Customised extruder feeder design *No Moving Parts*

[Eurisko84]

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
UPDATE 08/03/2012

New Version 2 has been loaded! New features include adjusted positions for the filament feed and air line, and dual 50mm fan mounts!

http://www.thingiverse.com/thing:18010

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

G'day All!

Just wanted to share my trials and trib....trials of designing and working this extruder.

I'm new to the whole 3D printing community, but not new to engineering. As a Mechatronics engineer I have worked in the field from small scale electronics to bucket excavators for mines. I recently purchased the UP! through my company to build small scale working models of our equipment. Its a great bit of kit, and its been a joy to play with, and have been successful in creating some custom parts.



To the point - being based in Sydney, Australia, the choices of filament are limited (as Madox can attest) and moreso when my requirement is to have Purple models. The only place that sold the colour I requried was Bilby CNC, and they have a range at a reasonable cost (see here http://bilbycnc.com.au/DisplayCat.asp?CatID=9&SubCatID=88).

This was for a good reason too - the quality of the filament is... poor. I have measured the filament from 1.4 to 2.1mm! There are also significant and visible bulges in the filament that cause serious issues, and prevent extrusion... and hence my requirement for a new feeder design.




One of my inspirations was Drew and his constant pressure roller design. It was, however, expensive to machine if you had access to the equipment which I don't have. I had to think outside the square, and with my dislike of moving parts, found this sitting infront of me.



And it struck me - the spring action of a leaf spring of a car. The principle is simple, springs have a force that varies depending on material and displacement, so depending on the thickness of the steel and how far it has to move, the more force it can apply. ABS is not an abrasive material, and with a few iterations I came up with this.



It has a few elements taken from Drew's design (the gear head cooler mainly) but utilises the spring steel from the Artline pen.



The air cooler is sized to take a silicon hose from a fish tank pump, and I am currently running a dual outlet 3.5l/m pump into here which cools it adequately.



As the spring is about 0.3mm away from the gear head, it will always be applying pressure to the gear, regardless. There is enough space to accomodate for the largest of thickness variations also, and works with the Bilby purple plastic!



I have successfully printed numerous parts, but until its working for a few days, I wont say its 100% as yet. Once I test it thoroughly, I will upload the files to Thingiverse - but for now I just want your input if you see any disasters that are lurking around the corners.


Cheers guys!

Eddie

[Madox]

Nice use of the pen clip!

I have an extruder project that i'll probably never get around to.... if it isn't too much to ask, can you adapt your design to work around the current overheating issue as well? i.e. perhaps a fan mount on the other side?

Pleaseeeeeeeeeee?

PS : Do you know about 'Robots and Dinosaurs'?

[Madox]

PPS : I did see the fan hose, but I wanted something on board rather than a hose from a air pump...

[darkhouse]

WOW, great design!! I like it, simple and effective

darkhouse

[Billzilla]

Aussie ingenuity.

[pleppik]

I love this concept, especially the simplicity of scavenging the spring clip. I have a reel of troublesome filament which would make a great test case--I'll give this a whirl as soon as you post the STL.

[DrewPetitclerc]

Very nice work.
Thanks for the reference to my work.
The only thing I would be concerned about is the friction forces, when the motor has been running awhile it is weaker due to heat/magnet stress and heat softening of the spring element, so that friction with the spring element could stall out the motor.
I have seen this happen when I recently tried some Polycarbonate, at 260c temp it needed my help (pushing the filament by hand) feeding as it has a higher viscosity and the motor had trouble turning even stalling completely.
This also why I run PLA at the 260c ABS temps as its higher viscosity at the 200c temp makes it stall or slip (clicking) when extruding and the nozzle to build plate distance can setup a back pressure causing stoppage.
All that being said.
I think your design is inventive and great, please keep up the great work and I look forward to seeing more.

[Eurisko84]

Cheers for the input guys!

Nice use of the pen clip!

I have an extruder project that i'll probably never get around to.... if it isn't too much to ask, can you adapt your design to work around the current overheating issue as well? i.e. perhaps a fan mount on the other side?

Pleaseeeeeeeeeee?

PS : Do you know about 'Robots and Dinosaurs'?

I can put a fan on the other side, but I agree with Drew - increasing the mass of the head with only the bent angle may introduce inaccuracies when making fast movements. We can give it a crack but! (Ill head to my PC store and find a fan or two to use).

I'm assuming you mean Hackerspace? I have seen it, but am not a member (as yet!).

I love this concept, especially the simplicity of scavenging the spring clip. I have a reel of troublesome filament which would make a great test case--I'll give this a whirl as soon as you post the STL.

I can certainly upload the file - but its still in a "alpha" stage at this point. In other words, it does have some problems, such as the air pump design is a prick to remove the internal support (due to the crossmember) but I wouldnt want to remove more material from it as it does assist in preventing flex. There are no dimples to insert the cover, and the gap is specifically designed for the artline pen at 0.7mm thick.

Very nice work.
Thanks for the reference to my work.
The only thing I would be concerned about is the friction forces, when the motor has been running awhile it is weaker due to heat/magnet stress and heat softening of the spring element, so that friction with the spring element could stall out the motor.
I have seen this happen when I recently tried some Polycarbonate, at 260c temp it needed my help (pushing the filament by hand) feeding as it has a higher viscosity and the motor had trouble turning even stalling completely.
This also why I run PLA at the 260c ABS temps as its higher viscosity at the 200c temp makes it stall or slip (clicking) when extruding and the nozzle to build plate distance can setup a back pressure causing stoppage.
All that being said.
I think your design is inventive and great, please keep up the great work and I look forward to seeing more.

Ill keep an eye out, but I dont think the heat softening should be an issue - the spring is effectively insulated as it doesn't make any contact with the servo (only the abs). The fish pump pumps a lot of air into that spot, which should keep the clip cool.

I have printed some larger parts (this one took about 3 hours to print) without any issues, but im still not confident in saying its "tired and true".





Ill keep tweaking - if anyone has any ideas to add let me know!

Thanks for the good points guys, appreciate the support!

Eddie

[Eurisko84]

Oh, and not to say I havent had any other issues with sourcing ABS...

I got this "ABS" from another mob here - its melting point seems to be extraordinarily low - it reacts with the heat from the platform!

It printed half a part then jammed... when I cracked her open I found this little jem...





*note this was before I was using my custom cable cover

[pleppik]

Cheers for the input guys!

I love this concept, especially the simplicity of scavenging the spring clip. I have a reel of troublesome filament which would make a great test case--I'll give this a whirl as soon as you post the STL.

I can certainly upload the file - but its still in a "alpha" stage at this point. In other words, it does have some problems, such as the air pump design is a prick to remove the internal support (due to the crossmember) but I wouldnt want to remove more material from it as it does assist in preventing flex. There are no dimples to insert the cover, and the gap is specifically designed for the artline pen at 0.7mm thick.

Still volunteering to help alpha-test

I don't see matching the model of pen as a major issue. Artline pens are carried by some art supply chains here in the U.S., and the clip looks very similar to what I've seen on other brands. A trip to the office supply store could probably identify a half-dozen candidates, though I don't imagine the store would appreciate my opening the packages to measure the clip thickness with my caliper.

[Eurisko84]

Still volunteering to help alpha-test

I don't see matching the model of pen as a major issue. Artline pens are carried by some art supply chains here in the U.S., and the clip looks very similar to what I've seen on other brands. A trip to the office supply store could probably identify a half-dozen candidates, though I don't imagine the store would appreciate my opening the packages to measure the clip thickness with my caliper.

Uploaded to Thingiverse - http://www.thingiverse.com/thing:18010

STL and SLDPRT files are there... now don't judge my modelling... I'm pretty ordinary at Solidworks and I don't use it often enough to know the best practice for it!

*stops covering his tracks*

Feel free to modify and adjust! I'm going to work on installing a fan I salvaged from a RAM cooler I had lying around

[pleppik]

Uploaded to Thingiverse - http://www.thingiverse.com/thing:18010

Doing a test model right now, using my grodiest reel of plastic. This one varies from 1.60mm to 2.0mm, and jams the stock feeder every time. So far so good, but I want to get a few models pushed out before I declare victory.

I found that the clip from a Pilot "Precise V7 Premium Rolling Ball" pen fits perfectly. In the U.S. these are readily available at most office supply stores and cost about a buck. They write well, too.

[pleppik]

Doing a test model right now, using my grodiest reel of plastic. This one varies from 1.60mm to 2.0mm, and jams the stock feeder every time. So far so good, but I want to get a few models pushed out before I declare victory.

The filament jammed when it hit a spot where it was around 1.9mm diameter. I think it may be jamming against the metal tube leading to the nozzle.

So this isn't a cure-all, but it did work a lot better than the stock component, especially with the undersized filament. With this reel and the stock parts it would jam every 2-4 grams. I got out about 15 grams on this test before it failed.

[odiegel]

Hi All,

Here's my little contribution, though not quite finished/tested yet. Have added baby fan for cooling instead of compressed air, and made a few slight mods to Drew's design to all for more pressure (as I intend to print in thermoplastic rubbber). Will do a bit more work on it over the weekend and try printing and testing in the next week, or so. Once design is finalozed, I'll upload the files...

Eurisko, I suspect the problem with the filament wrapping around the gear may have been due to you not having the plastic 'block' bit that sits around the gear which helps direct the filament the way it's supposed to go. So if using the pen clip (nice idea, by the way!), it would still pay to have something to guide the filament, particulalry as it leaves the gear/spring?

Incidentally, I printed a version very close to Drew's original mod in South Africa last week on a plastic laser sintering system and it 'appears' to work wihtout needing to be machined out of aluminium. Also meant I could do a few minor mods as machining complexity wasnt an issue. Will do some more, now that I am back in NZ, to confirm that Nylon will last. If it does, it will make a much easier way of manufacturng the parts than machining, which equals cheaper.

Cheers
Olaf

[Eurisko84]

Doing a test model right now, using my grodiest reel of plastic. This one varies from 1.60mm to 2.0mm, and jams the stock feeder every time. So far so good, but I want to get a few models pushed out before I declare victory.

The filament jammed when it hit a spot where it was around 1.9mm diameter. I think it may be jamming against the metal tube leading to the nozzle.

So this isn't a cure-all, but it did work a lot better than the stock component, especially with the undersized filament. With this reel and the stock parts it would jam every 2-4 grams. I got out about 15 grams on this test before it failed.

This was expected, and I thought it was only an issue with my extruder. But you are correct, if you measure the internal diameter of the extruder pipe, you will see its about 1.8mm, so anything bigger gets jammed.

I have run a 2mm drill bit through and shaved it a tad larger - that has allowed me to run even the largest blobs in my purple filament.

I wouldnt recommend doing this, unless you can spare the extra cash for another extruder, should this cause issues later on with regular filament!

[Eurisko84]

Eurisko, I suspect the problem with the filament wrapping around the gear may have been due to you not having the plastic 'block' bit that sits around the gear which helps direct the filament the way it's supposed to go. So if using the pen clip (nice idea, by the way!), it would still pay to have something to guide the filament, particulalry as it leaves the gear/spring?

Actually, that filament was warped using the standard bearing/housing, which has a guide built in to both ends! I will try this again with that particular plastic this weekend, with the new design and see how it goes - but you are right, my intention was to have a guide straight after the clip. It is a work in progress, and the more we all play with it, the better it will become!

Keep us all informed on the nylon parts! Still extremely keen on getting this to work 100%

Love the interest this is sparking

[Madox]

Hi All,

Here's my little contribution, though not quite finished/tested yet. Have added baby fan for cooling instead of compressed air, and made a few slight mods to Drew's design to all for more pressure (as I intend to print in thermoplastic rubbber). Will do a bit more work on it over the weekend and try printing and testing in the next week, or so. Once design is finalozed, I'll upload the files...

Incidentally, I printed a version very close to Drew's original mod in South Africa last week on a plastic laser sintering system and it 'appears' to work wihtout needing to be machined out of aluminium. Also meant I could do a few minor mods as machining complexity wasnt an issue. Will do some more, now that I am back in NZ, to confirm that Nylon will last. If it does, it will make a much easier way of manufacturng the parts than machining, which equals cheaper.

Cheers
Olaf

Can I beg you for SLS printed Drew mods? Would impregnating with CA (S glue) help with the strength of the SLS parts?

I'll would like the baby fan mod too That was basically what I was after, except in the unmodded extruders there is that guide/cover that you mention which blocks where I would like the air to flow.

Will be sending you an E-mail

[Eurisko84]

Just a quick update guys, I have recently revised and upgraded the cover to include a few more mods. I will post the pics later tonight if I get a chance, but heres a quick rundown-

1. Added two fan mounting points to mount on the back and side of the servomotor (50mm fans were all I had available - salvaged from a RAM cooler kit I had).
2. Increased the diameter of the fillament feeder and exit holes, and moved the exit of both much closer to the gear (to cool and prevent mis-feeding)

I have also added the heatsinks on the servo back and side to aid in cooling - this coupled with the fans have made a massive difference in the overall temp of the servo, I will measure with my thermocouple tonight.

The increased mass hasnt made any visible change to the accuracy of the head (I always print on Normal speed) and the prints are still straight and accurate.

I will keep testing and let you know the outcome.

As of this point, this extruder has not jammed on me once, but as I mentioned, I did bore out the extruder tube to 2mm.

Pics to come.

Eddie

[pleppik]

Just a quick update guys, I have recently revised and upgraded the cover to include a few more mods. I will post the pics later tonight if I get a chance, but heres a quick rundown-

1. Added two fan mounting points to mount on the back and side of the servomotor (50mm fans were all I had available - salvaged from a RAM cooler kit I had).
2. Increased the diameter of the fillament feeder and exit holes, and moved the exit of both much closer to the gear (to cool and prevent mis-feeding)

Let us know when you post the STL and I'll print a copy right away. The last version has been working well for me.

[Eurisko84]

Here you go - took some pics of the version 2 design - to suit a 50mm fan, but can be made to suit other sizes.

I initially designed it to carry two fans, but I did notice bobbing during fast movements - reduced to one fan, and its not as serious, but I will look into getting a reinforced locking arm for the X axis.

Things to note - I have been running this for a significant period, I just finished a 16 hour print of the Terminator Exoskull and it turned out great.



This shows the lowering of the fillament extrusion and air line closer to the gear head. The outlet guide has also been opened up slightly (the air line no longer has a taper for simplicity).

You can see the mount for a 50mm fan on the right hand side, along with a lip on the bottom. This is to force the air upwards and sideways, to avoid cooling the heater element. I no longer use this mounting position due to the bobbing, but can be used I think if the mounting bracket was stronger.



The rear shows the modified bracket for the rear 50mm fan. It is offset ~15mm to accomidate for the larger passive heatsinks I had (came with the ram cooler kit).

The heatsinks and fan have kept temps significantly lower than before. I can actually touch and hold the underside of the servo without burning my fingers.



This shows the current setup I am using - note the notch taken out of the bottom was due to the cable for the heater snapping, and needing to re-solder a connection (hence reducing the length).



Closeup of the heatsinks and fan.... and fancy blue light

Here is my latest creation, to give you guys an idea of the size and how well it is working. I have not had a single jam up since the mod, and its going strong.






I will upload the STL now on thingiverse. Updated in the first post also.

Cheers guys, and let me know your thoughts.

Eddie