Extruder Update

The new extruder parts have arrived and I have built up my new extruder. The build was uneventful, which was nice. Initial tests show that it works reliably, but I still have to dial-in all the settings again as the nozzle appears to be marginally bigger than the previous one. Once I've got it dialed back in I'll be able to continue experimenting with the various skeinforge settings.

I'll document the dial-in process but there is something that's going to interrupt that process. An upcoming house move that is going to disrupt my workshop for a few weeks while it gets packed, moved and unpacked.

PreprocessGcodes.py

import sys


# Preferred Start Codes
startCode = [
'G21\n'
'G90\n'
'T0\n'
'G28 Z0\n'
'G28 X0 Y0\n'
'G92 E0\n'
]


def AppendCode(fh, codes):
   for code in codes:
      fh.write(code)




def Main(fileName):
   # Counters   
   M101Cnt = 0
   M103Cnt = 0


   # Open Input File
   try:
      fh = open(fileName, 'r')
   except:
      print 'Cannot open file: %s' % (fileName)
   
   print '\nProcessing %s' % (fileName)
   
   # Create Output Filename
   name, ext = fileName.split('.')
   name = name+'_mod'
   opFile = name+'.'+ext
   
   # Open Output File
   op = open(opFile, 'w')
   


   # Add preferred Start-up code
   AppendCode(op, startCode)


   # Remove first X lines
   # Skeinforge produces strange start-up code that isn't useful.
   X = 12
   for i in range(X):
      fh.readline()
      
   # Process remaining lines in turn
   for line in fh:
      # Remove 'M101' and 'M103' commands, Mendel doesn't use these.
      if line.find('M101')==0:
         M101Cnt +=1
         line = ''
      if line.find('M103')==0:
         M103Cnt+=1
         line = ''
      # Remove 'M113' command, stepper based extruder doesn't use this
      if line.find('M113')==0:
         M101Cnt +=1
         line = ''
         
      # Write to output file
      op.write(line)


   print 'Generated file: %s' % (opFile)
   print 
   print 'Removed %s old M101 Codes' % (M101Cnt)
   print 'Removed %s old M103 Codes' % (M103Cnt)
   
if __name__ == '__main__':
   Main(sys.argv[1])

Rebuilding the Original Heater

I've rebored the PTFE to remove the notches I previously made on the inside of the barrel. It now has a diameter of 5.5mm and is completely smooth again. I've remounted it on the extruder base and attached the nozzle.


I've also rebuilt the original nichrome wire heater wrapped around the nozzle, just in case the new heater block I've made is causing the problem with molten PLA backing up into the PTFE barrel.


The heater part works fine, unfortunately the nozzle is still filling with molten PLA. Since it never used to do this before the original catastrophic failure of the PTFE barrel that this is the cause. I'll need to make a new barrel.


The problem is I don't have a lathe, so do I spend $95 on the techshop lathe class so I can use their lathes and then pay $125 for the one month membership, or pay ~$500 for a mini-lathe for my workshop. 


I'd like a lathe, but I doubt I'd use it again after this. Perhaps Santa will bring me one?

Extruder Problems Continue

After rebuilding the extruder and running some test prints the extruder is still having problems. After about 10-minutes of printing the filament stalled again. The PTFE of the nozzle had backed up with molten PLA again.


I think I'll try two more things tomorrow. Firstly, I'll rebuild the extruder with the original nichrome wire heater - perhaps the block heater is more suited to the new style extruder where the brass doesn't get inserted as far into the PTFE. I think that perhaps because the brass nozzle extends into the PTFE 15mm or so, it is transferring heat further up the insulator than with the new style extruder.


Secondly, I'll try boring out the PTFE insulator. There is a slight notch on the inside where the drill slipped when I was drilling out molten PLA previously. This notch appears to be catching the molten PLA, which could be causing the feed problems. This will increase the internal bore diameter slightly which hopefully won't cause any problems.


If none of this works, I'll be rdering some PTFE and joining TechShop and using their lathes.

Building the Block Heater

As part of rebuilding the extruder I thought it would be a good idea to build the block style heater, rather than than the original heater. I made a couple of modifications to the build, we'll see how they work out.

The basic components of a block heater.

I followed the mechanical drawings and plans on the RepRap Site. It did require me to buy a drill press, metric drills and a metric tap and die set, but I do enjoy having new power tools!


I used a 3/8" ultra-machinable brass bar from McMaster. Cutting and machining it was fairly straightforward. The new drills and drill press did their job well.


I haven't tapped anything since I was at school, but the technique came back quite quickly. A quick test fit showed that everything went together easily.


I used 1mm PTFE tubing I found at Frys to insulate the leads for the thermistor.


To increase the thermal conduction between the thermistor and the resistor and the block I used Biostar Nano Diamond thermal grease. It is good up to 240c and is an electrical insulator. I filled the thermistor hole with the grease before pushing it home. The resistor was coated with the grease before sliding it in to its hole. 

With all the parts inserted.

While I was shopping for the fire cement I found a Rutland caulk good to 600F at Ace. It doesn't require heat curing and is flexible. I'll give it a try and see how it works out. It was a little messy to apply but should be better than fire cement.


After wiring it up and waiting for it to cure first tests appeared to be promising. It heated up quickly to the specified temperatures.


First tests with filament did not go so well. The filament did not feed very well and the output from the nozzle curled up! - Not sure why?


Then catastrophe struck again, the nozzle appeared to seize up. It appeared that molten PLA had backed up the PTFE and solidified.


It was another late night dismantling the nozzle and cleaning it out and drilling out the solidified PLA out of the PTFE. After it was all clean I checked the nozzle opening in case there was something stuck in there causing the curling.

Complete and back on the extruder. Note the new sprung loaded mech.

Using another thermistor as a temperature probe inside the nozzle it I checked the temperature inside the nozzle. There appears to be a significant difference between the block temperature and the nozzle temperature. The block has to register between 250 - 260c for the internal temperature of the nozzle to register between 205 - 220c. This may have been the cause of the feed problems. At a block temperature of 220c the internal nozzle temperature was only at 185c.


Cranking the temperature up to 260c and trying some manual filament feed tests shows this temperature setting was much better. Tomorrow I'll try rebuilding it completely and testing it with the motor feed.

Fixing the Extruder

It looks like it might be possible to salvage my current extruder. I have successfully managed to clean out the brass nozzle by heating it up with a butane torch and pushing the stuck filament through with a drill. It's currently going through a final cleaning by spending the night in a glass full of acetone.


The PTFE insulator was a bit tougher. Due to my earlier problems with the hot-end, it is well glued into the extruder base and it is half-full of PLA that followed the nozzle down as it was pushed out. I managed to get this clear by screwing the nozzle back in as far as it would go then heating it with the torch, then screwing it in some more. This pushed the excess into the nozzle and I could pull it all out with the nozzle.


The nozzle appears to fit still, so I'm not sure why it decided to drop out during the print process, unless the PTFE deformed. I plan to wrap the nozzle in a few layers of PTFE tape and screw it back in, then re-build the whole extruder again. But first, I need a new thermistor, it got destroyed whilst trying to clean the nozzle up. I'll check Fry's tomorrow or place an order with Digikey.


However, I doubt the extruder will last much longer after it is rebuilt so a complete new extruder is on order from Mendel-Parts with the PEEK style nozzle. However, judging by the latest update message from Camiel on his site, I doubt I'll see it before February. Hopefully his assistant is alright, it sounds like they had quite a week!


I also plan on building the brass block style heater instead of the wound wire version. It will give me something to do while I can't print. It's also a great excuse to buy a drill-press. I found that the kapton tape on my current extruder would melt gunk into the objects that I was printing.


I'm wondering if this is an indication the extruder was getting too hot? When I rebuild it I'll put the thermistor slightly further up the nozzle, rather than close to the tip.


Fingers crossed I can at least print something for the next few weeks, otherwise I'll be writing applications or playing with CAD packages.

Extruder Problems

The most frustrating thing about my Mendel is my extruder. I've had lots of problems with it since I completed the build. It's Adrian's Geared extruder with the hot-end epoxied into the base rather than held in by a PEEK block.


The first problem that I had was caused by my wimpy stepper motors getting too hot. The excess heat caused both the extruder base and the X-carriage to distort. This has since caused all sorts of meshing problems for the extruder gears as they no longer line up. 


The second problem that I had was the hot-end dropped out completely while working on a print. Frustrating, but easily fixed by liberal application of more epoxy.


Today I resolved to fix the meshing issue which causes the gears to skip, resulting in poor extrusion. I completely dismantled the extruder body to fix the alignment between the motor mounting plate and the extruder base. After they had become distorted they developed a slight kink in both parts, and whenever I tried to align the gears the parts would prefer to slip back to align with the kink. 

Note the kink in both parts and the buried washer.

The solution was to file the kink flat on both parts and make up for the missing material with an M4 washer. I also found that the excess heat had partially closed the bolt slots on the base, preventing the bolts from sliding as needed. After opening these back up again I could position the motor so that both gears engaged successfully.


Several tests showed that everything was working perfectly, so back to printing. In this case some new frame vertex with feet, since my Z-axis pulleys stand a little proud for the current feet to handle.

Previously the wires were flush with the PTFE.

Unfortunately, half-way through the print another extruder problem developed. The threaded brass insert that makes up the extruder was pushed out of the PTFE block. Shifting it down about 5mm and leaking molten PLA out over the nozzle.


I'm not sure if I can salvage this extruder hot end, I might need to build a new one using the PEEK block to attach the nozzle to the PTFE. This should solve at least two of the extruder problems I've experienced in the past two weeks. It wouldn't be so bad if I've been churning out hundreds of Mendel parts, but I've only managed to print a couple of small objects and two vertex's. Reading other RepRappers blogs, I can see that others have far more success with their extruder than I do with mine.


Oh well, time to research new hot-ends!