Re: Replicator 2 - 3in1 Extruder Head Upgrade

[John Watson]

dude check in the putting out fellers thread and other threads ..we have been  selling and creating upgrades for a while now
plus there are various versions on thingiverse :-)

heres a link to a fine functional aluminum upgrade

http://www.wotz101.com/wotzbotz/index.html


On Friday, 12 April 2013 15:41:55 UTC-4, DHeadrick wrote:

I have read many posts about issues people have experienced with the Makerbot 2 shortly after receiving one.  I was only reading for interest until a few of these issues began to rear their head.  It is disappointing to have spend $2500 on a 3D printer and not be able to use it after as little as a week of printing.  My problems showed up within 12 hours of elapsed printing time!  This is not acceptable.  Unfortunately, my printer was no longer capable of printing out good enough parts to try any of the many fixes available on the web.

Luckily, I had access to other methods of manufacturing and have been able to design a few replacement parts which have completely resolved a few issues.  Here are the issues I found as I took my extruder head apart:

1)  The filament tension mechanism is an appalling design and it is obvious that it will never allow for consistent printing of quality parts.
2)  The gripper wheel that advances the filament does not engage the flat on the stepper motor shaft.
3)  The excess heat from the melter is allowed to travel away through the heat sink (where it should) but also the stepper motor housing (where it should not).

I have a designed upgraded parts which can be swapped into place in under 10 minutes which improve the consistency of the extruder.  Here are some sample images of the assembly in place:

The features this design incorporates are:

1)  It moves the stepper motor rearward slightly to allow the gripper wheel to engage the flat on the shaft.

2)  It adds an auto-tensioning system which includes a bearing and never needs adjustment.

3)  It also moves the fan closer to the heatsink so that the airflow must flow through the heatsink instead of past it through a large gap.  Also, the motor being spaced off from the heater bar reduces the amount of heat that is transferred directly to the stepper motor.

The materials for these parts are not set in stone yet because only prototypes have been built and tested so far.  The plan is to make the bearing mounting components from stainless steel and the filament guide block from either aluminum or delrin for durability.  The intention is to find out if there is any interest for a kit like this.  The kit would include ALL the parts required for the upgrade (not just a list of components and suggestions where they can be purchased separately).  None of the parts will require a functioning 3D printer to install which means you can get a non-functioning or misbehaving printer back up and running easily.

What I need to know is if there is any interest in this upgrade so that I can get a feel for if it is worth making.  The prototype has solved my issues and I am happy.  I want to allow others to enjoy the full functionality of their investment.

[DHeadrick]

I have seen that upgrade and it doesn't address the other 3 items my upgrade does.  Also, all the thingyverse upgrades require a working printer or you to source your own parts from a supplier (not easily done in the UK or Austrailia from what I've read).  Also, you cannot modify STL files to work if you find a similar part but want it to fit.

[brain...@comcast.net]

Sure you can modify an STL:
http://www.thingiverse.com/thing:64919

[DHeadrick]

Thanks... good to know :)

[brain...@comcast.net]

I am in the middle of many upgrades to my Replicator Dual right now.  I think I like your idea of moving the stepper motor back.  Will have to investigate.

[DHeadrick]

It allows the set screw on the gripper wheel to land on the flat of the motor shaft (as it was intended).  The prototype that I have running currently has this and during the filament load operation... it is very difficult to hold the fillament back from being fed into the extruder once the gripper has it.

[DHeadrick]

Another thing that I've noticed on a lot of lever type bearing tensioners (which this design is not), is that there is a little extension which sticks out the side to allow the user to release the bearing (this is not needed by the way).  Doesn't this interfere with the Y-Direction stepper motor and circuit board when the head moves to home position?  The max travel is determined by the extruder stepper motor touching the limit switch with the casing.  This does not leave much room for a bearing release lever to stick out and not cause problems.

Am I wrong?

[DHeadrick]

Are there other upgrade kits (that do not use printed parts) other than the wotzbotz one or this one:

https://groups.google.com/forum/#!searchin/makerbot/John$20watson/makerbot/L_SAu7I_avs/Q0lETfsR2A4J

On Friday, 12 April 2013 16:17:40 UTC-4, John Watson wrote:

[Eighty]

Yes, you're wrong. The lever goes over the top of the limit switch. And it IS necessary to allow the blob to pass out during unload. If you don't release the compression, the blob could break off between the pinch wheel and the barrel.

[brain...@comcast.net]

I don't know about a Rep2, but it doesn't get in the way for the Rep Dual.  I don't use the level during load, but it is handy while unloading.  Having the option is great, even when you don't use it.  The original Replicator does not have an issue with the drive wheel not locking onto the flat of the stepper shaft.

[Eighty]

That probably came too harshly.

I'm curious about what makes your design better. I'd like to see a pic of the guts, as I can't tell how yours is any different (aside from the setbacks and missing lever).

[DHeadrick]

Ah... well the replicator 2 (which this is for) does not engage the flat on the shaft.

[DHeadrick]

I have to leave at the moment, but I'll post some more images when I get back that might show how this is not a derivative of the other designs out there.

[Eighty]

Provided that the gear is tightened down well, I don't really know why it needs to be on the flat of the shaft. Can't say I've heard but a handful of complaints about anything ever slipping there. And IIRC, MBI had some explanation as to why they didn't want it that way. Anyone else know what that logic was?

[Eighty]

Depends on how tightly it was adjusted. Mine looked like that after about 40 hrs (IIRC), when I called MBI for a replacement.

[Joseph Chiu]

Wow.  Before PLA, I ran lots of ABS and my plunger only had faint grooves.  When I went to PLA, the plunger started to have much deeper grooving, but nowhere near that bad, and not so quickly.  I am impressed.

On Fri, Apr 12, 2013 at 5:43 PM, Eighty <adun...@gmail.com> wrote:

Depends on how tightly it was adjusted.  Mine looked like that after about 40 hrs (IIRC), when I called MBI for a replacement.

--
You received this message because you are subscribed to the Google Groups "MakerBot Operators" group.
To unsubscribe from this group and stop receiving emails from it, send an email to makerbot+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Eighty]

I imagine it has something to do with the filament too. Some of my PLA has little bumps, kinda like embedded grains of sand. That probably cuts right into the Delrin. Joseph, did you notice that on the samples I sent? I was surprised when I got your samples, as they were much smoother (and a little translucent).

[DHeadrick]

I was surprised also.  Even if it does take 40hrs plus to duplicate that groove, is that acceptable?  And regardless of whether or not you can get free replacements from MBI, is that acceptable with the install time and adjustment required?

Also, the plunger head is larger than the width of the gripper wheel slot so eventually it will wear and not be able to be adjusted any closer to the filament.

On Friday, 12 April 2013 20:26:16 UTC-4, James Spencer wrote:

I don't see how you can wear that kind of groove into the plunger in 12 hours. That should take more time than that.

[DHeadrick]

I would be interested in knowing why they wouldn't want to prevent relative motion between the gripper wheel and the motor shaft.  They should turn as one unit.

[DHeadrick]

As promised, here are some images to explain the merits of the design.

These first two images are just the full assembly and then the extruder without the heatsink and fan attached.

The follwoing image shows a section view to illustrate how the motor is moved rearward to allow the setscrew to land on the flat of the shaft.

This next image shows a view down the filament feed path showing how the filament is guided the entire way to the gripper wheel.  Also shown is how the bearing (red) is sized so that it can fit into the groove in the gripper wheel enabling it to adjust for a large range of filament thickness variation.

These next two images show the extents of the mechanisms auto-adjustment range.  It can adjust down to filament that is as small as 0.5mm in diameter and open up to completely clear the path (which is the same as stock - 2.25mm).

I welcome comments.

[DHeadrick]

Even a blind man could load filament in this design.  Cut an angle on the tip of the new filament to be fed and insert.  Let the machine do its thing.

[Eighty]

Interesting take on the design. Couple of questions:

1.) Have you logged many hours on this design? I'm wondering how the enclosed nature of this affects the heat creep issue. I've been operating under the assumption that we'd want more open air to dissipate heat above the barrel, but you never know until you try something different.
2.) One concern I have is if/when there's a jam that requires disassembly with the filament still loaded. It doesn't seem like you can access the area below the pinch wheel, to cut off the filament. I've had to do that before, so that I could remove the extruder. So you could end up with a situation that requires a lot of yanking to rip the stuck filament out along with the extruder assembly. Did that make sense?
3.) Have you performed a lot of unload procedures where a blob comes out at the end? As we discussed earlier, I'm still concerned about that blob getting torn off because you can't release the spring pressure.
4.) Looks like you have a much shorter moment arm that other versions, but compensate by compressing the spring to its fullest. I wonder how the result compares - i.e., total compressive force on the pinch wheel.

[DHeadrick]

Thanks for your feedback.  There is always room for improvement which is why I wanted to get feedback regarding interest as well as questions like these because I can't experience all all the ways problems can arise myself.  Let me see if I can address some of these:

1)  I have logged only a few hours of printing so far on the design but continue to run it looking for a failure.  I haven't experienced any problems but only time will tell.  The fan/heatsink combo with the reduced gap as well as the small spacers between the heater block and the stepper motor help prevent heat from conducting into unwanted areas.  I have personally found that printing with an extruder temp higher than the default works better (235*C - 240*C).

2)  There IS a gap between the guide block and the heater block inlet nut to allow scissors to cut the filament in the case of a major jam.  To do this (with most designs - stock included) you have to remove the fan/heatsink to get to this area, but it is possible.

3)  In order to test the pull strength of the gripper/tensioner combo, I had to load and unload filament.  Sometimes there was a blob on the end but never a huge gob.  Do you have some possible steps to produce this so I could try it and get back to you?

4)  I have had to try multiple spring rates to produce the desired force.  Too much force and the bearing pinches so hard that the stepper can't increment to the next step and just clicks in place.  Too weak and there isn't enough pinch force to feed properly.  I think I've zeroed in on the ideal force but I need to continuously run the printer to test it out.  I think I will start building a block which is the maximum volume of the build space to give it a sufficient test (it will have to be over a weekend).

[DHeadrick]

I have installed OpenSCAD and am still unclear as to how to open an STL file to convert it back to something that is usable by CAD software for modification... I'll keep reading :)

On Friday, 12 April 2013 16:59:18 UTC-4, brain...@comcast.net wrote:

[Eighty]

I hate STL files, wish the slicers accepted something else. Thumbs up to people who post the source files too (like whpthomas with his upgrade).

The only suggestion I have for creating blobs is to heat up the extruder (loaded), and let it sit for a few minutes. Then unload. You'll usually find that the plastic above the barrel slouches down into a bit of a pile.

Glad to hear that you can snip off the filament in place. Sure you have to remove the heat sink/fan, but anything more would be a pita.

From my experiences, prints that go past 8-10 hours usually have heat issues (which I have yet to completely identify). You don't necessarily need a huge model, just take a smaller one and slice it at 100um. That's a good test too, because the filament isn't being expelled as quickly as a 200+um print.

[Dan Newman]

On 12 Apr 2013 , at 8:19 PM, DHeadrick wrote:

> I have installed OpenSCAD and am still unclear as to how to open an STL
> file to convert it back to something that is usable by CAD software for
> modification... I'll keep reading :)

If you look at the .scad file for Emmett's Heart Gears at Thingiverse,
you will see that the OpenSCAD commands load an STL file and then use
it. (In the case of the heart gears, it intersects it with algorithmetically
generated cube gears.) Anyhow, that .scad file will show you how to
open an STL file in OpenSCAD.

Dan

[Dan Newman]

> The only suggestion I have for creating blobs is to heat up the extruder (loaded), and let it sit for a few minutes.

Better yet, print for a bit so things are good and hot. Then turn the bot OFF
that way, the heatsink cooling fan is not even running. That has the best odds
of giving you one of those big blobs. (Happens to me from time to time: I have
to run and I don't want to leave the printer on, even not printing. You get
gunshy that way with your Rep 1 after you've replaced the motherboard once
already.)

Dan

[DHeadrick]

OK... got OpenSCAD to import an STL file I made but still see no easy way to modify it.  I could use boolean operations with geometric shapes to modify it within OpenSCAD but no way to export it to anything useful to other programs.  Is there something I'm missing?

[Dan Newman]

On 12 Apr 2013 , at 9:25 PM, DHeadrick wrote:

> OK... got OpenSCAD to import an STL file I made but still see no easy way
> to modify it. I could use boolean operations with geometric shapes to
> modify it within OpenSCAD but no way to export it to anything useful to
> other programs.

OpenSCAD is not an STL converter. It's not a tool you want to use to
convert STL to other formats. If that's your goal -- convert an STL
to other formats -- I myself do not have any good suggestions for you.
(Not too sure why it was suggested to you if that was your goal.)

Dan

[Dan Newman]

On 12 Apr 2013 , at 8:19 PM, DHeadrick wrote:

> I have installed OpenSCAD and am still unclear as to how to open an STL
> file to convert it back to something that is usable by CAD software for
> modification... I'll keep reading :)

Others may have a different opinion, but one of the more common ways to
convert STLs to other formats is to use Blender. Blender, which is free,
happens to have a steep learning curve. But there tends to be a problem
even after you convert the STL to another file: it's still just a bunch
of triangles with no underlying nice CAD rational behind them. That is,
the CAD program will still just have a mesh of triangles and not some
nice recognizeable union/intersection/whatever of primitive solids. Even
a simple cylinder will no longer be composed of some circles: it's just
a bunch of triangular facets approximating a cylinder.

So, you may then be able to convert the STL to another format and then
resize, punch holes through, cut sections off of, etc. the result. But
you won't have a suddenly nicer model with which to work in the CAD
program.

Dan

[Bottleworks]

You turn your bot off? I never do that. My fears are damage during cold starts.

[David Celento]

The design efforts are to be applauded -- especially if the work better than existing upgrades. I like the adjustable spring idea (even though I'm having trouble deciphering "exactly" how it works. Echo the concerns about The Blob, which is a serious jamming threat -- especially for new users who are easily caught out.

YES. Imperative to reduce heat to drive motor (and diminish internal drive motor heat). In fact, the Rep2x moved it back. Personally, I'm beginning to think the heat sink may be a bigger problem than it is a solution as it blocks air to the filament AND creates a chimney for heat to rise in and around the filament. Cut it in half horizontally? Drill holes in it? I dunno. Testing needed.

Most seem happy with the upgrades by users (and possibly MBI's, if the hole in the bottom is big enough and the shoulder screw seats on the motor). Whpthomas' Rep 2 (in ABS if possible) in particular has proven to work well. To prove your design is better -- and it may very well be, which would be great! -- you'll likely need to print 2-3 hundred hours of varied sized prints, up to a few 24 hour prints using no accelerated AND accelerated speeds. If you convince the experienced users here with results, you've succeeded!

As for STL to editable NURBS -- this is black magic. Very few programs do this. The one that does it best (that I know of) is Geomagic. It takes an (STL) pointcloud and wraps it with NURBS surfaces. At about $30k, bring your wallet! Netfabb might work (have not tried) and maybe Rhinoceros (perhaps aided by plugins).

Eager to learn of other's success stories converting polygons to NURBS.

Good luck w your design project and keep us posted!

[David Celento]

Hey Eighty: you're really OK with the drive wheel set screw on a hardened round drive shaft?

No doughnut for you! ;-P

[Eighty]

I admit that it isn't the proper way to install, but I just don't see the need to modify it. The gear will strip the PLA (or the stepper will just click) before you'll have problems with the shaft. I guess I'm coming from the "if it ain't broke, don't fix it" camp.

On a side note, Autocad (and Inventor?) have a shrink-wrap function.

[David Celento]

Of the many "birth defects" my Rep2 showed up with, my drive wheel slipping was one of them. The shaft had rotational score marks on it where the set screw was located.

For me, peace of mind -- and the elimination of one of many variables preventing proper printing out of the box -- took 5 minutes with a file. Perhaps it's just cheap insurance for those running out of swear words! :-)

[Joseph Chiu]

Eighty, the samples you sent me varied in smoothness, but I don't notice any one being particularly worse than others.  The MBI group and ProtoParadigm group of filaments were mostly smoother than the Ultimachine and MatterHackers group of filament.  But they each had their (relatively) rough and smooth filament.

I find ABS to be much smoother regardless of source -- if you shine light on the filament, you can actually see that ABS has a much more mirror-like finish, while PLA generally has a slightly textured finish. 

 

On Fri, Apr 12, 2013 at 6:11 PM, Eighty <adun...@gmail.com> wrote:

I imagine it has something to do with the filament too.  Some of my PLA has little bumps, kinda like embedded grains of sand.  That probably cuts right into the Delrin.  Joseph, did you notice that on the samples I sent?  I was surprised when I got your samples, as they were much smoother (and a little translucent).

[Dan Newman]

On 12 Apr 2013 , at 11:01 PM, Bottleworks wrote:

> You turn your bot off? I never do that. My fears are damage during cold starts.

I'd like to say that electronics are normally very robust and tough; however,
that assumes that the design is sound. And my MightyBoard did blow on power up.
But I have no way of knowing when the underlying bad event happened which led
to the 5V reg committing suicide by (likely) thermal run away followed by self
explosion. The Bad Deed could have happened when I powered the bot off previously.
I don't know. (As most/all of the logic chips were toasted, it's a safe bet that
at some point 24V spilled onto the 5V rail. That alone -- Vin and Vout shorting
together on that linear regulator -- is enough to take out the regulator in such
a fashion. Typically those regulators can handle a short of either Vin or Vout
to ground as it's part of their built-in short circuit protection. BUT, I don't
know what exactly caused my motherboard to die: the cause may have been something
else entirely such as a failed stepper driver chip.)

At any rate, I turn my bots off when they are not in use. There are components
in them -- the electrolytic caps in the PSUs for example -- which do have rated
lifetimes. OTOH, the PSUs are easier to replace than the motherboards. So, do
what you are most comfortable with.

Dan

[Richard]

Just jumping in here but this looks like a nice design with the cam
spring combo that keep pressure on the bearing while allowing it to
adjust.

On Apr 12, 9:19 pm, DHeadrick <davidheadric...@gmail.com> wrote:
> As promised, here are some images to explain the merits of the design.
>
> These first two images are just the full assembly and then the extruder
> without the heatsink and fan attached.
>

> <https://lh6.googleusercontent.com/-oHh_Bcp7PTA/UWi9byrm_aI/AAAAAAAAAK...><https://lh3.googleusercontent.com/-w-UWKJ0Q2O4/UWi9iw5fLlI/AAAAAAAAAK...>

>
> The follwoing image shows a section view to illustrate how the motor is
> moved rearward to allow the setscrew to land on the flat of the shaft.
>

> <https://lh5.googleusercontent.com/-jpLUMNalj8I/UWi9oT8_kNI/AAAAAAAAAK...>

>
> This next image shows a view down the filament feed path showing how the
> filament is guided the entire way to the gripper wheel.  Also shown is how
> the bearing (red) is sized so that it can fit into the groove in the
> gripper wheel enabling it to adjust for a large range of filament thickness
> variation.
>

> <https://lh3.googleusercontent.com/-WXUV0jVwwDA/UWi9utKkxqI/AAAAAAAAAK...>

>
> These next two images show the extents of the mechanisms auto-adjustment
> range.  It can adjust down to filament that is as small as 0.5mm in
> diameter and open up to completely clear the path (which is the same as
> stock - 2.25mm).
>

> <https://lh4.googleusercontent.com/-hoJgj_kECsc/UWjAVwjovWI/AAAAAAAAAL...><https://lh4.googleusercontent.com/-aHE_iUaRqcM/UWjAR9R4tqI/AAAAAAAAAK...>
>
> I welcome comments.

[Bradley Pearce]

Isn't this the same style setup that MBI used on the 2X? They just
added a lever to assist in filament removal?

> --
> You received this message because you are subscribed to a topic in the

> Google Groups "MakerBot Operators" group.

> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/makerbot/m2X36BIqEWE/unsubscribe?hl=en-US.
> To unsubscribe from this group and all its topics, send an email to

[Eighty]

Hmm, good question Bottleworks. I haven't seen the guts of the 2X extruder yet, although I'm eagerly awaiting the details from Kobus and Whpthomas. Have you seen inside one yet?

[Bradley Pearce]

I've only seen the pictures posted. It looks like they were aiming
for this concept. Maybe?

http://www.flickr.com/photos/92380311@N04/8636511313/in/photostream

[Carl]

The easiest way to 'fix' the original Rep 2 filament feeder is to make sure that the gear is not able to touch anything on the side opposite the filament path.

There are two instances where you will hear the clicking sound... the first (and most obvious) is when the filament is slipping... the second is when the pressure put onto the gear by the delrin plunger forces the gear to touch the other side and stop.

One of my first mods to my machine was to cut away almost everything to the right of the gear - to ensure nothing was able to stop the gear turning - messy, but it worked for hundreds of hours...

The second mod I made was to add a little spring behind the plunger, between the grub screw and nut... this eliminated the need to adjust the tension.

[DHeadrick]

The fact that the gripper wheel touches the non-filament side of the plastic housing should indicate to you that you are side-loading the motor shaft far too much.  Exhausting material from the block so that it doesn't touch is not the solution.  This only allows you to side-load the motor even more.

[Carl]

Agreed... If you are causing too much side force... the stepper will stop and no filament will extrude... regardless of the casing! :-)

Firstly, it depends on what material you are printing with as to the amount of pressure that is required...

The gear needs to be in the correct alignment - the gear teeth on the stepper side are what you want to bite the filament with... not the middle section...

The casing of the standard Rep 2 feeder has warping tendencies when exposed to heat for long durations ie: 10 hr + prints at 235+ deg C. This leads to the parts softening and the sagging when hot - and when cool the parts almost return to their original form. The casing does not allow enough play to accommodate the minute flex on the stepper shaft - even when cool - and definitely not when hot...

If you machine is set up properly PLA can be printed at pretty low temps (around 220 deg C) but that is still hot! :-)

The stepper motor operates at temperature of up to 60 deg C... I'd advice mounting the heatsink to the back of the stepper, the fan to the aluminium cooling block, cutting away any part of the casing that could possibly touch the gear on the side opposite the filament and finding a decent spring to automate the plunger pressure...

This is what worked for me for over 400 hours before I designed my own replacement! :-) Almost everyone with a Rep 2 will eventually have to find their own solution - or use one as suggested above... Pity something that worked did not come installed on my MakerBot Replicator 2...

[DHeadrick]

Could you elaborate further on the quoted part of your reply below:  Thanks.

[David Kessner]

I fixed the set screw issue by making the flat part of the shaft longer with a Dremel.  Now it is on the flat.

[DHeadrick]

I have gone over all the feedback you guys have provided and also read a lot of the other discussions on the Makerbot Operators group and have reworked the design.  I am not at home right now, but will post up images shortly.  I have been able to add a filament guide tunnel after the gripper wheel as well as before to help prevent buckling.  The exit guide tube is 3mm in diameter.  I have also added the ability to manually release the bearing in the case where the filament gobs up into a large ball and will not unload.

I am running the printer now with these mods and everything is running smoothly so far.  I intend to run a 13hr build overnight tonight and should have something to report in the morning.

Thanks again for the useful input.

[John Watson]

man just get the wotzbotz upgrade hehehe its all done in lovely aluminum...been printing for over 60 hours now without a hiccup :-)

we already have a back order of 40 :-) and its gonna be a loverly piece :-)

plus it wont soften with heat ..and the spring mechanism is solid :-)
:-)

and there is noooo parrots involved whatsover

[PhGeis]

any wotzbotz r2x experience so far?

[John Watson]

not yet ...looking into it tho .:-)

gotta get all the others sorted and out the door :-)

for all those rep2's and rep 1's

[DHeadrick]

No thank you ;)

[John Watson]

come on mr heady the headster :-) you know it makes sense :-P

all that hard work your doing and its already made :-)

I will even give you a  rowntrees fruit gum

[Carl]

LOL! John... If everyone had that logic... we would still be bashing rocks together to build things! :-)

[John Watson]

good Call I rest your Case :-)

[DHeadrick]

I'll have to pass though that's a very appealing offer :).  I enjoy design (it's why I became an Engineer) and choice/competition is good.  It breeds innovation.

[John Watson]

very tru ...so lets make a completely new machine  with  auto mood sensing and  wifi extrusion based on (insert comment here)

with  stabilizers to stop my bloody glass plate from wibbling my machine to death .. zoinks

a way to heat and cool the build plate as necessary  a telescopic system so we can just insert a bigger build plate when needed

a pat on the back arm for a job well done

an automated way to finish pla ..instead of spending eternity trying to sand the crap

an automated  remove tool like what  large format printers use . a knife

and any other list of things fun or otherwise we can think of ..

we should make a wish list

[Carl]

Have fun with that... I'm off to sleep now! :-)

[DHeadrick]

In the left image you can see the slot for a screwdriver that will rotate the blue part to release the bearing pressure for disassembly and also if a gob gets caught below it and won't feed back through.  In the right image you can see that the filament travels through an exit hole also which may help prevent buckling.

I will report back in the morning regarding the outcome of the 13hr print (9 cubic inch block)

[DHeadrick]

Build lasted 11 hrs 15 minutes with no issues.  No defects anywhere in the print that I can see.

Also, I changed filament from opaque white PLA to clear PLA just before the build and the extruder needed no adjustments for the different filament hardness (I find the clear is harder and more brittle than the white).

[DHeadrick]

I guess I should also give the settings I used to print:

Block created in Solidworks 2012 with fine resolution STL output.
Replicator is using stock firmware 7.3.
Part was exported with Makerware 2.0.2.9

Makerware settings as follows (no custom profile used so all the variables are stock):
Quality High
With raft (no supports)

100% infill
3 shells
0.1 layer height
240*C temperature
80 mm/s extruding speed
150 mm/s travel speed

[David Celento]

Very promising!!!  

Next test is Cells Bowl by Dizingof!!! :-)

BTW: how does one adjust the cam (which adjust the spring pressure)?

[DHeadrick]

Currently it isn't adjustable but I could look into adding that.  I had to play with locating the right spring force using multiple springs.  Right now there are 2 cavities for springs and they can both be used or just one as required.  I will try that cell bowl to see how it goes.

And a correction to my post above... I'm using bot firmware 7.2 (I'm not from the future).

[DHeadrick]

Just to keep you updated.  I have modified the design to add spring adjustment.  This can be done on the fly now without taking anything apart.

I tried printing the "cells bowl" but couldn't get makerware to slice it... so no luck there.

I will post an image of the updated design so that others can see the improvement.  The next step is making drawings of the 2 components and get quotes :).

[DHeadrick]

Here is the final design.  It is and adjustable spring mechanism that will maintain even pressure on the filament even if it varies.  The pressure can be adjusted without taking anything apart so it can be done on the fly if you change filaments.  From what I am reading, the material surface hardness varies depending on PLA/ABS and even between the difference colours of PLA (I've noticed clear seems to be harder than white).  This will allow you to dial in the amount of grip you want.

For those of you interested in math/physics, here are a few noteworthy features of the design.  The bearing carriage (blue) rotates about the center of its circular feature.  The distance at which the spring acts from this center is constant.  The moment distance at which the bearing acts varies depending on the filament thickness.  The thinner the filament the smaller the moment distance, but you will also notice that the spring had to extend to move the bearing here thus reducing the force it exerts.  Conversely, as the filament gets thicker, the bearing moment arm increases but the spring force also increases to compensate.  This design keeps gripper wheel pressure constant throughout the range of movement of the bearing.

You may note that the dozens of lever arm designs on the Thingiverse do not compensate in this manner and can allow pressure to vary with filament thickness and potentially cause problems.

Food for thought.

[David Celento]

Conceptually, some interesting points about varying pressure. Many aspects to like, provided it works as desired. Looking forward to seeing Cells Bowl printed...or something even more wicked! :-)

[David Celento]

FWIW, some method to permit removal of The Blob seems prudent, otherwise it may break off and jam.

[DHeadrick]

What is normally required if a blob needs help escaping the extruder?  Do people usually have to remove the fan and heatsink?

I've never had this issue yet with my printer.  I've had where the filament gets stripped and won't unload without a good tug.  Also sometimes I have to begin the load function and push the filament then run the unload function and pull to get it to release, but it always comes free.  Once I cut the melted filament away, it always loads fine right after this.  The exit hole in my block is 0.125" (just over 3mm) so if the blob is larger than this, it won't feed through that (but the heater inlet pipe is smaller).

I keep reading conflicting information.  Some people want it all open so they can get scissors in to cut filament.  Others want an exit tunnel to keep filament on target as it leaves the gripper wheel on its way to the nozzle.  The nozzle inlet is not large so how large a blob can form and where?

[Eighty]

The blob forms above the nozzle. To describe it more anatomically correctly, it's a mushroomed end. It typically can't be pushed through the nozzle, so it has to come up and out.

It's one thing to disassemble the heat sink and fan to snip off the stuck part, but you only want to go through all that trouble when it's truly stuck.

When it's just a mushroomed blob, the easiest thing is to release pinch wheel pressure and let it pass up and out. Provided, of course, that the exitway will let it pass out.

[DHeadrick]

I suppose this design will allow that also.  If you record the setscrew depth so that you can return it to the setting you like, you could remove the spring entirely allowing the filament mushroom to pass.  The inlet is just over 2mm in diameter, so that may also be an issue (the stock guide block was not any larger).

When this mushroom happens, the unload filament function cannot accomplish this?  What prevents the blob from passing?

- is the filament wedged into the nozzle inlet tube?
- is the mushroom too large to force the bearing out of the way?

There is very little room between the gripper wheel and the nut around the heater inlet.  If you were able to cut the filament with scissors, what then?  Do you then have to remove the extruder assembly and pull out the filament?

[Eighty]

> When this mushroom happens, the unload filament function cannot accomplish this? What prevents the blob from passing?

> - is the filament wedged into the nozzle inlet tube?
> - is the mushroom too large to force the bearing out of the way?

Usually, you unload until it hits a necked-down section. The neck-down happens when the blob is adhering to the tube below, but the top is being extracted (thereby stretching the filament). Eventually, the blob lets go, but you have a stringy part with the blob at the end. At this point in the unload procedure, the pinch wheel can't grab the stringie because its too narrow. You can pull on the free end, but you risk breaking it off along the stringy section leaving the blob inside. So conventionally, we release pressure and tease it out until the blob passes through gently.

> There is very little room between the gripper wheel and the nut around the heater inlet. If you were able to cut the filament with scissors, what then? Do you then have to remove the extruder assembly and pull out the filament?

As I said, snipping is only required when something is jammed and you can't extract it from the top. So at this point, you take off the heat sink and snip it from the front (below the pinch wheel). Then slide the stepper back and use needle nose pliers to pull out the stuck-in blob. The concern is that I'd you can't snip from the front, you have to take apart the top shroud so as to remove the stepper vertically. That's just more work.

Keep in mind that the competing system has both an easy-release lever and a front-facing port for snipping. So giving up either of those options puts your system at a disadvantage. Unless, of course, you've offset that with a huge advantage elsewhere in the functionality.

[David Celento]

Agreed with 99% of Eighty's thoughts. The 1% difference is where the blob forms and its diameter.

I suspect the blob forms somewhere deeper in the feed nozzle and the diameter is generally not larger than the filament. It appears to me that this element (which happens pretty near 100% of the time on unloads) is a function of stretch on filament at the edge of the melt zone. A small portion of the filament starts to detache slightly from the strand upon retraction. Possible resistance in the tube as it withdraws causes it to neck (meaning, from top down: filament, slight neck, small chunk of filament -- aka "the blob" -- then thin thready wisps) or it simple necks because it's starting to melt a couple mm from the tip. Hard to say.

The central point (where Eighty and I completely agree) is that the pinch wheel needs to facilitate COMPLETE pressure removal on unload (and for every unload) so the bearing does not apply pressure to the neck, causing it to fracture off and jam. Only some kind of lever style mechanism for pressure release seems to offer the ease of use and functionality needed.

As for being open, versus closed, below the drive wheel, I see two arguments. Open is nice "if" it jams (but this may actually "cause" jams). If closed, jams may be eliminated -- but only if nothing unexpected happens! :-)

Your inquiry is applauded!

[DHeadrick]

Here is an image I took of a plug that happened when the rafting did not stick to the platform and when I checked on it, it had balled up pretty good on the outside of the nozzle and caused a jam.  If I'm understanding the terminology correctly:

A - Stripping (caused by gripper slipping and grinding away the filament)
B - Teeth marks made by the gripper as it meets resistance and tries to force more material into a plugged nozzle
C - Stringy Neck due to stretching of the hot filament upon unloading
D - "The Blob"

I have thought of a way to allow access to the filament if it needs to be cut but haven't quite devised a way to release pressure entirely short of removing the spring (which means having to dial in the pressure again - which I don't want to force on the user because it takes time which could be spent happily printing).

My idea for accessing the filament would also allow you to brush off the gripper wheel if it is experiencing a build up of filament bits which are causing slipping.  I have read where people have had to increase the extrusion stepper speed just to make up for lost traction due to the teeth getting filled with filament buildup.  This filament access would allow you to look and see if this is the case.

Now I just need to allow the user to release pressure.

[TobyCWoods]

Would sure love to see pics!

[David Celento]

Good example!

I suspect that "B" may be normal and probably is desired. Positive grip with no slip.

If one is speeding up the feed due to particle buildup in the wheel, the problem which really should becaddressed is that the wheel has not enough pressure applied to the filament (creating small shavings) and it needs to be cleaned.

[DHeadrick]

Here is the modified design.  Does anyone have a problem with the release needing a tool to operate (one of the hex keys that came with the makerbot)?  You push the hex key to disengage the bearing pressure.  The finished assembly would looks like this:

The large opening allows visual confirmation that your tension is right by allowing you to verify that the gripper wheel is not building up with small shavings.  If this is not needed, I can fill in this port and that would allow a guide tunnel for the hex key.  However, I think that the access port could be useful to some having issues.

I am going to print one out and install it to ensure it functions as well as the current prototype.

Thoughts?

 On Saturday, 20 April 2013 23:43:49 UTC-4, David Celento wrote:

If one is speeding up the feed due to particle buildup in the wheel, the problem which really should be addressed is that the wheel has not enough pressure applied to the filament (creating small shavings) and it needs to be cleaned.

[David Celento]

A tool inserted (as long as it's easy to engage blindly) is fine.

Giver 'er a shot!

[Eighty]

Hang on.

If you're pausing mid-print to swap colors, you can't fit a tool in there. The bot will have homed, and that side will be up against the endstop.
So with this setup, the only way to do it is to jog the X axis over a bit. Not the end of the world, but I'd prefer not to jog during a pause. I don't want another thing to worry about (losing stepper registration). And I don't even know if the MBI firmware will let you jog during a pause.

[DHeadrick]

I'll build and install it and check how easy/horrible it is. If the print head returns home, how do you access the lever type bearing release? Isn't it located above the Y-pos stepper motor and control board?

[Eighty]

Yes, but it is high enough to clear the switch, even when depressed.

[Kerry Chin]

My jams look like this too!

I just printed myself a new Makerbot extruder upgrade and I still can get the thing to run for more than 60 minutes before this happens. I even installed a new heater block assembly at the cool cost of $120.

I used to run 15+ hrs prints no problem, but even now, after new heater/nozzle and extruder upgrade, this thing jams. It's spinning out and getting stuck. Ahhh! So frustrating.

Is it normal to have the stepper motor getting so HOT? I don't remember it so hot before...

[DHeadrick]

If you mean pics of the "real parts", they don't exist yet.  I am sticking to printed parts until the bugs are worked out.  I can take a pic of the prototype parts if you wish :)

[DHeadrick]

Got it!  I have figured a way to get access to the bearing release when the unit is installed an at home.  I am going to install it tomorrow morning and take some photos showing the access.

[Wingcommander whpthomas]

DHeadrick, while I applaud the effort you are putting into this, there are already proven, reliable Mk8 upgrade designs available, which you could machine up in your workshop.  

[DHeadrick]

Thanks for your applause.  If everyone stuck to what was available we would all be driving cars from the 1950s because they do the job.  My intention is to fix multiple problems with the unit with a quick and easy replacement part.

This design allows all the features of the existing solutions and:

1 - Allows positive engagement of the gripper on the motor flat.
2 - Adjustable spring tension with release capability.
3 - Allow better heat dissipation during long prints by having a large port for heat to convect away.
4 - Allow visual access to the gripper wheel and even the heater inlet port.
5 - Allow scissor access in the case of a filament jam
6 - No additional cost (this is my very attainable goal)
7 - Aluminum guide block for heat resistance
8 - All parts required included in the kit (including the spring adjustment hex key)

Below are some photos of the final prototype which will be send out for manufacturing quotes.

[DHeadrick]

[DHeadrick]

After I finished installing it, I put the bot into load filament mode and played around with releasing spring pressure and pulling the filament out manually while it was still trying to load.  After doing this for 5 minutes or so, I caused a fairly major jam up.  I made a blob big enough that it would not come out of the inlet hole before the gripper wheel.  I was able to snip the filament without disassembling anything and pull the feed line free.  Then I noticed the blob I snipped off sitting in the top of the heater inlet pipe.  If I had long enouth tweezers, I would have been able to remove it and reload filament without having to take it apart.

It was the first time I had experienced what people were describing in the previous comments.  I was impressed that the jam could be so easily fixed.

[DHeadrick]

If I made mirrored versions of the parts, this would also work for the Replicator 2X I would think.

[Wingcommander whpthomas]

Thanks for your applause.  If everyone stuck to what was available we would all be driving cars from the 1950s because they do the job.

In all seriousness - the MBI plunger design was rather tragic and did need improving - and I really hope what you have come up with provides additional benefits over the minimal Mk8 design, but there are a few things in your list here that seem overstated either as problems or potential benefits. If I were you and trying to differentiate your upgrade kit, I would stick to the key benefits it provides and focus your attention on those - rather than naming lots of potential benefits with claim that are unsupported.

BTW: old technology that works reliably can easily trump new technology - the russian Soyuz capsules re-supplied the ISS quite reliably as I recall =)

---- ~ ----

So a quick critique of you list

1 - Allows positive engagement of the gripper on the motor flat.

Not sure exactly what you mean by this one, but if you think pulling power is an issue, the Minimal au Mk8 will lift a tangled spool without slipping - it grips about as strong as the stepper can push or pull.

 

2 - Adjustable spring tension with release capability.

If it has the right spring tension from the start then for PLA adjustment is not necessary - but for ABS and Nylon this could be an important feature to have [TICK].

 

3 - Allow better heat dissipation during long prints by having a large port for heat to convect away.

Well speaking from experience, the aluminium works as a large heat soak, air flows in from the left side through the cowling and inside the top covering and leaves through the gaps in the spring. The stepper is also cooled by the cooling block - consequently the stepper runs cool to touch even after 22hours of non-stop printing. Not sure this is a real problem in practice.

 

4 - Allow visual access to the gripper wheel and even the heater inlet port.

No argument here [TICK]

 

5 - Allow scissor access in the case of a filament jam

No argument here either [TICK]

However, you can prevent blobs on the end of the filament forming by doing a quick load before unloading so in practice this is not really a problem I have encountered. The upside of the Mk8 is that manual filament changes are really quick. To do this, I tend to grip the under side edge of the extruder with my fingers, depress the lever with my thumb, push the filament down to clear any blobs, and then pull it out. It never jams, and filament changes are really quick - you to the same thing to reload - insert filament, release tension with thumb, then push it all the way down until a small amount leaves the tip and resume print. If you have two feeder tubes and the second real is already prepped, you can do a colour change in a mater of seconds. The minimal Mk8 has a real benefit here.

 

6 - No additional cost (this is my very attainable goal)

I place a value on my time of about $100 per hour - your calculation might be different - but contributing to this community is fun, so I do it pro-bono ;) - but I still value the time I contribute non-the-less.

 

7 - Aluminum guide block for heat resistance

Aluminium is a relatively efficient thermal conductor?

 

8 - All parts required included in the kit (including the spring adjustment hex key)

 I thought no cost was a feature i.e. item 6 above?

---- ~ ----

So from this list I would focus on tension adjustment for different filaments that might need more fine tuning like ABS, Nylon, Wood and PVA - but these claims would need to be supported with some evidence of the benefits of you design compared to the alternatives. The ability to access and visually inspect the top of the hot end is a real advantage. While for me this is not really an issue in practice, every now and then it would be really handy - I have on a very few occasions had to remove the extruder due to filament that was molten getting carelessly squished by the pinch arm and not being able to pull it out. 

The downside of your approach is that mid-print filament changes may be potentially slower.

So what would really trump the Minimal Mk8 is if you could add a quick release cam to unlock your tensioner - that I would applaud, very loudly - food for thought =)

[DHeadrick]

Thank you for your valuable feedback.  I value good designs that work well (regardless of vintage).  I feed that my design is an improvement over what is out there, but there is no way to prove this until I run 100+ hours of filament through my prototype.

On Tuesday, 23 April 2013 20:52:46 UTC-4, Wingcommander whpthomas wrote:

So a quick critique of you list

1 - Allows positive engagement of the gripper on the motor flat.

Not sure exactly what you mean by this one, but if you think pulling power is an issue, the Minimal au Mk8 will lift a tangled spool without slipping - it grips about as strong as the stepper can push or pull.

What I mean is if a motor shaft has a flat and the gripper has a set screw, they are only there to serve each other.  Putting those features there and then not properly utilizing them is sad.  I have read numerous posts where people have had their gripper slip on the hardened motor shaft.  I have corrected this oversight so it never becomes an issue again.  Just because yours can pull a car out of the mud doesn't mean everyone's can or will or may not have that problem in the future.
 

2 - Adjustable spring tension with release capability.

If it has the right spring tension from the start then for PLA adjustment is not necessary - but for ABS and Nylon this could be an important feature to have [TICK].  Thanks :)

 

3 - Allow better heat dissipation during long prints by having a large port for heat to convect away.

Well speaking from experience, the aluminium works as a large heat soak, air flows in from the left side through the cowling and inside the top covering and leaves through the gaps in the spring. The stepper is also cooled by the cooling block - consequently the stepper runs cool to touch even after 22hours of non-stop printing. Not sure this is a real problem in practice.

My guide block will also be aluminum and there is sufficient clearance that the potential for heat to be contained and build up is mitigated.  I'm not sure there is a real problem here either, but I don't print with ABS or Nylon yet.

 

4 - Allow visual access to the gripper wheel and even the heater inlet port.

No argument here [TICK].  Thanks again :)

 

5 - Allow scissor access in the case of a filament jam

No argument here either [TICK].  Thanks again :)

However, you can prevent blobs on the end of the filament forming by doing a quick load before unloading so in practice this is not really a problem I have encountered. The upside of the Mk8 is that manual filament changes are really quick. To do this, I tend to grip the under side edge of the extruder with my fingers, depress the lever with my thumb, push the filament down to clear any blobs, and then pull it out. It never jams, and filament changes are really quick - you to the same thing to reload - insert filament, release tension with thumb, then push it all the way down until a small amount leaves the tip and resume print. If you have two feeder tubes and the second real is already prepped, you can do a colour change in a mater of seconds. The minimal Mk8 has a real benefit here.

My design also has a quick release of the spring tension so that you can pull the filament out and insert a new one in a matter of seconds.  The only difference is you have to use the 2mm hex key that came with your Makerbot to do it.  You insert it in the hole as shown in the images in the previous post.  Same as your lever except you push on a hex key (or screwdriver or rod if you desire).
 

6 - No additional cost (this is my very attainable goal)

I place a value on my time of about $100 per hour - your calculation might be different - but contributing to this community is fun, so I do it pro-bono ;) - but I still value the time I contribute non-the-less.

I also value my time as most do.  I was disappointed at all the time I had to spend dicking around tuning the Makerbot just to get decent prints.  If there was a replacement that would fix most of the known issues in one easy kit, I would have bought it.  From what I've read, people buy/print a lever arm.  Then they source the bearing and spring (if they had to print the arm) and assemble it.  Then they may have to use a file to extend the motor shaft flat (if they have that issue).  These are all factors that are a pain in the ass and have all been addressed by this design!  And I believe it can be done at the same cost as the currently available aluminum kits (~$99) yet solve more issues at the same time.

 

7 - Aluminum guide block for heat resistance

Aluminium is a relatively efficient thermal conductor?

I should clarify.  It will be made of aluminum so that it won't warp if the extruder does overheat.  I've read about people using the PLA upgrades and having them warp under heat (some people are on their 4th replacement).
 

 

8 - All parts required included in the kit (including the spring adjustment hex key)

 I thought no cost was a feature i.e. item 6 above?

I will throw in the hex key to adjust the spring tension (because none of the keys that came with the Makerbot will fit the set screw).  The hex key costs 17 cents to the parts... I think I can handle that to ease the buyers frustration.  My goal is to offer this kit for the same price as the competing aluminum upgrade kit.
 

---- ~ ----

So from this list I would focus on tension adjustment for different filaments that might need more fine tuning like ABS, Nylon, Wood and PVA (Done - spring force adjustment is possible and easy) - but these claims would need to be supported with some evidence of the benefits of you design compared to the alternatives (agreed). The ability to access and visually inspect the top of the hot end is a real advantage. While for me this is not really an issue in practice, every now and then it would be really handy - I have on a very few occasions had to remove the extruder due to filament that was molten getting carelessly squished by the pinch arm and not being able to pull it out (I agree it is a unexpected bonus of the part layout of this design). 

The downside of your approach is that mid-print filament changes may be potentially slower (Not any slower than the lever arm style - I will post a video to demonstrate in the near future).

So what would really trump the Minimal Mk8 is if you could add a quick release cam to unlock your tensioner - that I would applaud, very loudly - food for thought =)

Thanks for your insight.  This discussion thread has really helped direct the design of this upgrade into something I think will make the Makerbot much easier to use.

[Wingcommander whpthomas]

What I mean is if a motor shaft has a flat and the gripper has a set screw, they are only there to serve each other.  Putting those features there and then not properly utilizing them is sad.  I have read numerous posts where people have had their gripper slip on the hardened motor shaft.  I have corrected this oversight so it never becomes an issue again.  Just because yours can pull a car out of the mud doesn't mean everyone's can or will or may not have that problem in the future.

Very good point [TICK here too] ;)

Just for future reference I think most call it a hobbed drive gear (hobbing is the process that gives it the little indentations that make it grip the filament).

[DHeadrick]

On Wednesday, 24 April 2013 00:22:02 UTC-4, Wingcommander whpthomas wrote:

Just for future reference I think most call it a hobbed drive gear (hobbing is the process that gives it the little indentations that make it grip the filament).

Ah... I was aware hobbing was used to produce involute gear teeth but didn't realize it was used for this.  I was under the assumption that it was a straight knurl in a groove.

Just finished a 15.5 hr print of a solid PLA block (3.375" x 3.375" x 1.5") at 100% fill unattended overnight with no defects.  Between this prototype and the previous one that was installed (same design but different spring layout), the design has logged around 70hrs of printing.  Everything was cool to the touch while it was still printing at the 15hr mark (I touched the motor, cooling block, heatsink).

[TobyCWoods]

I've been curious since you started posting. At first I was a bit skeptical... kinda thinking along the same lines as Wingcommander... but... One thing (of many) I learned after working with designers and MEs for over 25 years as a PM is the concepts of manufacturing costs and part simplicity. I do not recall ever seeing the actual parts you have in there, but if all is equal regarding it's reliability and it adds to ease of use, the factor I'm very curious about is how easy is it to make. If it can be made faster and cheaper... 

[DHeadrick]

One thing I adhered to while designing it was to ensure that all the processes could be easily done by machining.  Because a 3D printer can always make things which can be machined but not the other way around.  Also, I wanted the minimum number of machine setups (because those are the steps that cost the most money - the machine will happily work away, but it's when it stops to wait for human intervention that you waste time/money).  The assembly is very easy to put together and the parts are all contained so they cannot explode and get lost during assembly/dis-assembly (ever try to re-assemble a mechanical egg timer - I did once).  Actually, the unit can be shipped assembled, so that it is just a bolt on replacement.  And it will fit in a bubble-wrap envelope so shipping costs would be minimal to other countries.

[DHeadrick]

I am hoping to have them up for sale in a week.  I am closing in on 100hrs of successful printing with the current design and I'm going to install it in another Replicator2 tomorrow so that I can expedite the validation process.  I really want to make sure that there are no problems before I begin offering it up for money.

I am really excited to get this out there so more people can immediately start to use their printers without having to perform multiple tweaks just to get up and running.  I know how frustrating it can be when you first receive the printer and start printing and something goes wrong.  Your first thought is that you set something up wrong or entered wrong parameters.

I will be sure to post as soon as I have more details.

On Sunday, 28 April 2013 09:32:11 UTC-4, Steve Bartholomew wrote:

I'd buy one today.

On Friday, April 12, 2013 3:41:55 PM UTC-4, DHeadrick wrote:

I have read many posts about issues people have experienced with the Makerbot 2 shortly after receiving one.  I was only reading for interest until a few of these issues began to rear their head.  It is disappointing to have spend $2500 on a 3D printer and not be able to use it after as little as a week of printing.  My problems showed up within 12 hours of elapsed printing time!  This is not acceptable.  Unfortunately, my printer was no longer capable of printing out good enough parts to try any of the many fixes available on the web.

Luckily, I had access to other methods of manufacturing and have been able to design a few replacement parts which have completely resolved a few issues.  Here are the issues I found as I took my extruder head apart:

1)  The filament tension mechanism is an appalling design and it is obvious that it will never allow for consistent printing of quality parts.
2)  The gripper wheel that advances the filament does not engage the flat on the stepper motor shaft.
3)  The excess heat from the melter is allowed to travel away through the heat sink (where it should) but also the stepper motor housing (where it should not).

I have a designed upgraded parts which can be swapped into place in under 10 minutes which improve the consistency of the extruder.  Here are some sample images of the assembly in place:

The features this design incorporates are:

1)  It moves the stepper motor rearward slightly to allow the gripper wheel to engage the flat on the shaft.

2)  It adds an auto-tensioning system which includes a bearing and never needs adjustment.

3)  It also moves the fan closer to the heatsink so that the airflow must flow through the heatsink instead of past it through a large gap.  Also, the motor being spaced off from the heater bar reduces the amount of heat that is transferred directly to the stepper motor.

The materials for these parts are not set in stone yet because only prototypes have been built and tested so far.  The plan is to make the bearing mounting components from stainless steel and the filament guide block from either aluminum or delrin for durability.  The intention is to find out if there is any interest for a kit like this.  The kit would include ALL the parts required for the upgrade (not just a list of components and suggestions where they can be purchased separately).  None of the parts will require a functioning 3D printer to install which means you can get a non-functioning or misbehaving printer back up and running easily.

What I need to know is if there is any interest in this upgrade so that I can get a feel for if it is worth making.  The prototype has solved my issues and I am happy.  I want to allow others to enjoy the full functionality of their investment.

[David Celento]

Hi David H,

This all sounds quite promising. 

Out of curiosity, what's the longest print you've done?  As you know, heat soak is not a trifling phenomenon; thus, short tests may work well, but not longer ones.

Have you had a whack at the Cells Bowl by Dizingof?

[John Watson]

still weirded out by the  points you are making  as there are already upgrades that are proven and functioning great

really don't understand  what this upgrade has over the wingcommander spring tensioned version is there a way to enlighten us as to the advantages of this upgrade over the other versions

as you seem to be completely ingnoring all the hard work that has been done initially to get a high percentage of rep2 owners printing happily

please dont think me knocking your design in any way I am just intrigued as to the benefits of both :-)

--

[DHeadrick]

The longest print I have done so far is 15.5 hrs.  I started it as I left work one night and it ran unattended and was still printing when I came in the next morning.  Everything was cool to the touch and there were no issues.  I have printed a few 10+ hr prints in both clear PLA and white PLA.  Tomorrow I'm going to print for the first time with black PLA.  I'm curious how it handles the variations between the different colours.  The second Replicator2 that I will install the upgrade into tomorrow routinely prints 20+ hour jobs.

How long typically before heat soak rears its ugly head?

I would love to print that cell bowl but have had no luck slicing it.  I've tried Makerware and Repetier and both have issues (Repetier uses both Skeinforge and Slic3r).  I don't have ReplicatorG configured to the geometry of my bot yet.  Can you give me any tips?