Firmware settings, first prints, slicer settings...

Discussion in 'OpenBeam Kossel Reprap and OpenBeam Kossel Pro' started by Terence Tam, Dec 27, 2014.

  1. Rick Gordon

    Rick Gordon Member

    Wow, that's an idea. And here I've been using "bead tweezers" to clean off the extra bit manually before it starts printing.
    (BTW "bead tweezers" have been very handy when dealing with this stuff - they're normally sold for people building beaded jewelry - very long thin points and roughened on the inside surface to make things easier to pick up.)
  2. Rick Gordon

    Rick Gordon Member

    Had time to look at this more closely yesterday - and eventually realizeed that the extruder was "skipping back" - it would run a few steps forward and then roll back slightly.
    Kind of like skipping, but it would retract filament at the same time. it would appear that the stepper motor is not holding its position against the pressure of the filament.
    For a 100mm extrusion, it was extruding only 60-70mm! I've been extruding with MHP plain PLA filament at 195C.
    Running tests at incrementally higher temperatures, I needed to get to 220C to get smooth extrusion again.
    I then experimented with extrusion rate, and dialed that in to 193.1 in order to truly get 100mm of filament in.
    But - I'd really like to extrude at cooler temperatures - trying to solve some of the bridging problems - I'm thinking that I need to increase the stepper motor current to compensate.
    I'm curious though - do steppers typically behave in this way? Is it possible the filament's elasticity has changed in the past few weeks?
  3. KDog

    KDog Member

    When my extruder "skips" it is always the stepper slipping. The only way I've found to fix this is raise the hotend temperature and slow the print down. Neither are a great solution. I do have a larger diameter bowden tube on order as others have reported the tube could be binding the filament and increasing pressure on the stepper. Would a higher torque stepper help?
  4. Rick Gordon

    Rick Gordon Member

    Is yours skipping in the same way? I'm not that "skip" is the right term - because its not just spinning against the filament - it's like its pushing (rotating CCW) and then bounces back (rotates CW) bringing the filament with it , then extruding again. But I'm doing anything other than a simple extrusion.

    Also I have to say - poking around in other forums leads me to concerns that such "bounceback" is caused by too much current, causing the driver to overheat. I don't know though.
    I did check the temperature on the extruder motor and it wasn't really warm.
  5. RicerX1

    RicerX1 New Member

    My extruder motor was slipping also. 2mm id bowden tube fixed it for me. Slipping is what happens when the pressure of the filament in bowden tube from the extruder to the hot end exceeds the holding torque of the extruder. 2 ways to fix increase extruder torque or decrease fiament pressure.

    Don't try a geared stepper as the problem seems to lie in the bowden tube and maybe hotend resistance / friction. When I tried a geared stepper all it was doing was bitiing and grinding away at the filament.

    After the bowden tube I was able to print down to around 195 - 200 C on my ultimachine filament (actual 1.74 diameter). I just picked up a roll or MatterHacker Pro PLA (actual 1.71mm daimeter) and that will print down to 180 C

    I am very excited for the new hot end as I believe that it will fix some of the filament pressure issues.

    The attached is MatterHacker Pro Light Blue PLA at 5 degree increments every 10mm.

    Attached Files:

  6. Rick Gordon

    Rick Gordon Member

    Thanks. I have to admit that in previous experiments with the tube removed I didn't notice any specific friction - but I didn't measure the ID either. The one that shipped is not 2mm ID?
    that's another clue at least.
    there seem to be multiple ways to reduce this back pressure:
    - reduce friction in the Bowden tube
    - increase hot-end temperature (allowing more material to flow at given speed)
    - slower print speed (thus calling for less material to be extruded)
    - maybe (i'm not sure) adjust the current applied to the stepper motor
    - replace stepper motor with one having a higher torque

    Also I know that we can change the stepper motor "rate" in mm/sec using M92 - that's how I was able to get it to accurately extrude 100mm measured.
    I don't know how that stepper motor rate translates to actual extrusion. I guess I could time it - or figure out how to calculate it.
    What I'm getting at - is there a maximum extrusion rate calculable for a given plastic and temperature - from which you can calculate the maximum print speed?

    for a 0.35 mm head, and a 0.15mm layer height, at 30 mm/sec - the required plastic volume rate would be:

    Extruder-Width * Layer-Height * Print-Speed = Extrusion-Volume/sec
    0.35 mm * 0.15 mm * 30 mm/sec = 1.575 cubic mm/sec

    You could divide that by the nominal filament thickness (say 1.72 mm) to get the required rate: 1.575 / 1.72 =~ 0.92 mm/sec

    That seems pretty reachable. What's the max?
    Can the Maximum Extrusion Rate (Max ER) can we then calculate the theoretical max print speed?

    Max Print-Speed = Max Extrusion-Volume / (Extruder-Width * Layer-Height)

    the trick is figure out that max extrusion -volume; i'm not sure how to do that yet.
  7. RicerX1

    RicerX1 New Member

    Check the filament in the tube on bends after it has passed through the extruder. The teeth on the hobbed gear sometimes distorst the filament enough to make it a bit tight. This was the case for me on certain filament. Virgin filament off the roll was fine.
  8. Wacky

    Wacky Member

    In remembrance of Saturday morning, (3/14/15 9:26), I should remind you that, in theory, the filament is a cylinder of diameter 1.72mm.

    The volume is given by the formula V = pi*D^2/4*H. So the calculation is 1.575/1.72/1.72*4/3.1416 = ~0.678 mm/sec
  9. Rick Gordon

    Rick Gordon Member

    Ha, fair enough, I was just doing a quick calculation. I'm surprised though that your number is lower - since a cylinder would have less cross-sectional area.

    On a related note, after some offline discussion I'm going to try increasing the stepper motor current a bit. I can just barely reach it with a screw-driver - enough to turn it 1/8 CW.
    I'll try that and see how things go.
  10. Wacky

    Wacky Member

    Comparing cross sections, you are correct. The area of a circle is only (very very roughly) 3/4 that of a square.
    However, you didn't compute the area of a square because you divided by the diameter only once.

    Dimensionally, you computed (mm^3/sec)/(mm) which is (mm^2/sec) and not the units that you quoted and desired (mm/sec)

    Thus, the cross sectional adjustment increase (4/pi) is more than offset by the diameter correction reduction 1/(2 mm)
  11. Rick Gordon

    Rick Gordon Member

    Thanks. I was actually trying to pay attention to the dimensional analysis but I overlooked that.
    It was just an idea - realizing that there are variables such as the max extruder speed, filament type, and hot end temp, as wellas unknowns within the slicer.

    anyway, i never got a chance to try it with the increased stepper motor current today -maybe tomorrow night.
  12. RicerX1

    RicerX1 New Member

    Pie R square . I believe area and volume calculations use radius.
    pi*R^2*H. Which will give you 3.14 * (1.72/2)^2 * 30 = 69.67 mm^2/sec

    Maximum extrusion rate will depend on the hotend resistance nozzle size and the specific brand, color, type of filament used. There are too many variables to accurately predict.

    I have never came across a hot end that can not extrude at least 50mm/s at 200 C. makergear, jhead, jhead clone, Prometheus, e3d v5, e3d v6, pico, and kossel pro jhead. All in .35 or .40 on a geared extruder setup. on the 2 finger test the pressure on the kossel pro jhead pressure is higher but I think it maybe tighter tolerances used in machining them. (yes i have an addiction to hotends)

    Max extrusion rate should not be an issue untill you start to go with very small nozzles.
    Last edited: Mar 16, 2015
  13. Wacky

    Wacky Member

    pi*R^2*H is not different from pi*D^2*H/4 because D=2*R.

    Which is nonsense. 30 mm/sec is the OUTPUT rate. And 1.72mm is the INPUT diameter.
    Remember that he was trying to compute the INPUT feed rate. That rate will depend on the nozzle diameter which appears nowhere in your computation.
  14. AndyG

    AndyG Member

    Have you attempted to integrate the pico into your KP? I was looking at the size of it the other day and wondering if it might be a drop in replacement with some hot end re-wiring.
  15. RicerX1

    RicerX1 New Member

    Unfortunately no. The barrel is offset from the mounting. And the heater wire comes out the middle.

    Attached Files:

  16. Terence Tam

    Terence Tam Mr. OpenBeam Staff Member Vendor

    Wierd. Are you saying that the ID of the bowden tube isn't 2mm? Has it been recut?

    That's the bowden tube that we use. Modifiedfluoroalkoxy (MFA) for slipperiness, 2mm ID.

    We should be getting the new heat breaks in 2 weeks, and we have prototype quantities (20-30) of the heater blocks and nozzles on their way to us. Unfortunately I've tried a bunch of places to get a prototype of the heat sink made - no dice, so I'll end up machining it myself to test then releasing the heat sink and trying to expedite it this coming Monday. Then we should be able to start shipping the new hot end (in another 2.5 weeks)

    -=- Terence
  17. RicerX1

    RicerX1 New Member

    The tube was not recut and I tried to stick the 2mm drill rod from the hot end into each side and it wouldn't fit, would not even slip in a little. I just assumed that it was 1.8mm ID since the drill rod wouldn't fit.

    So you got me thinking if it was maybe crushed by the bowden fitting. I used a sharp razor blade and cut off about an inch from one end and I was able to get the drill rod to slide in. It was tight fit but it went in with out too much force. So the supplied tube is 2mm ID its just much tighter tolerance then the 2mm ID PTFE tube I got from tridprinting.

    Strange but the ptfe tube off my Kossel Mini fixed my problems. Maybe I'll revisit the MFA tube and cut off the crushed ends when the new metal hotends are released.
  18. Rick Gordon

    Rick Gordon Member

    Going back to my original problem - I had a chance to do some testing, and will do some more.
    First I experimented with raising the temperature. This required going up to 220C to get a smooth flow. Now - I know the hot-end's temp is lower than what is reported from the thermistor, but its not that much lower! But at least I got somewhere.
    Next - I increased the stepper motor current by "1/8 turn" - not sure how much I increased it. I then tried printing something at 215C and aside from problems bridging it worked well enough.

    Next step is, I think, seeing how far I can lower the temp while getting a smooth extrusion at this current, and then deciding whether to increase the current further.
  19. RicerX1

    RicerX1 New Member

    Be careful with turning up the current. I looked up the data sheet for the supplied steppers and the max amp is 1.65 i think. The DRV8825 stepper drivers are only good for 1.5a with out a heatsink for fan cooling. There may be no immediate issues with running a higher current but I worry about the long term effect.

    You would need to know the sense resistor value to safely turn up the current. The vref can be measured between the pot and ground so you know where your at, so your not just blindly turning the pot. The pololu drv8825 board use a 0.10 sense resistor which makes Vref/0.50 = I, but I think the brainwave pro uses a different resistor. The vref is 1.5v at default settings when you turn it 1/4 turn it puts it at 2.5v (66% increase) and 1/8 turn will put you at around 2.0v (33% increase).

    With all that in mind I turned my pots back to default settings and explore other solutions. I was not in the mood to remove the brainwave and probe the sense resistors. Maybe @Terence Tam can chime in about the sense resistor value and turning up the stepper current. Would hate to see people destroy their boards / printers but I do not have the knowledge to be certain.

    The formula on the datasheet says V(xref) / (5 x Risense) = Ichop
  20. Rick Gordon

    Rick Gordon Member

    thanks. well, that's a bummer. I thought 1/8 turn would be "safe". so far, i've only printed with it at that setting for approx 1 hr, and i was monitoring the stepper motor for temp increase - it stayed approx room temp the entire time. so its not overheating the motor at least.

    so what other solutions did you come up with? just leave the temp high?

Share This Page