94, 3% other ABS, 1% PLA, 1% ABS+black or green colorant. Prior to the clog, nothing my MG94 had been run for several hundred hours. Disassembled, removed plastic plug, reassembled. Problem has not resurfaced for ~200 hours since, extruding MG94 at temperatures from 180 to 190C. In the link, I detailed everything I observed - including the fact that after the barrel cooled, I could rotate the auger with a wrench - indicating the plastic was not even touching the walls of the barrel anymore. The jam occurred after heating up from cold - not after running for a period of time.
IanJohnson (v1.0) - problem happened after running Asaclean. Replaced barrel, problem did not resurface.
insta (v1.0) - problem happened after ~100 hours (5kg). Ran nylon pellets which seem to have pushed clog out.
digitydogs (v1.3?) Ran MG94 ABS and PLA for 200-500 hours, several jam/clean cycles to no avail. Two machines exhibit this behavior, two do not. One of the two exhibiting the jam ran GITD filament right before the jam, one of the machines that does not exhibit this behavior has run 50lbs of GITD filament.
kitcarguy (?) Ran GITD filament in the first few hours of use, jam occurred. Disassembled and cleaned twice, problem went away.

Very slow extrusion:
Ggalisky (v1.5)- Experienced 1-4"/min with MG94 and GP35 at 1.75mm diameter. Went up to 16"/min after clearing the clog, no issues since.
ericpaulbishop (v1.4) Experienced "dramatic slowdown", found that higher temperatures triggered it sooner, found that wrapping the barrel helped.

Slow extrusion, seemingly unrelated to this problem:
spencermatousek Experienced 6"/min at 1.75mm filament. Changed to 185C extrusion temps, jumped to 12-14"/min (normal). Probably a case of thermocouple/controller not well calibrated.
borreas Experienced 7"/min at 1.75mm filament. I made some suggestions, never heard back. Not sure if solved, but 7"/min is not characteristic of what we're discussing here.
wahada - experienced 4.5"/min at 1.75mm filament, in vertical configuration. Problem was poor pellet feed/hopper design/configuration, as confirmed here: http://www.soliforum.com/topic/6976/ver … er-shaker/ he is now achieving 30"min.
thormj Nozzle was blocked with debris, cleaned nozzle, problem went away. Now "very happy".
Hargert Barrel was not tight enough and rotated durign operation, blocking pellet feed.

That's what I'm aware of so far. That's 5 users with complete stoppages, and two with very slow extrusion which may be related to the same issue, for a total of 7. I also added some users who had troubles that were resolved by other issues - pellet feed, temperature, debris in nozzle, etc.

What I'm finding is that the filament output frequently slows to nothing.  When I disassemble the unit, I find that there is a solid block of plastic around the 4" ed of the auger that I presume is just going round and round instead of pushing out the pipe (though I would expect the plastic to be there normally as well).  When I remove it, the unit functions again, typically for 10 feet or so before clogging again, though sometimes it's gone longer.  When it clogs, it won't "unclog" no matter how long I've let the heat "soak in", tweak the temperatures, etc, but in between "clogs" it runs like a charm.

When my unit first failed, I noticed that there were gouges around the auger at the heated end.  I wasn't sure if this was normal, as I hadn't disassembled it before, but the auger isn't supported at this end either, so I wasn't surprised.  I'm thinking that maybe the end of the pipe has gotten "polished" by contact with the auger and that this is perhaps causing the clogs?  Has this happened to anyone else?

Last night, however, after finally getting a new locking pin to hold briefly, the auger cracked the two plywood (I read recalled somehow?) backing plates, destroying the thrust bearing in the process.

I'm now in the process of rebuilding the unit, and perhaps upgrading components to prevent this from happening again.  The auger has some gouges, but I hope is still salvageable, and I'm planning to replace the backing plates with a pipe flange and a heavy duty fender washer and thrust bearing.

I'd like to replace the shaft coupler with something more precise, if possible, and with mechanism whose failure point won't change if it slips.  Are there linkages designed to give way that are resettable or easily repaired (e.g. with a new shear pin) that are readily available?  If so, what torque rating should I be looking for?  For that matter, what does the thrust bearing need to be rated for?