INTEGRATING WIRE FEEDER TO THE DUET CONTROLLED WELDER
-
Hi Owen,
Good to hear from you
Separate Wire feeder is not used for the designed setup.
I have built an custom wire feeder using a stepper motor and configured it as an U axis, hence using multiple motion system
-
@SANJR said in INTEGRATING WIRE FEEDER TO THE DUET CONTROLLED WELDER:
I have built an custom wire feeder using a stepper motor and configured it as an U axis, hence using multiple motion system
Just configure it as an extruder like @dc42 suggested, rather than as a separate U axis. A normal gcode move such as G1 Xn Yn Enn Fn means move a distance of "n" in the X and Y directions at a feedrate (speed) of "Fn" while extruding (or feeding wire) a distance of Enn. That effectively gives you independent feedrate for the wire feeder because the move will be synchronised. That is to say the wire feed (the E move) will start and end at the same time as the XY move and because the "E" distance will be different to the XY distance, so the feedrate for the wire feeder must vary proportionally.
-
Thanks
For a better understanding, lets consider for example
G1 X100 E2000 F50
As per the above the code the X axis moves 100 mm at a rate of 50 mm/min, the extruder feeds the wire for 2000 mm but at a higher feed rate since the 2000 has to be completed at the same time of X100.
Is my understanding correct?
-
As has been stated.
If you're going to try to work this way then the wire feeder must be considered an extruder and configured exactly that way in both RRF and your slicer.
I.e set your extruder steps per mm the same way.
In your slicer, set your filament diameter to your wire diameter (say 1.2mm)
Set you nozzle width to approximately the bead width your trying to achieve (say 3mm)
Set your layer height to the bead height (say 1.5mm)Then the slicer will calculate the correct E distance for the volume required.
There are MANY problems with this methodology.
Firstly you will need to insert your macro to start and stop the welder before and after every move that contains an E value.
Secondly and more importantly, the resulting layer height will not be uniform or exact so cumulative errors in height will result.
An FDM printer assures the height by wiping the top layer level and achieving square edges.
You will have a rounded profile.You may be able to periodically probe the last layer and readjust the Z height but you'd have to make sure it didn't result in the next layer being too far below what you've adjusted it to.
Then you still have to overcome the requirement for a delay at the start of the weld and a crater fill at the end of the weld.
-
I also suggest you watch this video.
Apart from detailed explanations energy transfer you should consider especially the section on omnidirectional processes.
To achieve this you would need something similar to what is used on Cartesian systems used for concrete applications.
https://youtu.be/HGII_91SWeA?si=9-rUoEhlGSPu5n37 -
Thanks Owen
The Video explanation was really helpful in understanding few areas where I was sceptical. So once again thanks very much.
Now as rightly said in the video and explained in this post, In order to have a either a droplet transfer or bridge transfer at any cost the wire plunging to be avoided. In addition, unlike an extruder the wire is feed at an angle to the torch. Finally as Owen rightly pointed out the issues of omnidirectionality is a key issue.
So in order to control these issues one of the way is to play with the WIRE FEED RATE.
Based on the discussions in this post summarized the possible option each with their limitations
- Configure the WIRE FEEDER as an extruder and use Open Source Slicer to predict the profile/geometry of the bead.
Limitation : for the below code
G1 X100 E2000 F80
for this code the X axis moves 100 mm at a feedrate of 80mm/min. The WIRE FEEDER extrudes wire at a rate of 80 mm/min for the 2000 mm that is the extruder would be feeding the wire even after the X axis would have stopped. This means that the motion of X axis and the WIRE FEEDER would be coordinated and synchronized at the beginning but not at the end. Is my understanding of this option correct?
-
Configure as an U axis by multiple motion system
Limitation : A) daemon.g file needs to rename as .PART to update the F parameter and the machine needs to switched
OFF and ON for a certain time like 15-20 mins for the feed rate to update.
B) Not able to vary the WIRE FEEDER FEED RATE in a continuous tool path.
C) Not able to expand for 2nd WIRE FEEDER, only 2 multiple motion systems are possible -
Install separate WIRE FEEDER systems and configure as IO through a 5V relay to control the ON/OFF operation through DUET. And the speed can be adjusted by WIRE FEEDER Potentiometer. By this option additional WIRE FEEDERs could be added if required.
Limitation : -
Coordinated and Synchronized movement at the beginning and end are required. In such a scenario option 1 doesn´t suit the purpose. Option 2 has limited flexibility
Any corrections or suggestion would be most welcome plz........
- Configure the WIRE FEEDER as an extruder and use Open Source Slicer to predict the profile/geometry of the bead.
-
@SANJR said in INTEGRATING WIRE FEEDER TO THE DUET CONTROLLED WELDER:
Thanks
For a better understanding, lets consider for example
G1 X100 E2000 F50
As per the above the code the X axis moves 100 mm at a rate of 50 mm/min, the extruder feeds the wire for 2000 mm but at a higher feed rate since the 2000 has to be completed at the same time of X100.
Is my understanding correct?
Yes, that correct. You can actually work out the what the feedrate for the extruder must be (in simple terms ignoring accelerations and decelerations). If the head moves 100mm @ 50mm/minute then the time for the move to complete is 2 minutes. If the extruder has to move 2000 mm then it must move at 1000mm/minute (2000mm divided by 2 minutes).
-
-
@SANJR said in INTEGRATING WIRE FEEDER TO THE DUET CONTROLLED WELDER:
- Configure the WIRE FEEDER as an extruder and use Open Source Slicer to predict the profile/geometry of the bead.
Limitation : for the below code
G1 X100 E2000 F80
for this code the X axis moves 100 mm at a feedrate of 80mm/min. The WIRE FEEDER extrudes wire at a rate of 80 mm/min for the 2000 mm that is the extruder would be feeding the wire even after the X axis would have stopped. This means that the motion of X axis and the WIRE FEEDER would be coordinated and synchronized at the beginning but not at the end. Is my understanding of this option correct?
Your understanding is incorrect.
It is as @deckingman has already stated.
If configured as an extruder, the feed rate of the extruder is not the same as the feed rate of the other axes.
In you example the feed rate of the X axis is 80mm/min and the distance 100mm so the move will take 1.25 minutes (ignoring acceleration)
The extruder deed rate will be whatever it takes to extrude the nominated amount in the same time it takes the X axis to travel 100mm.
2000/1.25= 1600 mm/min (26.66mm/sec)The only time the extruder feed rate equals the F value is if only the extruder is commanded to move.
G1 E100 F120 - Configure the WIRE FEEDER as an extruder and use Open Source Slicer to predict the profile/geometry of the bead.
-
@SANJR said in INTEGRATING WIRE FEEDER TO THE DUET CONTROLLED WELDER:
Configure the WIRE FEEDER as an extruder and use Open Source Slicer to predict the profile/geometry of the bead.
Limitation : for the below code
G1 X100 E2000 F80
for this code the X axis moves 100 mm at a feedrate of 80mm/min. The WIRE FEEDER extrudes wire at a rate of 80 mm/min for the 2000 mm that is the extruder would be feeding the wire even after the X axis would have stopped. This means that the motion of X axis and the WIRE FEEDER would be coordinated and synchronized at the beginning but not at the end. Is my understanding of this option correct?Not correct. When a move has both axis motion and extrusion, the feed rate refers to the overall axis motion. The extruder feed rate will be adjusted automatically so that the extrusion starts and stops at the same time as the axis motion, with the specified amount having been extruded..