Sudden Print Failure (Page 1 of 2)

I got a few more pawns printed correctly, having increased the extrusion multiplier from 0.8 to 1.2, but now even that is failing (in the same way as explained and pictured - insufficient filament extrusion).

I did first look for tips/help on the Solidoodle Wiki but did not find anything that seemed applicable. The closest described problem I found was the "hot clog" but, as I mentioned, I'd really rather not try to disassemble the extruder.

Start out by performing an extruder calibration.  you really need to command 100mm to see what the real variance is and then make the adjustments to the steps/mm in the firmware:

https://solidoodletips.wordpress.com/20 … libration/

Then you can dial in your extrusion multiplier, etc.

EDIT:  And if it is in fact a partial clog, the procedure above will help you to determine that.  When the extruder properly calibrated, I would be surprised if your extrusion multiplier exceeds 10% (.90 - 1.0)

Okay, I think I've broken the clog using the 1.5mm hex driver. I've also calibrated the EEPROM extruder steps per mm by following the instructions in the solidoodletips wordpress blog post "Extruder Calibration" on 2012/08/09; I ended up dropping the steps from 138 to 112-114. I've started trying to print the 30mm Test Cube by MaverickCNC on thingiverse (thing:193342), but if I set the extrusion multiplier any higher than .7 then the printer makes a clicking/popping noise which I'm pretty sure is coming from the extruder. However, it's *also* not extruding enough material.

http://imgur.com/gallery/bmedjwK/new

And what temperature are you using?  That sound you describe is most definitely your extruder stripping on the filament.  Make sure you have enough tension on the tension arm.  The process IronMan described is correct.  I am questioning your numbers though.  Call for and measure 100 mm's there is just too much room for error at 10mm.  The steps per mm should be closer to 105-107.  This is an independent calibration and calculation from extrusion multiplier and should be done first.  It will not change regardless of the small diameter differences seen from roll to roll of filament.
Extrusion multiplier is a function of Slic3r and therefor used to later fine tune the extrusion width.
And is only added to the extrusion process during slicing.
This is the next step in the calibration process and will change from roll to roll even if the filament diameter is exactly the same.  This is where you create different slice profiles in Slic3r for different types of filament.  Mark the extrusion multiplier and temp right on each spool as well.  For every roll the very first thing I do is print a single perimeter 20mm cube with no top.  I adjust and measure until I achieve .48mm wall thickness (if using a.4 nozzle). 
I am sure there are many effective ways to make a printer yield decent results but this is the only proper and well documented process that I am aware of.  It may seem like a pain and a waste of filament and time to do this for each roll.  I assure you it will save you time and filament in the long run.
More directly to your issue of the clog.  Watch the temp chart the next time and ensure the temp is stable.  If it is try running a little hotter to see if that changes anything. 
The harsh reality is you have clog and may never completely clear it.  If you can't get it cleared you will be forced to remove it and soak in acetone or torch it clean.  The problem will be upon reassembly and you most likely will have clogging problems again shortly.  If that's the case I would seriously recommend switching over to an all metal hot end such as the E3d.  It is well documented and supported on your printer making it a fairly straight forward swap.
I hope this helps

That is correct you set e-steps it is a function of EEPROM and simply overrides the set firmware values.  It is not a part of the slice profile so you won't see it on any of those tabs.
The extrusion multiplier is in slic3r and is set by printing a single perimeter and measuring until the actual width equals .48 as stated in my previous post.

I've tried using 202'C - 208'C for this. Normally I've used 200'C-202'C for printing.

I've taken off and returned the spring several times while trying to fix this. I currently have the end nut even with the end of the screw (as I believe I remember from ... something I read). Might the tension provided by that spring have changed? I could try tightening it to see what happens, although I would be concerned that if the clog isn't entirely cleared then forcing more filament into the nozzle will just make things worse.

I did command 100mm, measured using a straw as in the video in that solidoodletips blog post, twice. I did do that first before attempting to print the extrusion calibration cube (I appreciate the reinforcement that I understood the order correctly).

For the record: I have not seen any burnt/charred black come out of the extruder nozzle, although I might just not have been watching closely enough. What is the best way to know that it has been cleared?

About the E3D hotend replacement documentation, are you thinking of the E3D_extruder entry on soliwiki, or something else?

Thank you very much for your time and assistance.

I'm glad to hear that because it matches my current understanding.

Is there a best way to tell the printer to print a single perimeter? The only way I've encountered yet is to print a no-top cube or cylinder with a wall thickness of some target thickness. Is that the correct method? (i.e. should I create a cube with 0.48mm walls to print?)

Also, why 0.48mm when the default profiles use 0.42mm? Or have I misunderstood the "Extrusion Widths" under "Advanced" in Slic3r's "Print" profile configuration and that is actually a command of what it should be (aim for) rather than an input informing what it is?

When your done printing that the wall should measure .48 if not adjust the extrusion multiplier accordingly.

Yes, I've had mine since 2012 and this is the first trouble I've had with it, so I still have the 0.35mm nozzle. That doesn't explain why the e-steps are so far off though.

The slicing options on the otherwise solid cube is brilliant. So obvious after you know that trick - I love it. Thanks!

Okay. If I sufficiently increase the tension on the tension arm by screwing the nut further against the spring, then the clicking/popping noise stops. This also lets me repeatedly get 100mm of filament fed by 106 e-steps. I've since successfully printed out 2x 20mm single-perimeter wall extrusion calibration cubes.

The reason I printed 2 such cubes is that I found my Slic3r Printer profile had a 0.4mm nozzle while I was double checking. I'm pretty sure I remember ordering this thing with the .35mm nozzle. I think I may have increased it to 0.4mm so that the default 0.4mm Slic3r Print profile would work (since otherwise Slic3r crashes every time that the layer height is greater than the nozzle diameter) without having thought through why that number might be used. So I ran an extrusion calibration cube once each with 0.35mm nozzle + 0.42mm width setting and 0.4mm nozzle + 0.48 width setting. The 0.35mm nozzle settings produced a cube with 0.44mm walls whereas the 0.4mm nozzle settings produce a cube with 0.64mm walls (trying to carefully measure only the top 2-3 layers). I'm going to try sticking with the 0.35mm nozzle settings because the results look closest to what they should be.

However, the cubes do not look nearly as good as the test cube on cisana.net ("Extruder Calibration" Posted on August 16, 2013 by flurin). Both of my cubes start each layer under-extruded such that the "first" side is full of length-wise holes due to the extruded filament being too thing and spindly/wispy. The second and third sides look okay and the fourth side looks passable. I assume that is because the filament retracts and returns in between each layer. Is there a way to improve upon this, or is that a result of the likely-still-partially-clogged nozzle?

Very good; I'm glad I understood that correctly. So, since I measured 0.44mm width (average between the second, third, and fourth sides) I've reduced the Slicer Filament profile extrusion multiplier from 1.0 to 0.95 (0.42 target divided by 0.44 actual) for this particular spool of filament.

Unfortunately, the build quality is still extremely poor / sickly / problematic:
http://imgur.com/gallery/OLhBh9K - small hollow pyramid

http://imgur.com/gallery/vOgAAuN - 15mm solid cube (20% honeycomb infill, 3 perimeters horizontal and vertical)

Also want to mention that there are several other ways to attach an image that are more friendly than Pinterest.  So if you'd like me to be able to view them just attach the file or use imgur.