It's all about speed - printing faster than 50mm/s

Hi,

for quick "preview" prints I use X- and Y-axis speeds of 50mm/s for some of the settings in slic3r. The prints look good and I'm wondering what the speed limits for X- and Y-axis are (beside those #defined in the firmware). Any ideas?

Markus

---

XBee & electronics blog: [lookmanowire.blogspot.com]

At the moment, they are quite conservative. See this post: [forums.reprap.org]
dc42 went on to play around with M201 and M203, but set it back to standard, I think, after having problems. There is still the issue of running combined XYZ moves too fast, which we're working on (it means vectoring all moves, and limiting that move to the speed of the slowest axis, and certainly isn't simple...)

Ian
RepRapPro tech support

Quote
droftarts
dc42 went on to play around with M201 and M203, but set it back to standard, I think, after having problems.

I'm currently using this in config.g:

M201 X800 Y800 Z5 E250 ; set accelerations
M203 X3000 Y3000 Z180 E2700 ; increase max extruder speed

which increases the max extruder speed but sets the other speeds to the firmware values. Looks like I reduced the z-acceleration too, but I think this was because of a temporary problem I had, so I'll change it back.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thx Ian & dc42,

I'm not quite sure if I understood the G-Code RepRap comments:

"M201 - Set max printing acceleration
in units/s^2 for print moves (M201 X1000 Y1000) "

and

"M203 - Set maximum feedrate
that your machine can sustain (M203 X200 Y200 Z300 E10000) in mm/sec
Note: this should be in units/minute, just like the F code."

right. Sorry no native english maybe the question is dumb.
"Acceleration" in this context means: machine is speeding up with xx mm/s^2 according the setting for each axis. Does this mean if I have a X/Y-vector it is accelerating with their (X-, Y-xis) maxiumum/ set speed until reaching the "feedrate"?

And do I understand "feedrate" as the speed which is set when the acceeleration procedure is done (aka "feedrate" is established). For example: the printer is moving a vector, starting from idle mode first with acceleration speed and when "feedrate" is established continues with "feedrate" speed? And the problem with the z-axis moves < 0.24mm was that the combined X-, Y- and Z-axis (vector) move was misinterpreted?

Sorry have to get this right before tweaking the values and compiling firmware to test higher speeds.

Markus

---

XBee & electronics blog: [lookmanowire.blogspot.com]

I think your interpretation is correct, except that the movement starts not at zero speed but at the speed given by the appropriate entry in InstantsDvs, and it decelerates so as to end the move at this speed too. InstantDvs is not configurable via Gcodes, and is set up with these values:

#define INSTANT_DVS {15.0, 15.0, 0.2, 2.0} // (mm/sec)

Also there is some lookahead, so the speed at which a move starts and ends can be affected by the previous and following moves. It was this lookahead that was going wrong and messing up the handling of small Z-axis movements.

What is it that you want to change in the firmware, given that you can change the max speed and acceleration parameters by sending Gcodes?

Edited 1 time(s). Last edit at 01/13/2014 09:57AM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Hi dc42,

thanks for straighten this issue.
I want to play with the feedrate / acceleration, but when gcode overwrites those defined values there is no need to adjust the firmware. I had the wrong assumption the defined max feedrate in the firmware would be kind of hardcoded then.

Markus

---

XBee & electronics blog: [lookmanowire.blogspot.com]

If I set speeds greater than about 25mm/s (printing or non-printing), my Y axis skips occasionally.

Ok "already" at 60mm/sec (I did not alter acceleration speeds) and short moves there is audible rattling in the y-axis possible due to the mass inertia of the bed construction and type of drive. The print is noticeable "blurred" X- and Y-axis-wise in areas where short moves are carried out.

Looks like the bed construction might be an obstacle for higher speeds?

Markus

---

XBee & electronics blog: [lookmanowire.blogspot.com]

You may need to change the stepper driver current to achieve higher speeds. In config.g they are set to 800 milliamps, it would be worth setting them to 1000 milliamps. The motors should still not get warm.

Ian
RepRapPro tech support

@Ian

Ok, will try that. Thx.

---

XBee & electronics blog: [lookmanowire.blogspot.com]

Quote
markbee
Ok "already" at 60mm/sec (I did not alter acceleration speeds) and short moves there is audible rattling in the y-axis possible due to the mass inertia of the bed construction and type of drive. The print is noticeable "blurred" X- and Y-axis-wise in areas where short moves are carried out.

Looks like the bed construction might be an obstacle for higher speeds?

Markus

If you kept the acceleration the same, then the bed inertia is irrelevant. [I'm a physicist by training.] The factors I can think of that might affect performance at higher peak speeds are:

1. The centrifugal force on the belt where it goes round the pulley and the idler will increase;

2. The increased vibration frequency from the stepper might make something (e.g. the belt) resonate.

You could try reducing the acceleration, then do moves along the whole length of the y-axis with the max speed increased. Check that it accelerates and decelerates smoothly (to rule out a firmware issue), and see if there are particular regions of speed where there appears to be a problem. If it the belt is resonating, you could increase the tension to raise the resonant frequency.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Ok plenty of homework for tonight, thx

---

XBee & electronics blog: [lookmanowire.blogspot.com]

the inertia may be irrelevant, but the momentum and impulse are going to be higher at the higher speed whether acceleration is the same or not, surely, and I'd expect that to cause an increase in rattling on direction change (but then I just got a D in physics, which shows my physical expectations aren't to be trusted )

Ray

Quote
rayhicks
the inertia may be irrelevant, but the momentum and impulse are going to be higher at the higher speed whether acceleration is the same or not, surely, and I'd expect that to cause an increase in rattling on direction change

Momentum will increase, but that is not directly relevant. Impulse means the integral of force over time, but I don't see how that matters here either. At a direction change, the firmware is supposed to decelerate and accelerate the stepper smoothly at the specified acceleration. So there should be no difference in the amount of rattle on a direction change, if the acceleration is kept constant and the firmware is working as intended.

I've just tried doing full-length Y-moves at up to 20000mm/min feed rate, and it seems to be working OK, although it is certainly noisier.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

necessary to implement the structure ... have a Y bearing 623 in the plastic that flexes...
I can not tighten the belt more.....

Dario

There's a video on Youtube printing the infill at 200mm/s - see here

I don't think I'd subject my Ormerod to something like that - can't see it lasting more than an hour

There must be a 'sweet spot' for speed vs quality vs machine reliability.

Regards,

Simon

---

RS Ormerod No 192

Quote
Number 192
There's a video on Youtube printing the infill at 200mm/s - see here

I don't think I'd subject my Ormerod to something like that - can't see it lasting more than an hour

There must be a 'sweet spot' for speed vs quality vs machine reliability.

Regards,

Simon

Hi Simon, did you find the 'sweet spot' for speed and quality?

Cheers
Paul

Just a cross-post for you speed freaks - I've tweaked the code a bit to stop the machine banging on certain features, and it seems that it's now comfortable to print at speeds of 100mm/s or more, the firmware (experimental!!!) is here [www.dropbox.com] and the current discussion is in this thread [forums.reprap.org] - dc42 is showing interest, so hopefully there'll be a less experimental fix soon

Ray