3D Printer Build

Discussion in '3-D Printers' started by bmac2, Feb 3, 2019.

Help Support HMEM by donating:

  1. Mar 4, 2019 #41

    bmac2

    bmac2

    bmac2

    Well-Known Member Project of the Month Winner

    Joined:
    Aug 16, 2013
    Messages:
    995
    Likes Received:
    692
    Hi Don

    Yep that’s one of the things on my “Other Stuff To Think About” list. I was given some LED light strips and there ok but I just can’t think of a way to mount them where they would light up the build plate without just being in the way. They would probably work well for cabinet lighting but not for here.
    LEDIMG_3341.JPG

    LEDIMG_3340.JPG

    My wife had some of those solar Ikea garden lights where the charging circuit had packed it in and like I’ve said before very little goes into the bin without going through the workshop first. On these I’d salvaged the solar panel and the LED's. Those LED’s are very bright to the point of almost hard to look at. There 5mm and six of them only draw 0.04 amps at 12v with a 320 ohm resistor. I’m like yourself trying to figure out a way to get them mounted close to the head without being in the way. Looking around Thingiverse for cooling fans and other ideas it looks like the head can get pretty crowded.

    Fiber Optic light tubes mounted above the head?

    LED_IMG_3339.JPG

    LED IMG_3338.JPG
     
  2. Mar 5, 2019 #42

    ddmckee54

    ddmckee54

    ddmckee54

    Well-Known Member HMEM Supporter

    Joined:
    Jul 8, 2009
    Messages:
    192
    Likes Received:
    22
    Cable management and lights, these pictures show how I dealt with those issues on my printer.

    The first picture is what my printer looked like before I did anything more than using cable ties to bundle the wires. It was at best an organized rat's nest.
    upload_2019-3-5_11-30-32.png

    The next picture shows the under cabinet 12V LED that I added for some additional light on the build plate. This light also supports the cable chain to the print head. I've read complaints about cable chain being too noisy and that you shouldn't use it. I've found that is you take the time to ensure the chain is moving freely it is virtually silent. All I get is a slight click as the link sets down on the support, the stepper motors make more noise than the cable chain does.
    upload_2019-3-5_11-31-9.png

    The third picture shows what it looks like with the light on, and the light comes on whenever I power up the printer.
    upload_2019-3-5_11-31-40.png

    The last picture shows what things look like after I got done "Killing the rats" and eliminating their nest. All of the blue parts and the white parts have been printed by this printer to improve itself. Some of the stuff, the cable chain in particular, I found on Thingiverse and modified to suit. The rest of the stuff was my design and a way of learning how to use my 3D CAD package.
    upload_2019-3-5_11-32-49.png


    Don
     
  3. Mar 6, 2019 #43

    bmac2

    bmac2

    bmac2

    Well-Known Member Project of the Month Winner

    Joined:
    Aug 16, 2013
    Messages:
    995
    Likes Received:
    692
    Don thanks for the pics. You have a very nice tight setup. Your LED strips look a lot brighter than the ones I have and I like the cable chain. It seems like there’s little middle ground with cable chain either you love them or hate them.

    I’ve got a 3D CAD that I’ve been using for around 10 years and I dreaded the thought of having to learn another system. But a couple of weeks ago I dug out the manual and actually read the chapter on stl files. I was experimenting with exporting some of my existing files to Slic3r and at first couldn’t figure out what was going wrong until it dawned on me that Slic3r assumes that the file dimensions are in mm. Sort of explained why when I imported a part that was 2 inches long and a half inch wide it looked so tiny on the screen.

    When I started casting I had to rethink some of the ways I drew things up. Draft angle, wall thickness and could the pattern even be pulled from the sand. Now going through some tutorials on various slicer programs I have to start thinking in terms of “can it be printed?”, “will it need supports?”, “will the supports affect the print?” etc.

    Good thing this is a hobby . . . .
     
  4. Mar 6, 2019 #44

    ddmckee54

    ddmckee54

    ddmckee54

    Well-Known Member HMEM Supporter

    Joined:
    Jul 8, 2009
    Messages:
    192
    Likes Received:
    22
    Slic3r is good, and the Prusa version of Slic3r is even better. You sure can't beat the price and I always thought that Slic3r was better at recognizing where supports were needed than Cura was.

    Late last year I bit the bullet and bought a copy of Simplify3D. OMG what a difference.

    The biggest differences I noticed is in perimeter wall and the top layer quality. I've got a 0.4mm nozzle have always had trouble with Slic3r when printing objects with relatively thin walls, more than 0.4mm and less than 2.0mm thick. If Slic3r is not able to fit an even number of perimeters between objects, or in the perimeter wall, it leaves a gap. This isn't good for wall strength and looks like crap on the top layer. Simplify3D gives you the option of either filling this gap with the infill pattern, or extruding a single filament to fill the gap. I made the mistake of trying the infill option - once. You haven't seen a 3D printer in the spastic mode until you watch one trying to infill a gap less than 0.8mm wide on the top layer.

    The other big difference is printing with support. I used to avoid printing with supports like the plague, I ALWAYS used to have trouble removing supports. To get the supports where I thought they were needed, I had to put up with supports everywhere that Slic3r thought they were needed too. I used to add a single or a double perimeter sacrificial wall that I would later remove just to avoid printing using supports. Simplify3D allows you to really control the supports, adding or subtracting them where you need to. The supports are much easier to remove also. I've since been able to print complex cavities where I was able to remove all of the supports for that cavity in one piece.

    Not an ad for Simplify3D - just my observations,
    Don
     
    Preston Engebretson and bmac2 like this.
  5. Mar 7, 2019 #45

    ddmckee54

    ddmckee54

    ddmckee54

    Well-Known Member HMEM Supporter

    Joined:
    Jul 8, 2009
    Messages:
    192
    Likes Received:
    22
    Small parts, that's another nifty thing about Simplify3D. If you try printing small parts, and my rubber band gun design has several of them, Simplify3D asks you if you accidently tried feeding it imperial parts instead of metric parts. Then it offers to convert them for you.

    Don
     
  6. Mar 10, 2019 #46

    bmac2

    bmac2

    bmac2

    Well-Known Member Project of the Month Winner

    Joined:
    Aug 16, 2013
    Messages:
    995
    Likes Received:
    692
    I got one of those personalized ”recommendations for you” e-mails a while back from Amazon that had some PLA cheap ($9.00 for a 1 lb. spool). I knew I was going to need some and probably use most of it just getting this thing calibrated so I clicked on the link, read the reviews, and ordered it. When it arrived I thought I knew why it was on sale . . . . This stuff is PINK. And not just pink it’s so pink it almost glows.
    I’ve spent a lot of time reading on the internet the last few days and thankfully there is a ton of information on Marlin available. As usual my biggest problem was figuring out what things are called. I knew what I wanted to change but had no idea what they called it but I felt I’d gotten to where it was ready to print.
    The first attempt was a total failure with the nozzle too far from the bed so back to reading about bed levelling and setting nozzle height.
    IMG_3187.JPG

    The second attempt was more successful but the layers started shifting. The curious thing about this print was that it didn’t shift consistently. The “o” in Hello is good and the shifting got progressively worse doing to the left.
    IMG_3189.JPG

    After another night of reading I decided it was ready to print a 20x20 calibration cube. This is where I learned about infill. If you don’t have any infill there’s nothing to support the top layers. Idiot.
    IMG_3190.JPG

    It’s not pretty but as a tool I was happy with it. The Y axis was shifting badly and perhaps being printed too hot the X looked better. Time to read up on 20x20 Calibration Cube print defects.

    IMG_3191.JPG
    IMG_3192.JPG

    One thing that came up repeatable was that the belt tension wasn’t tight enough so I tightened the belts . . . . . . . . Ok. Apparently that wasn’t it. . . . . . Back to Google.

    IMG_3194.JPG
     
  7. Mar 11, 2019 #47

    XD351

    XD351

    XD351

    Well-Known Member HMEM Supporter

    Joined:
    Feb 17, 2015
    Messages:
    314
    Likes Received:
    58
    Looks like your x axis is jumping steps , could be the amount of current from the driver to the stepper or acceleration/ deceleration settings in marlin , what speed are you printing at ? What is your maximum travel speed set to ?
    What is the x& y jerk settings set at ? Are both axis nice and free to move with no binding ? Are the drive pulleys locked to the shaft ?
    Thomas sanladerer does a good tutorial on marlin settings on youtube .
     
  8. Mar 11, 2019 #48

    ddmckee54

    ddmckee54

    ddmckee54

    Well-Known Member HMEM Supporter

    Joined:
    Jul 8, 2009
    Messages:
    192
    Likes Received:
    22
    I would sometimes get same step shift as shown in several of your pictures, not that mess in last one picture though.

    I personally think the step shift is a Slic3r glitch. I could never find the "smoking gun", but I did find that I could re-slice the model, generate a fresh g-code file, and the problem would go away.

    That last picture looks like you definitely caused a problem with the Y-axis drive. When you took out the extra tension that you put in the belt did the problem go back to the single shift in the Y axis that you originally had?

    I have had models that have caused a physical shift in an axis like this. These were usually caused by trying to print too large of an unsupported overhang when printing multiple parts at once. The thin edge would curl up as it cooled and sometimes this lip would catch the print head as it traveled to print the next layer. If the curled edge was big enough and hard enough, it would cause the printer to lose steps. This is not the problem with your cube since you weren't printing any overhangs.

    If you haven't tried re-slicing the model yet, just for giggles and grins, give it a try.

    Don
     
  9. Mar 11, 2019 #49

    bmac2

    bmac2

    bmac2

    Well-Known Member Project of the Month Winner

    Joined:
    Aug 16, 2013
    Messages:
    995
    Likes Received:
    692
    Hi Don. Nope this wasn’t caused by a Slic3r glitch it was a definite Bob glitch.

    Bingo XT it was driver current. A lot of this thing is made up of bits and pieces I had around including the stepper motors. I’d pre-set the stepper drivers and in the process of building it and tearing it down I’d paid close attention that the two Z axis steppers where a match but mixed up a couple of the others. At some point the 1.5amp stepper on the Y axis was only been fed .75 amps. Tightening up the belt did not make it very happy. Went back through my stash and did a more selective job of picking the motors so all are now 1.5amp 1.8 degree. The rest of the machine checks out being square and true with no binding so back printing stuff.

    I have no idea what normal is for a 3D printer so I’m just using the default settings in Slic3r (perimeters 60mm/s, maximum travel speed 180). Marlin is yet to be tuned so that will be more trial and error but this thing didn’t come with a manual. Calibration is always a slow respective process of documenting changes and repeating the same test over and over (20mm cube) but I’m starting to make some real progress.

    The “Print Quality Troubleshooting Guide” at Simplify3d has proved invaluable getting to this point

    I’ve was messing around with layer height and temperature settings in Slic3r. I Changed layer height to 0.2mm and printing at 205c with everything else still at the defaults. To me it’s looking pretty good.

    Can you get filament with a decongestant? This things nozzle gets like having a three year old with a cold in the room.
    IMG_3197.JPG
    IMG_3198.JPG
    IMG_3202.JPG
    IMG_3201.JPG
    IMG_3203.JPG
    IMG_3201a.JPG
    IMG_3199.JPG
     
    minh-thanh likes this.
  10. Mar 11, 2019 #50

    XD351

    XD351

    XD351

    Well-Known Member HMEM Supporter

    Joined:
    Feb 17, 2015
    Messages:
    314
    Likes Received:
    58
    Looks good !
    The constant dribbling from the nozzle can be because the hot end temp is a little high , try 190c . I usually put my first layer down at 205 c then have my slicer settings back it off to 190 and the top layer is at 175 c . You will find different pla will like different temps but somewhere between 175 and 205 is a good starting point .
    My printer died last week - motherboard cooked itself , well the power section for the drivers and micro did anyhow . The micro seems ok ( removed some damaged components then powered it up via a remote supply and the lcd still shows the programme running but there was no power to the drivers . New motherboard ordered and i now have to make a circuit up to test the drivers as I don’t want to fry the new board if one of the drivers has a dead short ! Anyhow that is enough of me hijacking your post !
     
  11. Mar 12, 2019 #51

    ddmckee54

    ddmckee54

    ddmckee54

    Well-Known Member HMEM Supporter

    Joined:
    Jul 8, 2009
    Messages:
    192
    Likes Received:
    22
    For the initial prints on a new machine this is looking pretty good. A lot better than some of the prints you see on videos, some of those are just nasty.

    You'll have to play with the temperatures and print speeds to find the sweet spot for your machine. Every machine is different, you need to find what works best for yours. Even different colors of the same brand of filament will perform differently. I've had white filament from one manufacturer that wanted to print at 230° while black filament from the same outfit wanted to print at 210°.

    I've found that my machine is happier printing in the 45-50mm/s range for perimeters, just seems like I get better print quality at a slightly slower speed. It's a balancing act, you want a good looking print, but you don't want it to take forever to print the stinkin' thing.

    Like XD351 said if it's dribbling it's probably too hot. I have to run my temperatures in the upper end of the print range, I typically print PLA in the 215-230° range. At least that's what my printer says the temperature is, I haven't been able to figure out a way to verify the actual temperature with the hardware that I've got. My IR thermometer doesn't like shiny surfaces. Every time that I've torn the hot end apart I've been in a hurry to get the printer running again so I've never bothered to paint the hot block with BBQ paint to see if the flat black would give me a better reading.

    Extruding hot filament is good, because you get better adhesion between layers. But hot is also bad because you've got to get the heat out of the layer you've just printed so that you can print additional layers. If you haven't got a part/filament cooling fan that'll be one of the first things you want to add. I highly recommend a radial blower type fan as opposed to an axial fan. The blower will move a lot more air which is what you want. There are several cooling fan duct designs on Thingiverse, pick one you like and print it out. I printed mine out of PLA and a long as I keep it at least 1/4" away from the hot block I don't have any problem with it deforming. I picked up a trick from youtube for fine tuning the cooling fan duct design, use a shallow dish of water on the build plate. This allows you to lower the nozzle to the surface of the water and actually see the ripples where the cooling air is hitting the surface.

    Above all, remember this is SUPPOSED to be fun!

    Don
     
  12. Mar 12, 2019 #52

    Peter Randerson

    Peter Randerson

    Peter Randerson

    New Member

    Joined:
    Feb 3, 2019
    Messages:
    3
    Likes Received:
    0
    Gender:
    Male
    Location:
    Wellington New Zealand
    Dialing in an own design 3D printer can be a long and involved process given the number of variables. However, You Tube has some excellent tutorials and diagnostics available.
    It appears to me that you need to do a couple of initial steps:
    1. A PID tune (Google it) of your extruder temp control. Done in Marlin Firmware.
    2. Calibrate your Extruder stepper (Google it) so that when the Gcode asks for 10mm of filament it actually gets 10mm of filament.

    I would also recommend subscribing to the following You Tube Channels:
    Makers Muse
    3D Printing Nerd and
    Tom Salanderer
     
  13. Mar 12, 2019 #53

    XD351

    XD351

    XD351

    Well-Known Member HMEM Supporter

    Joined:
    Feb 17, 2015
    Messages:
    314
    Likes Received:
    58
    There is also the reprap calculator on the prusa website , this makes calculating the steps per mm or inch of travel easy .
     
  14. Mar 17, 2019 #54

    bmac2

    bmac2

    bmac2

    Well-Known Member Project of the Month Winner

    Joined:
    Aug 16, 2013
    Messages:
    995
    Likes Received:
    692
    XT, Don I think I have my post nozzle drip settled. Part of the problem was I’d been poking around with the “end” G-code and inadvertently deleted the M104 command to turn off the nozzle at the end of the print. Also I’d calibrated the extruder but I hadn’t checked the filament and my 1.75mm filament is more like 1.82mm so I was also over extruding. As far as “SUPPOSED to be fun” I love this sort of thing. I’m recently retired and I have a very understanding wife that doesn’t mind if I want to disappear into the shop for a couple of hours, or days, as long as I get my “chores” done.

    Peter I am defiantly a graduate of YouTube U. Most of this printer is ideas off YouTube and there was a lot of pink goo between the first cube and that last one.

    I’m trying to use up the pink stuff as fast as I can with a tool holder and a bird feeder. The bottoms warped on both but the chickadees don’t seem to mind. By time the feeder was done it was barley holding onto the bed. I guess I should get the bed heater installed.
    IMG_3220.JPG
    IMG_3217.JPG
    IMG_3219.JPG

    One thing. Is it normal to have to relevel the bed every half a dozen or so prints? I’ve been reading up on auto bed leveling and I’ve got a few ideas but I think it might be the way to go.
     
  15. Mar 17, 2019 #55

    XD351

    XD351

    XD351

    Well-Known Member HMEM Supporter

    Joined:
    Feb 17, 2015
    Messages:
    314
    Likes Received:
    58
    You might need to adjust your retract settings in your slicer , i would put a mark on the z axis feed screws ( i use a sharpie marker ) that way you can confirm that both of the z axis screws are homing in the same spot . My printer never seemed to home correctly after a print so I guess it was losing steps homing the z axis I don’t see why you should have to re level the bed although i check mine every print but rarely have to adjust it ,if you want auto bed levelling you will have to configure it in marlin and hook up a probe to either the servo port if your board has one or z minimum on your machine i have just fitted the new motherboard and loaded the latest edition of marlin and i had a hell of a time getting the z axis to home properly . It turned out that there was some noise causing it to false trigger which can be configured in the config.h settings . Still thinking about moving the bed heating mosfet off the board as it gets awfully hot ,i made up a simple driver tester. ( polulu or stepstick) so i can pre configure the current setting and test operation before plugging it into the motherboard and it made life much easier .
    Your prints look ok with no stringing so you can’t be far off with your settings .
     
  16. Mar 17, 2019 #56

    Peter Randerson

    Peter Randerson

    Peter Randerson

    New Member

    Joined:
    Feb 3, 2019
    Messages:
    3
    Likes Received:
    0
    Gender:
    Male
    Location:
    Wellington New Zealand
    Bob
    For me the heated bed is an essential element (pardon the pun) for keeping prints in place,
    You asked about bed leveling. My current printer is a cartesian (same as yours) with a 2020 ally frame. I am using a Duet wifi controller and a BL Touch Z probe. With the Duet software I can run a multi-point leveling sequence that provides a mesh that the software can use to adjust level. I run that sequence before every print. Even with that I find I am having to manually re-level the bed every couple of weeks.
    I have some theories on why the bed goes out of level but no answers on fixes:
    1. I am using 2 stepper motors for the Z axis. It takes very little to get these out of alignment. Even 0.1mm is enough to affect the print.
    2. The bed is mounted on 4 springs, and is in constant motion when printing. Ask yourself, is that the best design for maintaining an accurate level surface?
    3. I use a brass brush to clear ooze from around the nozzle. By design, the print head is far from rigid so rubbing a brush over it on a regular basis may knock things out of alignment.
    I have read that the bed should be mounted on 3 points rather than 4 to make it easier to attain and maintain level.
    My nest printer will be a Core XY with the bed only moving in the Z axis and using one stepper.
     
  17. Mar 30, 2019 #57

    bmac2

    bmac2

    bmac2

    Well-Known Member Project of the Month Winner

    Joined:
    Aug 16, 2013
    Messages:
    995
    Likes Received:
    692
    Just an update on the printer. I think I found the problem I was having with keeping the bed level. I noticed that the loops and brim on the prints where thin to the point of being transparent. Parts where sticking to the bed so hard that I had to use a lot of force to break them free consequently messing up the bed. After many hours of reading, fiddling around with Marlin and running test prints I finally have the nozzle height dialed in correctly. I must have missed that chapter in the setup manual . . . oh yeah no manual.

    I’ve added bed leveling, parts cooling fan, and lights. The bed leveling works pretty well for DIY. It’s loosely based on the BL Touch. In a lot of the advertising they say how the BL Touch they don’t use a servo they use a solenoid to lift and drop an aluminum rod with a magnet on the end that triggers a hall effect sensor.

    I remembered that when I bought my first Arduino Kit it came with one of those little 9G servos so mine has a servo lifting the probe. Drew up a simple mount and printed it out. The hall effect sensor is the same as I used for the limit switches I just made it about half the size so it would fit.

    IMG_3222.JPG

    IMG_3229.JPG

    At first I couldn’t get the servo arm to catch the probe reliably and after about an hour fiddling around trying to add a small lip to the arm it dawned on me that I was sitting in front of my laptop (with CAD) beside a working 3D printer (idiot). A half hour later I had it assembled and was working on setting the deployment and retraction angles.
    IMG_3224.JPG
    IMG_3225.JPG

    The probe is a 1/16” brass rod with a 1/8” neodymium magnet epoxied to the end. This one works pretty god but I’ve found one on Thingiverse the BFPTouch (https://www.thingiverse.com/thing:2890290) that uses an optical sensor and is a lot neater being enclosed that I might give a try. I scrapped the one in the picture and redid the mount so the probe was on the inside closer to the hot end.

    IMG_3231.JPG
    IMG_3242.JPG

    My parts cooling fan is definitely a version 1.0 but better than nothing for now. The lights are 4 LED’s mounted in the hot end cooling fan and put a lot of light right on the nozzle.

    As a test for all the tweaking and tuning I printed another small bird feeder. The wall's only 1.5mm thick and it’s about 170mm tall with no supports. It took about 5 hours to print but looks cool in the white PLA and the chickadees love it . . . . or maybe it’s just free food.

    IMG_3349.JPG
    IMG_3350.JPG

    IMG_3351.JPG

    IMG_3352.JPG
     
    gunna, Cogsy and minh-thanh like this.

Share This Page