Z feed rate in Netfabb

[berend 0]

Troubleshooting on the height deviation of my objects (Z axis problem) it looked at the generated code, in particular the Z moves.

The goal is to find out how a z move is defined in Netfabb gcode and what values are used

The approach was generating Gcode files, extract the lines where Z motion is defined and examen these.

The victim of my research was a cube of 1x1x1 mm. (no cubes were harmed in this research)

Result:

1. Z moves are classified as 'fast jumps' by Netfabb (Toolpath -> Visualisation -> Display fast jumps)

2. Moves (G1) with a new Z setpoint have allways the same X,Y and E as the previous step.

So: Z is always moving 'alone' and is not used for printing. (This is a difference with SF where Z is moving together with X and Y)

3. With 3 of Florians profiles (Hollow standardQ, Hollow HighQ and Hollow UltraQ) the feed rate for z was the same for all profiles: 1200

So: Z moves are profile independend (for the used profiles)

4. Another parameter: Machinetype -> ... -> General Settings --> Max Z acceleration (m/s2)

- can be adjusted from 0.01 to 10

-Effect of the 'Max acceleration' setting on the feed rate is displayed in the graph:

HipChat-3.0.13-139.zip

Conclusion: In Netfabb I can define my G1 Z feedrate from 200 to 1200 using Max Z acceleration

How this results in my printed object depends of ReplicatorG, Firmware and my mechanics.

 

Remarks:

1. Each Netfabb gcode line is filled with Z data despite its only relevant when Z motion is desired.

(if communication bandwith is a problem whil printing, these Z positioning data can be filtered away easily)

2. The Max Z acceleration parameter has no influence on the Z feedrate of the gcode header.

Have fun,

Berend

update: i just came into my mind that there is also an speed override function in Netfabb. That is my next focus.

[owen 19]

Some interesting findings there Berend. Could be a simple improvement to Bandwidth alright.

Owen

[floush 0]

Hi Berend,

great research topic!

Just a couple of thoughts:

- Finding the right Z speed/acceleration will improve the Z scar, in the best case let it disappear completely.

--> Choosing improper values could have an effect on the Z accurancy.

- Depending on your firmware the Z speed it completely overridden. E.g. Bernhard's Marlin uses a much too low Z speed per default.

- Changing the Z accerlation in Netfabb will break the print time calculation. This has to be fixed soon. The default value of 0.0150 (or was it 0.15?!) is way to low.

- At some point Paul and myself found that if you use the same Z speed as the perimeter is printed results in almost no Z scar. But at the moment this is overridden by bad firmware defaults.

That's it for now - back to work.

Cheers,

Flo

PS: To be political correct it's more like "Paul's profiles" than "Florian's profiles".

[joergen 2]

So, what value do I enter where in netfabb with which firmware to get rid of the z-scar (middle of nov 2011)?

[berend 0]

All,

I examined the influence of the Netfabb option 'Toolpath->Override->Speed' on the generated feedrate for Z.

Here my results:

1. The Z feedrate is affected by the Speed override.

2. The feedrate is override is calculated after Max. Z acceleration calculation.

vf-low-poly-bow-tie.zip

Next focus is on Z feedrate generated by SF.

Have fun,

Berend

[berend 0]

Hi Berend,

- Changing the Z accerlation in Netfabb will break the print time calculation. This has to be fixed soon. The default value of 0.0150 (or was it 0.15?!) is way to low.

 

I assume u mean 'Max Z acceleration' from the machine menu.

If this is not correct i'm curious where to find the 'Z acceleration' adjustment.

 

PS: To be political correct it's more like "Paul's profiles" than "Florian's profiles".

 

Paul, sry, next time it's 'Paul's profiles'

Flo, your such a modest guy

Greetz,

Berend

[catohagen 41]

Does the firmware support subprograms ? when the part are equal in same height (same xy code in several layers) using subprograms would create smaller gcode output

example

G1 Z0.0

M89 P02

G1 Z0.2

M89 P02

G1 Z0.4

M89 P02

O2

( SUB NUMBER, 02 )

G1X0Y0

...

...

M99 (return to mainprog)