Shop OBEX P1 Docs P2 Docs Learn Events
A New CNC Build - 3D Printer - El Cheapo - Page 8 — Parallax Forums

A New CNC Build - 3D Printer - El Cheapo

15681011

Comments

  • idbruceidbruce Posts: 6,197
    edited 2014-04-30 06:35
    ErNa

    Thanks for the reply.

    I suppose I should have been more specific. I apologize for that.

    My main goal is to either free up the Propeller resources or gain computational speed. I am wondering how this applies.
  • ErNaErNa Posts: 1,738
    edited 2014-04-30 08:35
    Maybe I can give better answers soon as the tbone board seems to be successfully funded http://www.tbone.cc/ So I will have one of the first boards and start to implement the drivers to the stepper on a propeller. With this board a lot of housekeeping like acceleration can be outsourced to the motion control chips
  • PublisonPublison Posts: 12,366
    edited 2014-04-30 08:46
    idbruce wrote: »
    Jim



    Have you ever used one of the FPU processors? I am just wondering how beneficial it would be to invest in one or more for all the floating point calculations.

    Bruce,

    I only used the 32 bit device with a GPS, (Application Note 36), which is pretty much a stand alone application without a microcontroller.

    Jim
  • ErNaErNa Posts: 1,738
    edited 2014-04-30 11:31
    oops, I just saw: such a chip is about 20$? Must be really fast?
  • idbruceidbruce Posts: 6,197
    edited 2014-04-30 16:33
    oops, I just saw: such a chip is about 20$? Must be really fast?

    Yeah, it isn't cheap. I don't know about the speed, but it sure has a lot of functions. I think it would remove quite a bit of burden in all the computations that must be made. However I am truly not sure.

    I think I will initially try to get by without it and if necessary, just buy it at a later date.
  • babinda01babinda01 Posts: 54
    edited 2014-05-01 04:19
    Hi Guys
    I was just doing a spring clean on my hard drive, and came across 2 projects by other people, that I had downloaded. The both have to do with the prop chip and 3d printers and CNC.
    I have absolutely no idea of what state the projects are in, but they may be of some use.
    I also do not know the people who wrote them, and I hope I don't offend anyone by posting the links here.

    The links to the projects are:
    https://github.com/gumbo/tihaerl
    https://github.com/rosco-pc/propmaker


    Best regards
    Andrew
  • idbruceidbruce Posts: 6,197
    edited 2014-05-01 04:41
    babinda01

    Thanks for the links, no offense here.

    I took a peek at some of the source code and found some code that interests me, whether it is useful or not, has yet to be determined.

    Anyhow thanks

    Bruce
  • Mark_TMark_T Posts: 1,981
    edited 2014-05-01 05:51
    ErNa wrote: »
    oops, I just saw: such a chip is about 20$? Must be really fast?

    There not particularly fast, they are PICs programmed with floating point routines I believe. 14us for
    a double multiply, 38us for a double divide.
  • idbruceidbruce Posts: 6,197
    edited 2014-05-01 10:10
    Instead of getting busy with the actual construction of the printer, I decided to finish the design of my extruder, since I still have some lingering pain and the extruder design still needs completion. After a brief discussion with Brian_B about his direct drive extruder, I decided to design and build a direct drive extruder, instead of a geared down extruder. The main benefit will be the speed obtained at the cost of losing torque for pushing the filament.

    Before diving deep into the design, I knew that one of the main considerations of building a direct drive extruder, would be the drive gear used to grip the filament, so I did a little research. During my research, I came across this article (http://airtripper.com/1676/3d-printer-extruder-filament-drive-gear-review-benchmark/), which discusses the review and benchmark testing of various drive gears. I would consider this article a must read for all those using or building direct drive extruders. While you are there, you may also want to read about his "Extruder Filament Force Sensor", which can be found here (http://airtripper.com/1338/airtripper-extruder-filament-force-sensor-introduction/).

    Considering that I am currently uncertain where the filament spool will reside, I know that I want as much torque as possible for this extruder, however I also want it to be as fast as possible. Without doing some bench testing of my own, I will simply be guessing which drive gear to select for building the extruder, however the choices have been narrowed down to a MK7 or a MK8. As mentioned in another thread, I will be supplying 48VDC to the motor drivers, which in turn should provide plenty of speed for the printer, providing that the X axis cooperates. The X axis has the burden of pushing the Y axis around, so a lot of the printers capability will depend on just how well the X axis handles the load. I believe that basically most of the parameters of this printer will depend upon the speed at which the X axis can accurately move the Y axis, however I could be wrong, and it may depend on how much torque I have available and at what speed I can push the filament. As also mentioned in another thread, I will be using (4) - Applied Motion High Torque NEMA 17 motors, with X, Y, and Z axes utilizing their HT17-275D model (specs. - http://www.applied-motion.com/products/stepper-motors/ht17-275) and their HT17-075D for the extruder (specs. - http://www.applied-motion.com/products/stepper-motors/ht17-075). For a NEMA 17, this motor is pretty hefty and it has quite a bit torque.

    As I am sitting here pondering my selection, it has dawned on me that another variable must be considered, especially when going after speed. Besides the limitation or ability of the X axis to reliably push the Y axis around, the speed at which the filament can be processed by the hot end also comes into play. Luckyily I purchased enough raw materials to experiment with making several hot ends, but I would imagine there is a maximum speed at which the plastic can be reliably extruded, regardless of the extruder motor speed or torque.

    Okay then.... I suppose I will stick with torque and select the MK8 drive gear.
  • idbruceidbruce Posts: 6,197
    edited 2014-05-01 16:22
    After doing a little more research in an attempt to find the true creator of the MK8, because I want the authentic gear, I discovered this place (http://reprap.me/extruder/mk8.html). Although not 100% certain, I assume this to be the original MK8 creator and distributor.

    Instead of ordering on the site, I went to eBay to look around. On eBay, these folks have several listings, which offer different shipping rates, and of course I chose the cheapest, because this gear is coming from Br
  • idbruceidbruce Posts: 6,197
    edited 2014-05-02 05:01
    As a side note for those folks that are truly following along, here is some pertinent information pertaining to the MK8 drive gear. After purchasing the drive gear, I sent an email asking a few questions about the gear. The content of the original email and the reply are shown below:
    My email:
    I just purchased an MK8 drive gear from you folks off of eBay, because I am designing and building a direct drive extruder. However I am finding no extruder design documentation pertaining to this drive gear. For instance, I have two very important questions.

    1. What is the exact required distance from the center of the drive gear to the center of the filament core when designing an extruder for 1.7 filament?
    2. What is the exact required distance from the center of the drive gear to the center of the filament core when designing an extruder for 3.0 filament?
    Any time that you may put forth into answering these two questions will be greatly appreciated.
    Their reply:
    Thank you for asking.

    The MK8 drive gear has an effective diameter of 7mm which in case of 1,7mm filmaent will give a theoretical offset of approximately 4.35 mm.
    Depending of applied pressure there is some deformation of the filamet which is in the order of 0.1- 0.2 mm approx. (D)

    Calculates as (7+1,7) / 2 - D

    The same goes for 3mm filament (actual measure 2.85mm)

    EDIT: I believe there is an error in their reply.
    EDIT: After examining this closely, I believe the correct formula would be similar to the following:
    (ED / 2) + FR - FD

    ED = Effective Diameter = 7MM
    FR = Filament Radius = 1.7 / 2 (for 1.7MM filament) or FR = Filament Radius = 2.85 / 2 (for 3.0MM filament)
    FD = Filament Deformation = Approximately from 0.1MM to 0.2MM depending upon idler pressure.

    *For 1.7MM diameter filament, this equates to a distance of 4.15 through 4.25MM from center of motor shaft to the center of the core.
    *For 3.0MM (2.85MM actual) diameter filament, this equates to a distance of 4.725 through 4.825MM from center of motor shaft to the center of the core.
  • idbruceidbruce Posts: 6,197
    edited 2014-05-03 06:22
    Concerning the previous post which pertains to alignment of the MK8 drive gear with the core of the extruder, particularly this information:
    *For 1.7MM diameter filament, this equates to a distance of 4.15 through 4.25MM from center of motor shaft to the center of the core.
    *For 3.0MM (2.85MM actual) diameter filament, this equates to a distance of 4.725 through 4.825MM from center of motor shaft to the center of the core.
    There is a theoretical potential core alignment difference of 0.575 MM (0.023 IN) when changing from 1.7MM filament to 3.0MM filament. To compensate for this difference, I could make two extruders, but I really don't want to be swapping out extruders when changing the size of the filament, however if I don't make two extruders, then that will introduce friction into the drive by having a filament offset during filament feed, which could lead to jambs. Of course there are several ways in which a person could approach this problem, but while considering this dilema, I came to the conclusion that I want at least one of the filaments to be as closely aligned as possible, and considering that the 3MM filament will have a much more difficult time making this offset, as compared to the 1.7MM filament, I am choosing to align the drive gear and core according 3MM filament usage.

    After computing this mathematically and verifying the results in a CAD drawing, by aligning the 1.7MM core with 3.0MM core, there is in fact an actual difference or offset of 0.575 MM (0.023 IN), as initially stated. In my current extruder design and as close as I can currently determine, this offset must occur within 13/16 of an inch. I don't suppose that is too horrible, expecting a 1.7MM diameter plastic filament to offset 0.023 IN within 0.8125 IN. If jambs do occur, due to increased friction, when using 1.7MM filament, I could always make another extruder for 1.7MM filament or I could simply settle into using only 3MM filament.

    Whatever the outcome, I do plan to make cores for both the 1.7MM and 3.0MM filament, and experiment a little. Of course I will have to post the results at a later date.

    EDIT: Now that I have basically decided to design a 3.00MM filament extruder and considering that I make most of my designs with imperial measurements, I must now convert the metric to imperial. Before converting, I must take into account the differences of filament deformation due to applied idler pressure. I do not want the maximum pressure, because that may stall the extruder motor, nor do I want the minimum pressure, because that may lead to filament slippage, so I will choose a happy medium by dividing the 0.1MM difference in half, and add that to the minimum distance of 4.725MM, which now totals 4.775MM.
    For my extruder, 4.775MM will be the total distance from the center of the motor shaft to the center of the core. Converting that metric measurement to imperial brings about a result of 3/16 of an inch (0.188" actual conversion rounded), which is simply perfect for my needs.
  • idbruceidbruce Posts: 6,197
    edited 2014-05-03 10:19
    Without going into too much detail at the moment, I figured I would post a front and side view of the extruder, in it's current state of design. Although not shown, two fans will be mounted on opposite ends of the square tubing, with one pushing air and the other pulling air. As represented in the side view, a small circuit board (1/16 X 2 X 2) will be mounted to the rear of the tubing with standoffs. Both fans, the core heater, and one or two thermistors will be wired to this circuit board. Additionally, there may also be a third fan mounted to the top of the tubing for blowing air into the core and cooling the filament, which if implemented, will also be connected to the circuit board. The last remaining problem is the design of the idler and tension system.

    It is worth noting that I lowered the drive gear closer to the core, which reduced my 13/16 clearance, so I may be stuck with using 3.0MM filament.

    Oh, and the motor outline is shown in the front view, but the motor will not actually be visible.

    Extruder.jpg
    474 x 417 - 16K
  • idbruceidbruce Posts: 6,197
    edited 2014-05-03 17:31
    To help provide a better understanding of the extruder design, I am now providing a new front view, with the motor outline removed. In this view, fans are attached to both ends of the tubing (represented by black rectangles), and the filament is shown in black, going through the top guide bar/mounting bar, passing through the drive gear (represented by gray circle), and into the core assembly. Please note that the idler and tension assembly is still not shown.

    Extruder2.jpg
    278 x 545 - 6K
  • $WMc%$WMc% Posts: 1,884
    edited 2014-05-03 22:28
    Bruce;
    I would add another feed roller (gear driven) like a MIG welder uses. One on each side.
    '
    This would deform the filament equally on both sides and lesson the filament to want to curl. It would also sink more heat going back to the spool.
  • idbruceidbruce Posts: 6,197
    edited 2014-05-04 01:33
    Walt
    I would add another feed roller (gear driven) like a MIG welder uses. One on each side. This would deform the filament equally on both sides and lesson the filament to want to curl.

    I agree with you 100%, but I just don't foresee it happening, mostly because of increased complexity and cost.

    To be perfectly honestly, while feed systems may look like simple devices, they are actually quite difficult to design and can be very complex, especially ones with dual rollers. When building the wire bender, I spent a lot of time investigating different wire feeders and various configurations. As you may have guessed, some were much easier to construct than others, with dual rollers being the most difficult, that is why the cheaper welders often use a single roller with an idler. Whereas with a welder, the rollers are much larger and bearings can be added or they can be motor driven or gear driven, but the increased size of the roller gives you much more room to design with. The MK8 drive gear is a tiny little thing, 9MM diameter X 11MM in length, with an effective diameter of 7MM, and the small size of the gear truly hinders design possibilities. You can't come around from the back side, because you have a motor in the way (unless you really want to get complex), so the only true option is to add the other roller from the front, which also has it's complexities. In addition to encasing the wire or filament with another roller or idler, for guidance and drive purposes, a tensioning device must also be incorporated to prevent motor stalls or slippage. From a design standpoint, adding a simple tensioning device is no easy task by itself, because the tensioning device should be spring loaded, to allow for deformations which may cause motor stalls.

    As it stands now, I got the easy part done, and I like the design so far, but the hard part remains. To add an effective roller with a proper tensioning device, I will truly have to think hard and be creative.
  • idbruceidbruce Posts: 6,197
    edited 2014-05-04 06:16
    Please allow me to introduce the first complete rough draft of the "Drummond Extruder".

    In combination with the information and images provided in posts #224 and 225, I believe this additional image should complete the picture for you. However I will give you some additional notes to paint a much clearer picture.
    1. The idler and tension lever will be constructed from aluminum angle.
    2. The idler bearing runs on a pin which is pressed into the idler and tension lever, and the idler bearing is secured to this pin with a snap ring.
    3. The idler and tension lever has a pressed in sleeve bearing which runs on a pin that is press into the motor mount bar. The idler and tension lever is secured to this pin with a snap ring.
    4. The top guide bar/mounting bar has a slot to allow rotation of the tension screw. The tension screw has an oversized hole for a drift pin which is pressed into the top guide bar/mounting bar.
    Extruder Rough Draft.jpg
    794 x 735 - 44K
  • kwinnkwinn Posts: 8,697
    edited 2014-05-04 10:53
    I wonder if anyone has considered a filament feed assembly like the attached diagram. A bit more complex than the current method but it applies more force more evenly and does not deform the round profile of the filament. If the filament roll is mounted in the orientation shown it would also help to remove the curvature of the filament.
    489 x 551 - 31K
  • idbruceidbruce Posts: 6,197
    edited 2014-05-04 15:42
    kwinn

    I must be having one of those days, because I cannot grasp the theory behind your drawing , so could you please explain it to me.
  • $WMc%$WMc% Posts: 1,884
    edited 2014-05-04 19:32
    Sorry Bruce.
    '
    What I meant was a idaler roller oppistit of the feed roller.
    Extuder_rollers_WMc.jpg
    1024 x 558 - 18K
  • idbruceidbruce Posts: 6,197
    edited 2014-05-04 20:28
    Walt

    No biggy. In fact, after I finished my reply to you, I dove right in to finish the design, and stuck with it, until it was done. I hope the drawing, in combination with the others, is easy to understand, because just a small sliver of the idler is showing. In the drawings, it appears to be much larger than it actually is, however in it's current state, it is 5-7/16 inches from the tip of the nozzle to the top of the mounting bar. When I am finished with it, It will be exactly 5 inches tall. Kind of taller than I prefer, but it should be rock solid, have plenty of cooling capacity, easy electrical hook-up, and the ability to add more cooling and an additional thermistor if necessary.

    I hope you like it.
  • idbruceidbruce Posts: 6,197
    edited 2014-05-05 06:38
    In order to provide a little more clarity, I am now providing a more descriptive front view of the extruder.

    I spent some time on the design last night, in an attempt to reduce the overall height, and I managed to get it down to 5-1/8 inches, however that is about all I can subtract.

    Some of my main goals while designing this extruder were as follows:
    1. Rigidity - This extruder is constructed mostly from extruded aluminum.
    2. Ease of assembly and disassembly. Experimental cores and hot ends can easily be swapped out.
    3. Ease of wiring - With a circuit board mounted to the rear of the extuder, individual connections may be made to the board, and a muli-conductor cable coming to the board should be able to service all connected items.
    4. Ease of clearing jambs - Simply disconnect the hot thermistor from the circuit board and unscrew the core and nozzle as an assembly from the PEEK insulator.
    5. Feed torque and grip on the filament - The selection of the MK8 gear should take care of this.
    6. Cooling of the cold end of the hot end - The wind tunnel created from the tubing and two fans should provide efficient cooling immediately above the heating area.
    7. Ease of loading filament - There should be ample room available to easily load and tension the filament.
    8. Provide a means to monitor the cooling above the heating area - A mounting hole for a cold thermistor will be provided in the PEEK insulator, which is located in the center of the wind tunnel.
    Disadvantages:
    1. Height - While height was never a major concern for me, because I am designing a 3D printer to accomadate the height of the extruder, I did try to keep and get the height of the extruder down to a bare minimum.
    2. Weight - Weight is a major concern for me, because there is already a heavy burden on the X axis. Considering that the extruder is mostly constructed from aluminum, instead of plastic, weight might be a hinderance in the speed of the printer, but at least it will be rigid and it won't melt :)
    3. Length of the core - The core is 1/2 an inch longer than I prefer and this may lead to jambs. However I will keep my fingers crossed.
    EDIT: It is also worth noting that the intended fans are 50MM X 50MM X 15MM. These fans will be secured to the tubing with (4) 4-40 X 3-1/2" cap screws, nuts, and washers, going from one fan to the other, and should fit snugly within the four corners of the tubing. Additionally, although not shown, support for the PEEK insulator must be provided on the inside of the tubing to prevent insulator rotation, during core and nozzle removal. This support should be located in a position that will not interfere with the (4) fan mounting screws.

    Descriptive Extruder Front View.jpg
  • $WMc%$WMc% Posts: 1,884
    edited 2014-05-06 18:36
    @Bruce
    '
    I'm working on an extruder my self.
    '
    The feed rollers are very similar to yours or that of a MIG or SUB ARC welder. Quick swing out roller to clear a jam.
    '
    I'm also working on a two piece heater. A clam-shell design to clear jams from the heater faster with out having to get a drill bit out for a one piece heater packed full of PLA/ABS/etc.
    '
    I think a steel heater body clam-shell with inductive heating would work great. The inductive coil would be totally none contact with the steel heater orifice/body/clam-shell and could easy be lifted out of the way to open the clam-shell(steel) heater.
    '
    I need to test the Dia. and length of the steel body. I'm gonna start with 5/8 x 1 inch.
    '
    The easiest Temp. reader seems to be a 2wire RTD or thermistor. I like T/C's but I can't see giving up all that RAM space for the look-up chart.
    '
    I'll push some drawings later this week.
  • idbruceidbruce Posts: 6,197
    edited 2014-05-06 19:09
    Walt
    I'm working on an extruder my self.
    WOW That's great. I am glad your sticking around and now you are making an extruder. Very Cool :)
    The feed rollers are very similar to yours or that of a MIG or SUB ARC welder.
    Are you making your own drive gear or are you going to purchase one?
    The easiest Temp. reader seems to be a 2wire RTD or thermistor. I like T/C's but I can't see giving up all that RAM space for the look-up chart.
    I don't know if you have been following the other thread (http://forums.parallax.com/showthread.php/155404-Input-Needed-Combining-Propeller-Proto-Board-Prop-DIP-40-and-ADC-for-3D-Printer), but in their I talk about my purchase of the Propeller Memory Card (http://www.parallax.com/product/40004). I believe it has a pretty big bang for the buck, but I am not positive yet. Either way, I wanted one on hand, so I wouldn't have to wait for it later.

    Had a major setback last night... The sander took another dump.... Looks like I might have to spring for a new one, just to eliminate the headache.

    Anyhow, I look forward to seeing your drawings. Good luck with it.
  • idbruceidbruce Posts: 6,197
    edited 2014-05-06 19:15
    Hey Walt

    One other thing.... Considering my extruder design and the fact that the core and nozzle can easily be removed...... I am going to make an airline adapter for the core..... If I get a jamb, I am hoping that I can simply heat up the core and nozzle without any cooling, to melt the jamb, and I am hoping I can clear it with low air pressure.

    What do you think?

    EDIT: And after clearing the jamb with air, perhaps chase the core with a reamer.
  • idbruceidbruce Posts: 6,197
    edited 2014-05-07 06:37
    Referring to the extruder design of Post #233, I am now considering a design alteration.

    Black PEEK Rod, 5/8" Diameter X 12" Length, Cost = $31.51

    That is a pretty huge investment for the 1-7/8" piece that I actually need. So I am now considering the fabrication of a 1" diameter aluminum heatsink, complete with heat disapation fins and thermistor mounting hole. Considering this will be placed directly in the wind tunnel of the extruder, I think it would probably work better and be less expensive than the PEEK insulator.

    Cold Finish Aluminum Round 6061 T651, 1" Diameter X 12" Length, Cost = $8.36

    As a side note, considering the conductivity of aluminum and brass, the inside areas of the resistance wire bobbin will have a thin layer of Kaptan tape to prevent short circuits.
  • idbruceidbruce Posts: 6,197
    edited 2014-05-08 21:42
    A side view of the 1/8 X 2 X 2 tubing wind tunnel is now presented, with the proposed heatsink of the previous post illustrated. Although not shown, the cold thermistor mounting hole mentioned in previous posts would also be included as part of the heatsink. The two heatsink mounting holes allow for attachment of the heatsink to the tubing, and prevent rotation during core and nozzle removal.

    If a jam occurs, it is envisioned, provided an air compressor is available, that the core and nozzle be loosened with a wrench, and then rapidly removed with an air ratchet or similar tool. At which point, the core and noozle are attached to a compressed air fitting and valve, the entire core is then heated until it reaches filament melting temperature, and then the plastic jamb is expelled with compressed air. Upon removal of the jamb, the core and nozzle are then reamed to remove any remaining plastic.

    Heatsink.jpg
    470 x 444 - 23K
  • idbruceidbruce Posts: 6,197
    edited 2014-05-09 01:18
    I am now leaning towards a hot end assembly that includes a heatsink.

    Basically the heatsink, core, and resistance wire bobbin would all be pressed together as an assembly, with only the nozzle being capable of being swapped out. This would eliminate threading the entire core. At which point, the entire assembly could be removed by removing the (2) heatsink mounting screws, or if a person simply just wanted to change a nozzle, simply unscrew it from the assembly.

    The new drawing below, shows the proposed changes.

    EDIT: It is worth noting that the heatsink mounting holes would actually be rotated 90 degrees from the position shown. The position of the holes shown was effort to provide clarity and avoid crowding of lines.

    Heatsink2.jpg
    470 x 442 - 24K
  • idbruceidbruce Posts: 6,197
    edited 2014-05-10 08:19
    A new descriptive front view is now provided to show all changes, including the heatsink.

    Please note that the core now extends above the 1/8 X 2 X 2 wind tunnel, and the core is now much closer to the drive gear and idler. This will slightly increase the amount of friction in the drive system, but I am hoping it will make it easier to load the filament. Additionally, the visible portion of the core will be threaded to allow attachment to a compressed air line for jamb removal, as discussed in previous posts.

    Descriptive Extruder Front View2.jpg
  • kwinnkwinn Posts: 8,697
    edited 2014-05-10 10:01
    idbruce wrote: »
    kwinn

    I must be having one of those days, because I cannot grasp the theory behind your drawing , so could you please explain it to me.

    Sorry for the long delay in responding. A bit busy these days.

    It is a large roller with a rubber surface. The rubber has a small V groove that the filament rides in, and an idler belt that wraps around the top half of the large roller. The top 2 small rollers are fixed, and the bottom pair are spring loaded to press the belt against the large roller and filament. It provides a good grip on the filament without unduly deforming it.
Sign In or Register to comment.