You are not logged in. Please login or register.
SoliForum - 3D Printing Community → Filastruder → Filament Extruder - Convert pellets to filament
just for information ..learn from my error... a filastruder is not a 3d printer.... if abs vapor is harmless from a solidoodle.... i don't think it's the same for an extruder that melt a lot more abs
i took a terrible headache for two days after my first 6 hours of running my filastruder..
so when you use it.... open area well ventilated.... be smart 
Tim,
and anyone else who is interested, here are some pictures and a quick video,
currently I haven't installed the fan (was missing from the kit, but don't worry as I have plenty) and guide, but I'm not to bothered as only running through contamination period for now.
A quick measurement shows 3.0mm +/- around 0.04mm, I used a 2.65mm drill bit that I had lying around!
Looks to be pushing out about 120mm a minute of filament,
https://www.dropbox.com/sh/1w27dxxrcsra88a/gzLJk78Odx
Cheers,
Rob
Another quick question, regarding temperature, does increasing the temperature increase the diameter or is it the other way around?
Cheers,
Rob
elmoret, first post wins it!! Classic!
elmoret, first post wins it!! Classic!
To be fair, they did register on the 24th, maybe they've been waiting...
How has the coloring of the filament Via master batch been going?
Another quick question, regarding temperature, does increasing the temperature increase the diameter or is it the other way around?
Cheers,
Rob
Higher temperatures results in smaller diameter, typically.
I have been wracking my brain trying to dream up an easy way to do spooling and make the output even more consistent (I wanted to make it the 1000th post, but I had not had any lightbulb go on last night). This morning an idea did come to me. Here are my thoughts:
The filament needs to drop a certain distance to the floor. The weight of the filament creates some pull on the extruder.
A winder that winds up the filament can change the "pull" on the extruder.
A winder would need more torque than could be tolerated as "pull" on the extruder.
One solution already being explored by others is to have a two stage setup that separates the pull on the extruder from the winding torque. This requires measuring the filament diameter to regulate how much pull.
Here is what I came up with for an alternative approach that has only one stage:
Let the filament drop off the table into the air.
Do not let it touch the floor, but loop it back up to a spool under the table that has a small motor.
Regulate the spool motor by measuring the "weight" of the suspended filament.
By making the weight constant, the "pull" force will be constant.
The weight can be measured by the guide that the filament rolls over the edge being on a "spring" or "counterbalance" arrangement.
As the weight increases, the balance will drop some, and a light beam interrupter arm can turn the winder motor on and off.
By adjusting the spring, or counterbalance weight, or the interrupter position, various "pull" forces can be achieved.
By controlling the cooling direction on the filament as it leaves the extruder (like cool from the bottom), a slight bias can be forced to make the filament want to curl in the direction of the winder spool.
The filament is not going to be wound very tight on the spool. This may be ok, but if not, it is easy to make a quick spool to spool transfer arrangement to rewind it tighter. If this approach is used, then a fancier winder spool can be made, since it is only made once. For instance a 3D printed spool with the drive teeth for a worm drive on it, and a printed bearing right in the spool sides.
I do not yet have a 3D printer yet (I'm thinking I will make a delta later this year), so I can not experiment with this 3D printed spooler idea yet.
Comments?
just for information ..learn from my error... a filastruder is not a 3d printer.... if abs vapor is harmless from a solidoodle.... i don't think it's the same for an extruder that melt a lot more abs
i took a terrible headache for two days after my first 6 hours of running my filastruder..
so when you use it.... open area well ventilated.... be smart
I find this depends a bit on the particular formulation used - the gray pellets seem to stink a bit more than the natural. Definitely use in a well-ventilated area, though!
Great work going ahead here. I'm looking at building the same sort of thing so all the ideas here are really helpful. I would love to help back the KS once it's up. Thanks for all the hard work you've put into this.
You were the 1000th post! Shoot me an email to claim your free roll of filastruder filament! filastruder at gmail.com.
Comments?
One of the issues here is that the loop of filament only weighs a few grams. Thus, if you control by weight, you need sub-gram resolution in your force sensor or spring setup. That's tough to do, rollin' on the budget we're aiming for.
I still have a lot of hope for those $14 photodiode arrays, just haven't had a chance to play with one yet.
I think there is a minimum amount of pull required to affect the diameter, which is greater than the pull from gravity. As filament spools on the floor, the length between the floor and extruder changes on its way down, and as it coils around, but the diameter stays pretty constant. When I ran the filament through a dropped loop to the puller the diameter stayed within tolerance as the loop went up and down. The challenge was getting the loop to stay at the same length.
I'm working on a free spinning roller guide to keep the filament straight for a bit coming out of the die before it drops to help maintain a round cross section. Any changes in weight of the hanging filament would pull down on the roller probably more than out from the die due to the stiffness of the filament. If the filament makes a straight line or shallow drop from the die to the puller, then the diameter is directly controlled by the force of the puller.
Whatever method I come up with to regulate the speed of the winding could be used to run a spool only, rather than two stage. However using the stepper has the advantage of being able to count the filament as it's spooled. A stepper and an arduino are actually cheaper than any mechanical wire counter I've been able to find. Also the first few feet of filament is often not usuable. It's hard to guide it to the winder without stretching it or getting kinks along the way, so it helps to be able to let everything stabilize before clipping it and tucking into the spool.
I looked up catenary, but I'm not sure it applies. It refers to the way a chain or cord drops, which is completely flexible. Filament is something between a solid rod and a rope. I'm sure there is a category of physics to describe how that moves, but I don't know what kind of keywords to look up.
A quick measurement shows 3.0mm +/- around 0.04mm, I used a 2.65mm drill bit that I had lying around
Good info, thanks! Let us know how it prints.
Dennis wrote:Comments?
One of the issues here is that the loop of filament only weighs a few grams. Thus, if you control by weight, you need sub-gram resolution in your force sensor or spring setup. That's tough to do, rollin' on the budget we're aiming for.
I still have a lot of hope for those $14 photodiode arrays, just haven't had a chance to play with one yet.
Actually, a balance is easily able to measure a few grams. The key is to have a friction free pivot, which is a no brainer with a knife edge pivot. It just consists a 3 blocks cut at 45 degrees and a knife edge, like the straight 4 inch wallpaper stripper blades you get at the hardware store. You just set the assembly attached to the blade on top of the blocks and there you have it. Picture attached (if I did it right):
Actually, a balance is easily able to measure a few grams. The key is to have a friction free pivot, which is a no brainer with a knife edge pivot. It just consists a 3 blocks cut at 45 degrees and a knife edge, like the straight 4 inch wallpaper stripper blades you get at the hardware store. You just set the assembly attached to the blade on top of the blocks and there you have it. Picture attached (if I did it right):
How do you make that? Print it all out? Blocks of wood and assemble it? How does the counterbalance slide? How is it affected by inconsistencies in the filament - when extruding from the die, sometimes the melt will curve up or down slightly, which could easily exert a fraction of a gram of force on the balance. Finally, if you're going to use photodiodes, why not just use them directly on the the filament at the bottom of the loop?
Just so all of you know this is what 50 pounds of PLA from Open Source Printing looks like:
Bob Teeter
Dennis wrote:Actually, a balance is easily able to measure a few grams. The key is to have a friction free pivot, which is a no brainer with a knife edge pivot. It just consists a 3 blocks cut at 45 degrees and a knife edge, like the straight 4 inch wallpaper stripper blades you get at the hardware store. You just set the assembly attached to the blade on top of the blocks and there you have it. Picture attached (if I did it right):
How do you make that? Print it all out? Blocks of wood and assemble it? How does the counterbalance slide? How is it affected by inconsistencies in the filament - when extruding from the die, sometimes the melt will curve up or down slightly, which could easily exert a fraction of a gram of force on the balance. Finally, if you're going to use photodiodes, why not just use them directly on the the filament at the bottom of the loop?
It could be made out of wood or printed. The balance beam could be a threaded rod, and the sliding weight could be a nut or a threaded weight, or just a "U" shaped piece that hangs off the threaded rod. It is just set once to the right position so that the filament does not hit the floor. It can be made out of just about anything lying around. The contact surfaces for the knife edge should be hard though, like a small piece flat metal or hard plastic glued to the block, rather than just a soft wood surface.
The inconsistencies in the filament should be within a reasonable tolerance for force of weight measured. You would want to have it far enough away from the extruder tip to have it able to be pulled down by gravity in the loop. The setup won't really work well until enough filament has been extruded to make a whole loop to the winder spool. Then the balance weight could be set to the loop weight if needed. But I have to admit that I have not played with any filament yet, so I don't have that intuitive feel for its physics.
One of the benefits of this is that if the weight is too low or high, it will always register that way. There is no sweet spot that the filament has to stay within to be measured. The place that is being measured can also wonder around on the roller a few inches without affecting the measurement. The photodiode is also contained with the main housing and close to the electronics, so it is neat and tidy. It does not have to be floating round in mid air at the bottom of the loop.
Why not just have the filament go through a 3x3 window opening on it's way to the spooler. On the top edge of the window you put a contact switch to turn the spool motor off, on the bottom edge you put another contact switch to turn the spool motor on. That's what is used in industrial re-reelers that we use where I work, but since it's usually steel we can just use wire contact to turn the motor on or off. You may have to rig up your own contact with two blades of paper thin copper to use the lightest pressure to turn on and off the switch, but I still don't see why it wouldn't work.
The problem with such a sensitive scale is that the filament doesn't always go straight out of the die. It can twist a little bit, and set with a slight wave every now and then. If there is a dip in the filament it could push down the scale as it passes over rather than the scale being pulled. Using a photodiode directly can be problematic for the same reason. Sometimes instead of dropping lower, the loop might widen.
I would rather have the feedback mechanism close to the extruder, but I'm thinking that having the bottom of the loop run through a guide roller that gets pulled up and let down might be the most reliable. I'm hoping it can be done with a loop short enough that the winder and extruder only need to be about 12" above the table so the measurement thing doesn't need to go on the floor or something.
The ideal method would still be an optical mouse sensor at the extruder. However the best arduino support is for discontinued sensors that need to be harvested from mice. There seems to have been some work done with the current ones you can get from Mouser but I don't understand it well enough to try and use it.
One other possibility is an optical encoder on a roller that the filament slides over as it bends toward the floor. That would be an IR LED and sensor pointed at a black and white disk mounted on one end of the roller.
The problem with such a sensitive scale is that the filament doesn't always go straight out of the die. It can twist a little bit, and set with a slight wave every now and then. If there is a dip in the filament it could push down the scale as it passes over rather than the scale being pulled. Using a photodiode directly can be problematic for the same reason. Sometimes instead of dropping lower, the loop might widen.
What about having a lightly weighted roller at the bottom of the filament loop? It could be heavy enough to keep the filament straight, but not so heavy as to overly "pull" the filament from the extruder. It would look like those high wire bicycle rider toys with the weights hanging down by wires on either side of the wheel to lower the center of gravity below the filament, making it stable. It would also be easier to detect its position, rather than a thin wiggly filament.
What I've been working on today:
Nice experiment!
How is it working for filament quality?
Could you replace the puller with a direct drove of a spool and kill 2 birds with 1 stone?
Nice experiment!
How is it working for filament quality?
Could you replace the puller with a direct drove of a spool and kill 2 birds with 1 stone?
If you do the spool without a puller, you still need something to keep it tight, if that makes sense - like pull through a springloaded clip or something. I don't see why not, though.
Filament diameter seems to be the same as previously. Too early to tell definitively, though.
Dennis wrote:Nice experiment!
How is it working for filament quality?
Could you replace the puller with a direct drove of a spool and kill 2 birds with 1 stone?If you do the spool without a puller, you still need something to keep it tight, if that makes sense - like pull through a springloaded clip or something. I don't see why not, though.
Filament diameter seems to be the same as previously. Too early to tell definitively, though.
Yes, or something like a spring clothes pin with felt lined jaws perhaps. It does not have to be extra fancy to work. ;-)
Brad, look away! I call it the ponystruder. Kickstarter to follow.
SoliForum - 3D Printing Community → Filastruder → Filament Extruder - Convert pellets to filament
Powered by PunBB, supported by Informer Technologies, Inc.
