Laser Cut Parts (Bits from Bytes)

What is the issue with the acrylic gears? Do they wear out or break?

---

[www.hydraraptor.blogspot.com]

Cast acrylic is REALLY brittle. I shudder to think of trying to use a gear made of it.

Tonys done many many hours with his and we don't use cast Acrylic but extruded (better thickness tolerance and strength) but we have now switched to Hi-Impact much stronger agian.

---

Ian
[www.bitsfrombytes.com]

After having a lot of trouble with my moulded BfB extruder, I'm having a lot of trouble with my BfB extruder. It could be because of me, of course

I'd like to share this so we can discuss this and maybe fix problems you may be having and some that I'm still having.

Anyway, the extruder is hard to get right. If anyone has a well working (laser-cut) extruder, please tell me what you did!

My biggest problem which is probably the root of all evil, is the pressure/resistance the filament has (I'll get back to that). I've been striking at the branches for a while:

I'd like to share it and see if others may have the same problem.

First, the filament kept slipping. I removed the reaction washer that was held by little plate on the side. I made a new threaded rod by cutting it out of a stock 8 mm bar. Interesting detail: This required a lot of force and the rod became very hot. But I got very sharp thread in the process.

I poor-man's-lathe-d myself a 6 mm diameter section on the drive shaft of about 2 mm high at the very bottom. Around this I could put the thin ball bearing that only 8mm outer diameter (so not wider than the drive shaft itself). I now have a screw that can push down on the bearing and push it towards the filament, pressing the drive shaft against the filament. On the other side I can have the springs tightened all the way, but it simply doesn't help. I think my extruder can now deliver a pretty good push on the filament, and one that should easily be enough... but it isn't.

I used a thermocouple which should be pretty accurate. Room temperature and 100'C are right. Still, even when I go up to 230 degrees with ABS (I know, way higher than the ~ 110'C needed), it seems that the tip doesn't really become hot. Thinking of it now, I'm afraid the PTFE thread insulates the head from the heated tube too much. Do we really need that stuff to avoid leaking? The ABS plastic comes out hardly fluid.

I'm going to try it without it and see...

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

When I was fine-tuning my lasercut bfb extruder, I tested it by removing the nozzle and then running things as I thought they were supposed to go. This let me see if there was someplace else where things were getting held up. Once that was working fine (basically all the channels for filament were big enough and lined up), I made sure that the plastic was coming out completely melted at full width at my desired operating temperature and extruder flow rate. This was a little easier to test with HDPE than ABS since HDPE turns completely clear when melted and stays that way for a while. ABS always looks about the same to me.

One thing to try- I'm extruding ABS at 250c according to my thermistor- if I were you I'd try your existing setup at 240c and see if it flows better without burning the ABS (as long as it stays white I think you're ok on temp). If that seems at all better and it doesn't burn then crank the temp to 250c and see how that is.

ahan, thanks. A higher temperature helped. I have some experience with HDPE but not ABS, just that my old extruder broke under the pressure... perhaps I should've tried higher temperatures back then.
I put heat conduction paste (silver paste) in the thread where the nozzle tip attaches. Maybe that also helps. I think the tip may have been about the coldest part of the extrusion mechanism. Can't be right. I hope it's better now.

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

Erik, I'm still running my original cast extruder pretty successfully. I bought a lasercut extruder too to be on the safe side, but the old one is still kicking so I'm still using it.

One thing about heat conduction - if you want to keep the tip of the nozzle hot, there's two ways. You can reduce the thermal resistance leading up to the tip, as with the thermal paste. Or, you can also increase the thermal resistance outside the tip - ie, insulate it to prevent heat leaving the tip. If you get insulation down around the tip, past the thread, it should keep the nozzle almost as hot as the heater barrel.

Mine sure worked a lot better with insulation - I think the tip temperature is pretty crucial, as the viscosity drops pretty quickly, making the extruder jam or break. With ABS, make sure you're getting hot enough that the extruder drools 1.5 or 2 cm of filament just sitting at 230 C for a few minutes - then you know it's nice and hot. Don't try and run too fast either; with ABS, 0.6mm nozzle and a GM3 I can't go much beyond 3 mm^3/s, which is about what Nophead is running at too.

If you get it running, try using plexiglass as a bed material too - it sticks like crazy! CD's work in a pinch.

Wade

Thanks for the tips Wade. I did both, insulated the outside and increased heat conduction towards the tip. But thanks for being complete.

I'm glad your original extruder works well. It's a good sign that it can be robust.

Describing what your ABS does makes things more clear: mine definitely doesn't drule. I thought HDPE did this, and ABS didn't. But it seems the tip is still not hot enough. I definitely can't stick it well to plexiglass or ABS itself. So my temperature is still too low (while the extruder is pretty hot inside).

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

Another thought just occured to me - where is your thermocouple located? My old extruder uses JBweld, but I hear that the fire cement on the new laser cut version is a fairly good insulator. That will make the placement of the thermocouple pretty critical.

I have my thermocouple JBwelded into the nook between the nozzle and the heater barrel, separate and downstream (plastic flow wise) from the heater itself. The idea is to thermally couple the thermistor to the nozzle temp as much as possible, as opposed to the nichrome wire temp. Depending how much material you have between your nichrome and the heater barrel, you can expect the nichrome to be at a much higher temp than the heater barrel.

A good test would be to drop another thermocouple down the barrel and compare temps at startup, and again after a few minutes. I make my Darwin wait for 3 minutes at 230 before it tries to build anything, to make sure everything is nice and hot.

Alternatively, you could just keep turning up the heat untill you get some plastic drooling, and turn it off if it starts smoking. Remeber to go outside for a while if it does.

Also, I didn't put any ptfe tape between my nozzle and heater barrel. I think ptfe is a decent insulator too, although the thickness is so small it shouldn't matter.

Finally, you could try making a one piece barrel. I made one out of 1/4" brass rod, a hand drill, a file, and two drill bits. You can even give it a nice conical tip. Took me less than an hour to build it. It looks like it will be awesome thermally, although the hole is a bit big; 0.8mm was the smallest drill I could find locally. It would probably increase my build rate though, might have to finish it and try it out.

Hope that helps!

Wade

I am also having trouble with my laser cut extruder. It ran well for two hours last night, then it suddenly jammed and would not extrude anything. It jammed right in the middle of printing an otherwise good corner block in ABS. :-( After that, I could not convince it to extrude for more than a couple of minutes at a time.

This morning, I partly disassembled the extruder and connected the drive screw to a variable speed drill, simulating what the motor does. As I drove the screw back and forth, I was able to observe the small threads that the drive screw carves into the filament. Then I applied enough pressure to the filament to prevent it from moving, simulating a temporary jam. The drive screw was still able to turn when the filament was jammed, but instead of carving threads, the screw ground an almost smooth surface into the filament. Once the filament has been ground this way, the drive screw has nothing to work against, so the temporary jam became a permanent jam. This is exactly what happens in operation.

I've observed that the extruder only needs to be jammed for a second or two before it jams permanently. Now that I think about it, that makes sense: it should only take one full revolution of the drive screw to grind away most of the contact surface.

What can we do about this? For me, temporary jams happen whenever I extrude in a place already filled, plugging up the hole. The machine can't see that the place it's extruding is already filled, and the extruder motor is too powerful to react to the jam, so the filament gets ground up and the print fails.

Now I'm looking for ideas on how to solve this. Because of warping, I don't think I can completely eliminate the extra pressure caused by attempts to extrude where material has already been deposited. Rather, I think I need the extruder to be able to recover from short jams.

Some ideas I've had:

- Maybe the drive screw should have only 1 or 2 edges contacting the filament. Perhaps that edge could be razor sharp and point downward.

- Maybe there should be some kind of channel so that the filament has somewhere to slip rather than get ground up.

- I could heat the nozzle directly with a second, smaller heater.

- Maybe I really do need the extruder to be hotter. It's already hot enough to drip 1-2 cm of ABS when the motor isn't running, which is a good sign, but maybe that's not enough. When I set the heater to 220C, the hottest part of the heater reaches 250C and the nozzle tip is 140C or so. Is 140C hot enough?

> The drive screw was still able to turn when the filament was jammed,
> but instead of carving threads, the screw ground an almost smooth
> surface into the filament. Once the filament has been ground this way,
> the drive screw has nothing to work against, so the temporary jam
> became a permanent jam.
>
Yup, that's what happens.

I don't think you need to make the extruder recover from jams. You need to prevent it jamming.

The object should not warp enough to block the nozzle. It has to be held down by being stuck to a suitable bed material with a raft. It then only warps significantly after the build it finished and it has been removed.

140C for the nozzle seem very low to me. My nozzle is only about 20C less than the brass heater barrel. The heating element is a bit hotter. [blog.reprap.org]

---

[www.hydraraptor.blogspot.com]

Ok nophead, based on your recommendation, I optimized my heater. My heater is a lot like the bitsfrombytes heater; the main difference is I used an acorn nut instead of a custom nozzle since I have no access to a lathe.

There was a washer and nut between the furnace cement and the nozzle. I removed those, shortening the shaft about 10mm. The acorn nut is now touching the cement directly. I replaced the nozzle made of a steel M6 acorn nut with a 1/4" brass acorn nut (I used an M6 tap to force it to fit). I replaced the 30mm washer at the top with a thin stainless steel 1-1/4" fender washer, which transmits much less heat to the mounting plate above. I was really fortunate not to break the heater as I did all this. :-) I also wrapped the heater with fiber glass insulation.

Now my heater can reach 240C easily (it used to be limited to 230C) and the tip of the nozzle is at least 175C. The extruder seems to be in good shape again. It seems like my heater is now a lot like the heater you described in your blog post, so I'm not sure how you were able to get a temperature drop of only 20C. I don't have a lot of confidence in my thermal measurements, however. I'm considering buying one of those non-contact infrared meters.

As suggested earlier in this thread, I also removed the inner mounting plate for the bearing pushing against the filament. Now, when the filament jams a little, the bearing causes the screw to carve a little deeper. This seems like an improvement.

Thanks for your help.

I carved the pressure bearing window out a bit on my laser cut extruder too to get more travel against the filament, in addition to moving the lower axle support plate above the axle. Haven't tested it yet though; still using my V1.0 extruder.

As for those infrared thermometers, they're slick, but mine isn't much use for measuring the nozzle temperature. I have a Sinometer MTC300C which works great for larger objects, like motors and power supplies. However, the stepper driver heatsinks are about the smallest object I can measure reliably; anything smaller doesn't seem to light up the whole sensor, and gives a low temperature reading. Even if I have the face of the thermometer touching the part under test, which is kind of silly.

It's labeled as having a 10:1 spot size; but that doesn't seem to work well at very short distances. It was pretty cheap though; a better unit might be more accurate. I used to use a Raytek for troubleshooting locomotive turbos; it was sweet, but those turbos are big. I see Raytek are now Flukes - probably means they're good.

Wade

I have been measuring ooze from the extruder. Using HDPE at 240C, in the 10 seconds or so after the drive stops, I get another 20mm of thread. My initial experience is that the oozebane tool in skeinforge does not cope well with this long ooze. Is this a common experience with the BitsFromBytes Laser-Cut extruder? Is it worth trying reverse drive to the extruder?

Edited 1 time(s). Last edit at 11/16/2008 04:07AM by jonwise.

You could try reversing it, might help. My version of the Gcode Interpreter doesn't support reverse right now, but it's an easy fix. Adrian built a valve into his nozzle to deal with that. I'm just ignoring it for now - I leave Oozebane off, and just trim any "hair" out of the dead spaces on the printed object. The Comb function gets rid of most of the hair for now anyways... Oozebane was causing way too much plastic at the start of paths for me, and I was also having trouble with too many short move commands anyway.

I'm about to test out my lasercut extruder now that my Mark II has given out. So far the pump is pumping, the motor speed controller is controlling, and the heater is heating, just have to try it all at once now.

Wade

If you play with reversing long enough you'll notice that you can significantly reduce the amount of oozing. It is very deterministic but the amount of extrudate will always lag behind the inserted amount of filament. These two amounts will converge because of the pressure. It would greatly help quality-wise if XY speed depended on an extrudate flow rate model. This model should use filament entry speed, but with at least lag and slightly more sophisticated. Oozebane is the best we have now.
Remember that Oozebane is mostly to get the right amount of plastic in the right spot. As a consequence this reduces oozing and unwanted 'skein'.


My main point is that I think that adjusting XY movement rate to nozzle flow rate is much easier than the other way around.

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

> in the 10 seconds or so after the drive stops, I get another 20mm of thread
Applying reverse drive to the extruder for about 3 seconds, I get only 4mm of ooze. This feels worth adding to the skeinforge tools. Is oozebane the place for it or is there somewhere better?

Jon

About the lasercut extruder - got mine working nicely now. Built 10 or so small parts with it so far.

A couple tricks; I threaded the top of the heater barrel into the PTFE, and the put a hose clamp on it to prevent leaks. I had tightened the hose clamp too much though, and that was stalling the filament, so I ended up removing that.

Oops, just looked at the extruder and I've now got a plastic leak there, will have to add that clamp back in I suppose.

I also had to turn the temp up to 240 (measured by a thermocouple threaded between the nozzle and the barrel), and reduce the flow rate to below 2 mm^3/s; otherwise after a few seconds of extruding the resistance would strip the cut threads out of the filament.

I also sharpened the M8 bolt with a die, and then with a needle file, using a drill and one of of the bearings to spin the bolt as I sharpened it. That seems to be key.

A bit of grease quieted up the gear train too; not sure how that will affect the life of the gears though. Silicone lube is probably best.

Wade

Wade: I also have to keep the flow rate low to prevent jamming, although I think mine does fine with 2.5 mm^3/s. I've sharpened the bolt with a die but haven't tried the needle file yet; I'll do that if the extruder jams again. The lube sounds like a good idea.

It surprises me that others haven't mentioned breaking the delicate nichrome wire. I broke the nichrome several times, forcing me to start over on the heater, before I added a bar that prevents strain. I haven't broken the nichrome since. I just thought I'd mention that for anyone who might be having the same trouble. More details in my blog.

[shane.willowrise.com]

I'm still using most of the heater pictured in that post, but now the top washer is much thinner, the bottom washer and nut have been removed, the barrel has been cut about 8mm shorter, the nozzle is now brass, and I've wrapped fiber glass around the copper.

I've been doing Nophead's trick of embedding the nichrome connectors in the fire cement and using copper wire outside; so far so good. That bar looks like it might sink a fair bit of heat; my lasercut extruder already takes a long time to heat up as it is - roughly 6 to 8 minutes, vs 2 for the old Mk II.

I resurrected my Mk II extruder with new PTFE, some fire cement and an a thermocouple trapped between the nozzle and the barrel; now I've at least got a backup for when my lasercut extruder fails. So far I'm having a lot of trouble keeping the new one running.

I like your gcode compensation idea; might have to try that! I've also been struggling a bit with too much plastic in the short moves, but it's manageable. I've minimized my short moves and upped the comm rate to 57600.

Wade

I've been embedding the nichrome as well in every heater. However, whenever I worked on the heater, enough of the cement would always flake off to expose the nichrome to mechanical stress. Perhaps you are using a different kind of cement. Mine is quite gritty, but I like it anyway because it does not irritate skin.

The bar does get hot. I've been thinking that a bit of fiberglass PCB might serve better as a strain relief if it can handle the heat. In any case, my heater works well enough now.

I'm glad you have both extruders working again. The extruder is fairly challenging to build and I'm interested in learning whether some of the issues are specific to the laser cut design. You're in a position to find out.

I did find some better fire cement locally; its very strong and smooth. It's labeled Imperial Hi-Temp Stove and Furnace Cement, and it's good to 1149 deg C. We'll see how it holds up over time.

One thing I do is zip-tie the wires to the extruder a few cm away from the connection; I also bury the connectors far enough inside the cement so that I only see copper wire outside the cement, no connector or nichrome. That didn't work so well on the MK II rebuild though, as it has a lot less cement on it.

The new lasercut extruder is leaking at the lower end of the PTFE again despite my best efforts; I think I'll try the MK II tomorrow with your compensation code. The nice thing about the MK II is that I have several sets of spare parts printed already, so when it breaks I don't get stuck. I did run it for several weeks without any problems, but a lack of PTFE stopped me for a while.