New build - Extruder issues ( solved - with equation for Estep)

Hi all,

I have just completed my first build of a mendel but I seam to have a few questions regarding the extruder side of things.

Electronics : Gen6
Mendel: Standard
Filament: 2.7dia ABS
Extruder: Own design, 1:5.8 main gear ratio, 17mm dia pinch wheel, nema17 stepper 1.8º
Bed: non heated 5mm Al 1050 plate.





all works fine regarding the coldend and from what I can tell the hotend. The extruded filament is smooth and free follwing with almost no ooze. It can be run then left and restarted without issue.

my questions may seam pretty simple but to a novice such as my self they may not be so apparent. My problem is getting the extruded filament to stick to the bed.

1. My current temp setting is 230ºC - is this to low?
2. Do it need to cover the bed with Kapton tape or should I be able to print to the bare metal?
3. Is there a reliable equation to calc the steps of the extruder motor?

I would be very greatful for any wisdom in these matters.

regards

J90

Edited 1 time(s). Last edit at 09/10/2011 07:37PM by Joe90.

1 - 230 - 240 should do (double check the thermistor table referenced in firmware corresponds to the thermistor you have)

2 - Don't expect ABS to adhere to bare metal. Kapton and a table at 110C, acrylic for small pieces at rooms temperature... clean surface with aceton before each print

3 - if using skeinforge (or even if not) have a look at this [josefprusa.cz]

even easier if you have the hot end disconnected from the pinch wheel/motor block

Quote
Joe90
3. Is there a reliable equation to calc the steps of the extruder motor?

For Skeinforge 40 and later, steps per mm = (steps per revolution) x (gear ratio) / Pi / (pinchwheel diameter).

brnrd Wrote:
-------------------------------------------------------
> > 3. Is there a reliable equation to calc the
> steps of the extruder motor?
>
>
> For Skeinforge 40 and later, steps per mm = (steps
> per revolution) x (gear ratio) / Pi / (pinchwheel
> diameter).

I have found it is more accurate to determine this parameter experimentally.

There are some unknowns that influence the "pinchwheel diameter", like the depth of the pinch wheel teeth sinking in the filament, depend on the characteristics of the pinchwheel, the hardness of the filament, the pressure applied etc.

Yes but it is a good starting point. I then extrude 100mm and measure the actual feed length and get that right.

---

[www.hydraraptor.blogspot.com]

I enter the calculated value in the firmware and then I use the measured feed length to calculate the Filament Packing Density ratio setting in Skeinforge 40 and 41 as the ratio of the measured over the entered length.

brnrd Wrote:
-------------------------------------------------------
> I enter the calculated value in the firmware and
> then I use the measured feed length to calculate
> the Filament Packing Density ratio setting in
> Skeinforge 40 and 41 as the ratio of the measured
> over the entered length.

That is a very good point!

Thanks guys,

I have increased my temp to 240ºc which has helped. Also tried the Kapton tape with no luck, however by placing some double sided sticky tape on top of it ( B&Q cheap stuff ) it now works a treat. Still getting a bit of warping but it's good enough to work with for the moment. A heated bed is the way to go I guess for zero warp or close to it?

I use Resnapper btw not Skienforge but will try it at a later date once I get to grips with it a bit better.

The prints are coming out well ( very tidy indeed), however each layer is not bonding 100% with the prev one and with a good flex they start to delaminate? I think the solution is to increase my E_step value slightly more as the extruded filament is pretty round ish at 0.4mm x 0.5mm. Currently at 9.0 on E_step.

The formula's don't seam to come close unless I'm being thick? Also I noticed no coefficient of expansion in any of the equations?

Driven gear: 100 teeth
Driver gear: 17 teeth
Pinch wheel outer dia: 20mm
Pinch wheel inner dia: 17.2mm
Filament: 2.7mm
Extruder nozzle: 0.5mm
Temp: 240ºC
Ambient Temp 23ºC


cheers

J90

For 1/8 steps, your e steps per mm should be around 174 if you use the inner diameter of 17.2 mm.

Thanks for the effort brnrd,

but my settings for E-Step are around 8.6 or were untill recently ( now 8.5). This setting produces very tidy parts using Gen6 and Resnapper, electronics and application. Also I'm well within 0.1mm on size ( all 3 axis).

Trying most of the wiki equations and the odd ones found on the net with no luck, I decided to give the equation ago myself. The most noticable thing in all of them was expasion was not taken into account. Hence I decided to include it and the result of my efforts after many hours was supprising to say the least! My setup once calculated gives me a figure of 8.5 which I tried this afternoon. To my joy the prints are even better. Also changing to a new layer height is now a breeze, and works perfectly rather than the trial and error method before. My head hurts but it was worth it.

The setup for my extruder is the same as the wade system, a small gear on stepper to larger gear on shaft, with a pinch wheel located on the large gear shaft. The nozzle of the extruder is 0.5mm but that is not required. The gears are from HPC gears and of the helical type with a mod of 0.5 ( almost no backlash).

M = Steps per full rotation of motor
W = Width of extrusion
H = Layer height
R = Filament radius
P = Pinchwheel Dia
r = Half layer height (H/2)
∆T = Change in temp of filament (melt tempºC - ambientºC)
G = Gear ratio multiplier ( Driven/Driver)
E = Coefficient of linear expansion ( in my case ABS )


single line version: Step=M((W-H^2 )+(πr^2 ) )/(πR^2 (πP/G)+3E(πR^2 (πP/G) )∆T)*2

If anyone else has a similar setup please give it a go and let me know if you have the same results? I would be really interested in seeing if this works for others as it has for me?

regards

J

I thought you were using Skeinforge 40 or 41. These later versions makes the math of figuring out the extruder steps per mm a lot easier since it's based on the feed and not the extruded filament. The SW does all the math for you once you set this and the diameter of the feed correctly.

If you re-factor the bottom line you get πR^2 (πP/G)(1+3E∆T), i.e. you are saying the thermal effect is (1+3E∆T).

Where does the 3 come from? Why is not simply E∆T?

---

[www.hydraraptor.blogspot.com]

No worries brnrd, I thought you might thinking it was Skienforge.


nophead - Thanks for re-factoring the bottom line I was having a little difficulty with it, my math's is a bit out of practice these days. The 3 is due to volumetric expansion, β=3α. If the cross section of the filament didn't change or was not important then I would of just used the linear figures. Since it dose I used volumetric. However I can't ignore the similarity of the result with my settings. If the layer height is changed and a new e_step is calculated the parts still have the same quality.

I'm just experimenting and far from an expert in such things.

Speaking of which - I decided to dip ( quick one's twice ) the finished parts in acetone to dissolve the odd string or overhang blob. The out come was rather nice with the part becoming smooth and very shiny, with well defined contours. Will post some pic's soon.

I have to say this is very addictive past time...

regards

J

Edited 2 time(s). Last edit at 09/13/2011 03:36AM by Joe90.