76

Re: Filament Winder

I've been busy and I've got all the parts printed and I even made some that IanJohnson hasn't posted yet so I can play around. I don't want to post my versions till after IanJohnson does because I don't want to derail his project or his direction. However if Ian want's my files just let me know. I'm happy to share. http://thingiverse-production.s3.amazonaws.com/renders/aa/3a/de/1b/1d/image_display_large.jpg

77

Re: Filament Winder

drandolph wrote:

I've been busy and I've got all the parts printed and I even made some that IanJohnson hasn't posted yet so I can play around. I don't want to post my versions till after IanJohnson does because I don't want to derail his project or his direction. However if Ian want's my files just let me know. I'm happy to share. http://thingiverse-production.s3.amazonaws.com/renders/aa/3a/de/1b/1d/image_display_large.jpg

You're not the DRandolph I'm thinking of are you? I'll take a guess and say yes. The last time I saw anything from you was when you where modifying your laser cutter I believe. Nice to see you around here too big_smile


Also, these filament winders are looking nice, It will probably be the first things I build once I get my 3D printer big_smile

78

Re: Filament Winder

Lasers and 3D printers and killer robots DRandolph? Yep that's me. I like the filistruder project and the winder project so its going to be a lot of fun to play and contribute around here.

79

Re: Filament Winder

One change I haven't got around to yet is the servo mount.  I couldn't really get it screwed in very well, so I want to go ahead and wrap it all the way around the top of the servo.  Also I posted a mount for the photo resistors which is untested, but should work ok.  The guides are made from some wire run inside PTFE tubing that is 4mm OD 2mm ID.  Forgot to include that in the BOM.

At some point I'd like to come up with a nicer control panel, maybe a couple of LEDs to indicate if you are setting the guide's min or max.  I'm trying to avoid adding an LCD to keep things simple, which might just mean a bunch of knobs and blinking lights like a control panel on the Enterprise.

80

Re: Filament Winder

Thanks for the update! I'll print out your tensioner parts and photo resistors gate tonight and see how it fits. On my little panel I put one master power switch and a power indicator. Nothing special. But to keep things simple what if you put in one RGB LED. That could give a lot of information without much complexity. Red/Red blink/Blue/Blue blink/Green/Green blink.

81

Re: Filament Winder

There is a lot of static build up on the filament as it spools.  I'm not sure if it comes from the extrusion or from the coils sliding against each other as they lay down on the spool.  Any ideas how it might be discharged along the way to cut down on dust?  It shouldn't be much more complicated that just using a filament wiper on the printer.

82

Re: Filament Winder

Get some ball chain like they use to hold pens down at the bank and have the filament go through it like a curtain and then ground the chain. Or if you want to be really cool get some of that conductive filament and make your idler out of that and then ground it.

83 (edited by bteeter 2013-05-14 21:37:59)

Re: Filament Winder

How about something simple like a metal ring in the filament guide off the servo for layup direction control.  Ground it and you have removed the static source.

Bob Teeter

84

Re: Filament Winder

bteeter wrote:

How about something simple like a metal ring in the filament guide off the servo for layup direction control.  Ground it and you have removed the static source.

Bob Teeter


I had been considering putting a washer into the filament guide to reduce friction as the spool drags the filament through it.  That might also be another source of the static.  Would it be enough to run a wire from the washer to one of the screws holding the servo mount to the wood, or would it need to do something like splice into the servo's ground wire?

85

Re: Filament Winder

Here are a couple of time lapses of the winder and the filament loop.  Getting the loop set up can be tricky, I found that the key seems to be moving the extruder back and forth to keep the fiament from getting pushed out of position at the nozzle until the loop can settle out.  A posted a more detailed description at the blog - http://solidoodletips.wordpress.com/201 … -the-loop/

86

Re: Filament Winder

Have you thought of some kind of guide that you can put down from the nozzle so it will keep the filament straight until it cools enough then starts the loop?

87

Re: Filament Winder

neat. 


Tyler McNaney must be puzzled as to how he is going to make this fit in his machine...

SD2 with E3D, SD Press, Form 1+
Filastruder
NYLON (taulman): http://www.soliforum.com/topic/466/nylon/

88

Re: Filament Winder

I've seen a few other extruder designs where they say "and then there will be a spool here which will turn and take up the filament"  as an afterthought to tackle once the extruder is done.  It's a lot more complicated than that.  Tyler's has the extruding filament turn the spool by pushing on it.  That only begins to work if the filament is solid when it comes out, which is great if he has cracked that.  I imagine it will be a loose, jumbled mess.  Makibox was planning on extruding downward, and up around a roller on an arm like my tension adjustment to regulate the speed of the spool.  They will quickly discover that asking the filament to do any mechanical work as it is extruded is a dead end.  It looks good enough on paper to start taking $150 preorders though.

Have you thought of some kind of guide that you can put down from the nozzle so it will keep the filament straight until it cools enough then starts the loop?

I've tried various kinds of guides when extruding horizontally, and they never worked because the slightest drag against the bottom of it from gravity, or the sides from sideways motion of the filament cause enough push back against the nozzle to screw everything up.  I haven't tried it extruding down however.  I think a lot of those problems won't be an issue while extruding along with gravity, so it is worth a try.

The most reliable solution would be a nozzle of some kind that will extrude filament that is already solid.  This would basically be a printer hotend in reverse, or at least the barrel, PEEK and PTFE portion.  It needs to be just a precise as well.  I'm still working on it, and nothing that relied on threading PTFE to a brass barrel has worked because the PTFE was too soft.  I have some PEEK on the way and will try making something that is basically a J-Head cold end with a 1.75 ID.   I screwed a Makergear PEEK onto the extruder once, and it worked, but the filament was too big because the ID was something like 2mm.  So I am trying to reproduce something like that in 1.75.  If it works, you should be able to do just about anything you want to the filament and it will still be consistent.

89

Re: Filament Winder

IanJohnson wrote:
bteeter wrote:

How about something simple like a metal ring in the filament guide off the servo for layup direction control.  Ground it and you have removed the static source.

Bob Teeter


I had been considering putting a washer into the filament guide to reduce friction as the spool drags the filament through it.  That might also be another source of the static.  Would it be enough to run a wire from the washer to one of the screws holding the servo mount to the wood, or would it need to do something like splice into the servo's ground wire?


I would tie it to earth ground.  The static electricity is just like walking across a carpet in winter.  ZAP-POW.  Especially when
I got to zap my brother.  What fun. I always could out run him.  If you tie to earth ground you will remove the problem.  I would also put a washer on the output of the initial puller so that there are 2 spots to remove static cling.

Bob Teeter

90

Re: Filament Winder

Ian - I was wondering if you printed the bins that your extruder and winder are on in the video.  If you did great HOW? If you did not does your wife know?

Bob Teeter

91

Re: Filament Winder

I bought the bins for the machines.  I figured the Sam Adams boxes weren't quite classy enough for a 3D Printing meetup.

92

Re: Filament Winder

Ok - to get your machines run perfectly at the meetup.  1 . give the bins to your wife she will love your thoughtfulness.  2. put some ice in glass bottles and put them in the Sam Adams boxes.  Around the edges.  Then when you setup the guys will hear the clinking and ask about that noise.  Just tell them that it is just some bottles with ice in them.  Put the equipment on top and start the demo.  After a while the guys will ask about the bottles again and just say that you cann't move anything until the setup runs perfectly.  You will have more help fixing it and turning it up than you can imagine.  After it is all done then take it apart and when the guys ask about the bottles in the Sam Adams box.  Open them up and give them each one.  Also have the equipment already loaded so that you just pick up the boxes and say "BYE".  tongue

Bob Teeter

93

Re: Filament Winder

OK, I've been waiting on parts and I just got them in today so I'm back on task for making the winder work. I finished the filastruder and even made a nice case for it. Now I'm soldering up a shield for my arduino and I got the laser assembly built so I'm getting really close to getting it all up and running. I can't find if I asked this before but you had a switch for setting min-max on the servo, what pin was that to? I also noticed on your breadboard that the hall sensors are one row off on the breadboard. http://i.imgur.com/Uek3gZR.jpg

94

Re: Filament Winder

plus one on getting the electronics up and running,

I'm expecting the motors to arrive this week and would love to get these winder up and running!

drandolph if you have any other wiring diagrams I'd appreciate them! ~(loving the shield btw)

Cheers,

Rob

95

Re: Filament Winder

I noticed the issue with the sensors when I was putting my breadboard together as well.  I have the guide set button on Pin 12.  I also added a button on Pin 1 for setting the setpoint for the spool tension PID.  You hold the button down and move the arm to the point you want it to maintain, then let go of the button.  I'm attaching the firmware as it stands for me.  Some of the PWM pins are probably moved around.  I was having trouble with the motors not working for some reason unless I moved them around to different PWM pins.  I tested all the PWM with the servo and they still worked, so I don't know what the issue was.  It's easy enough to change them around in the pin assignments to however you want to wire them up.

I'm not sure if I mentioned it before, but you will need to download the PIDl library.  There is a page about it here- http://playground.arduino.cc/Code/PIDLibrary

Polarity matters with the hall sensors, so you will need to test them to figure out which way the magnets need to point.  I stacked two magnets together to get them to stick out far enough to run close enough to the sensors.  To test them, you can add a line like

Serial.println(digitalRead(hall_a_Pin));

and watch for the value to change in the serial monitor.  I included a serial monitor function to make it easier to comment it in and out as you test stuff.  One thing to keep in mind is that the PID won't work very well if you are outputting data to the serial monitor, because it slows the program down so much.  At least it does at 9600 baud, I suppose it could be run faster.  I don't know if the Uno has an upper limit.

I uploaded a new sensor mount yesterday, I don't know if you saw.  The previous one had the laser too close to the sensors, and I wasn't getting very good reliability.  The new one has a notched base, and the sensor holder, laser holder, and guides can slide back and forth along it so you can find the best positions for all of them.  I would recommend screwing the base down to a small board to give the whole thing a little weight and stability.

I still want to add another button and switch.  The button would be for calibrating the photo sensors.  You would hold the button, then wave a piece of filament past them.  It would save the highest value read for each sensor, using that as the ambient, and save the lowest value for each sensor, making that the shaded value.  Then it would figure a value somewhere between those two points for being the trigger value.

The switch would be to set whether the speed knob was for the pull up speed or the let down speed.  Also I want to add in the RGB LED to indicate what setting mode you are in, and whether the filament is being pulled up or let down.

I changed the puller as well.  I kept having trouble with the filament slipping behind the bearing, so I switched to a MK3 jigsaw replacement for better filament guidance.  I changed the puller mount to be flat on the front with NEMA17 holes and the holes for the gearmotor.  This way you can attach whatever NEMA17 compatible direct drive extruder you want.

96

Re: Filament Winder

Hi Ian,

Have you got any ideas about the spec. of the photo resistors you used?

I'm trying to pick some up but there is a large range of resistance ranges to choose from!

Ideally I'd like to source the same ones that you used yourself.

Many Thanks,

Rob

97

Re: Filament Winder

I don't actually know the specs of the photo resistors.  They came in the Make Magazine intro to Arduino kit.  I'm guessing that it would be similar to this one from Adafruit - http://www.adafruit.com/products/161

I got around this morning to setting up my photocells on a bit of strip board.  I have the resistors with the photocells, since space is getting tight on the bread board.  There are 4 strips with a 4 pin header at the end.  Strip 1 is + with a leg of each photocell on it.  Strip 2 is GND which gets a leg from each 10k resistor.  Strip 3 is the other leg for the top photocell and its resistor, and Strip 4 is the other leg of the bottom photocell and its resistor.

http://solidoodletips.files.wordpress.com/2013/05/photo-copy.jpg

I made sure that the photocells extended far enough to clear both the resistors and the thickness of the mount.  I should have also cut a wider piece of strip board so there would be room to screw it onto the back of the mount rather than just taping it.

98

Re: Filament Winder

I didn't get the firmware attached to that previous post, so trying again-

Post's attachments

Winder_PID_v3.ino 8.79 kb, 20 downloads since 2013-06-04 

You don't have the permssions to download the attachments of this post.

99

Re: Filament Winder

Hello Ian,
Its been very interesting following your work on the winder, I was just wondering if you had made any recent progress at all that you can share?

100

Re: Filament Winder

I've sent a PCB design to OSHPark which I should have back at the end of the month.  It incorporates the arduino and has headers for all of the connections.  It's the first time I've done this, so it might be great, or ir might simply not work.   

At the moment I have the winder set up to use a stepper as the puller and draw the plastic directly from the extruder.  I hope to get the code loaded and have it running that way this morning.  This didn't work well enough for me the first time because the feedrate from the extruder is too variable.  There needs to be a feedback loop from the output for it to work, which is the way the pros do it.  There is a laser micrometer reading the diameter at the die, and the controller uses the reading to adjust the speed of the puller to keep the diameter within tolerance.

I've been trying to connect a linear CCD  to the arduino, but I haven't been able to get it to work.  I've read that they are more vulnerable to static, so I need to get another one and try again and be better about wearing a wristband.  If I can get a laser and CCD to work as part of the winder, it would be cool to use that as a basis for making a standalone laser micrometer that doesn't cost thousands.

Another method for measuring the filament is to use a simple LED and phototransistor.  As the filament changes diameter, the shadow changes size and alters the output of the phototransistor.   It might be hard to calibrate it to an actual size measurement, but I only need to see changes so the motor can adjust and try to minimize the change.  The challenge with that method is accounting for movement of the filament toward and away from the sensor.  I might be able to get it to work if I can place a phototransistor on both sides of the filament.  If the reading from one increases and the other decreases, then I know the filament has move sideways rather than growing.