M109 and G29

[wepollock]

Some questions;

1) Is it worthwhile to run M109 R200 T0 after G28; the intended result is to set a target temperate and wait until it is stable?

The Gcode generated by simiplify3d does an M109 xxx T0 -- xxx is the target temperature of the print job.


G90
M83
M106 S255
G4 P500
M106 S0
M140 S60
M190 S60
M104 S200 T0
M109 S200 T0
G28 ; home all
G1 E-3.2000 F500
G1 Z1.115 F1200

How is G29 used; or does it even have to be used?

G32; runs bedg
G29; saved bed map to file

[geneb]

I've never used G29 with a Duet-equipped printer. G32 is so quick I never bother saving the results - it calibrates before every print.

g.

[wepollock]

Gene I am just learning the Duet web interface. So I just figured this out now.

There is an area on the web interface where after you run G32 (click to do so) there is a pull down menu I just found which runs G29.. The G29 seems to start probing on one side of the bed and it works it way down.

I am just trying this feature now at this moment; it looks like you can save the heightmap and then load the heightmap. It will be interesting to see if there is what I expect to see; I expect to see minimal deviation because the glass is flat and the machine is built like a tank.

The one advantage there might be is that the 69 bed probes are saved to a heightmap in a csv file on the SD card. The heightmap can then be loaded before the next print. Although you can run G32 before every print job (which I do because what the #$## it is quick enough vs a 24 hour print job) you really don't have to. Might be better to run G32 followed by G29 every now and again.


I remember reading a white paper on delta printers which studied the error rate on beds. It had a similar error heatmap map that G29 produced. I think the wider the print bed is the more error correction may be necessary.


https://duet3d.dozuki.com/Wiki/Using_me ... mpensation

[img]http://i.imgur.com/73Jv4xN.jpg[/img]

[img]http://i.imgur.com/yB5anCe.jpg[/img]

[geneb]

Ah, ok. Yeah I 've used that but I either forgot or wasn't aware that G29 kicked it off.

What gives you the impression that a wider bed requires more error correction?

g.

[wepollock]

Why would there be more error correction on a wider bed?

Before 32bit ... I don't think the processor was powerful enough to do all the trigonometry involved in delta kinematic movement. Not that the math is very complicated; or intensive; but there is lots of math to do as opposed to a cartesian printer. My guess therefore is that the software simplified coordinate finding by breaking the bed into small triangular parts. (sort of like an aeronautical map) Hence more error potential.

With 32 bit there should be enough processing power to do the trigonometry. (I don't know where the software presently is - DC would know).

How to read the mesh grid map? (aka I don't know how to read the mesh grid map).

Is the hashed part of the grid map close to perfect level, and does the plain part indicate more deviation.

This is the white paper I remember reading with similar error-heat-maps A more accurate statement on my part might be not that a wider bed produces greater errors; but instead where are the most errors going to occur on the bed based the geometry and kinematics of a delta printer.

http://fab.cba.mit.edu/classes/863.15/s ... tics_3.pdf

I am back therefore to the 24v issue. I would expect the steppers to react better in terms of speed (voltage) and hold better (using less amperage).

PS: I really don't want to look at linear algebra again, it gave me a headache more than 40 years ago.. I don't even want to revisit the white paper.. it's far too egg-heady for me.

[dc42]

The motor current you need to get a certain resolution is independent of the supply voltage. The advantage of using a higher voltage (e.g. 24V instead of 12V) is that the stepper drivers can maintain the requested current up to higher speeds.

[wepollock]

The motor current you need to get a certain resolution is independent of the supply voltage. The advantage of using a higher voltage (e.g. 24V instead of 12V) is that the stepper drivers can maintain the requested current up to higher speeds.

As an "end user" I am more interested in "feel" but it's great to think about how/why things work.
In a few weeks I am going to try to move to 24v and I will report on whether or not the "feel" was better; and whether or not the "change" from 12 to 24 was justified.


As an end user I am thankful some very smart guys (DC JJ, Gene, Partdaddy et al) are behind the curtain getting headaches so I don't have to.

Double damn; Now I am thinking about Joules and Inductance.

[dc42]

The "feel" is unlikely to be any different using 24V, except for movement speeds at which 12V is inadequate. See http://reprapfirmware.org to calculate those speeds. However, at 24V the stepper motors may be either quieter or noisier when idle than they are at 12V. Which may depend on how old you are!

[geneb]

So you're saying that 24v enables the Anti-Teenager mode? These Duet boards get better every day.

g.