[Dan]

For stipple lately I like 2000-2500 pts, max size 4 min size 2 on A2 paper.

I haven’t experimented much with stippling on A4 paper.

[Dan]

Photograph the person on a clean white background. I’ve made a backdrop from a few sheets of paper I would have used for drawing. I strongly recommend A4 or US letter size paper for drawing because it will be ~15 minutes, instead of the more than 1h for A2 paper.

Inkscape can do edge tracing, which can be saved as DXF and then loaded in for quick simplified line drawings. Your inkscape settings will vary with lighting conditions, etc.

Higher contrast is usually better.

[Dan]

Aaah…. The new D10 system makes sure that the software knows which hardware model is connected.
V2 and V3 Makelangelos only have one pulley size, so I hid the option. Less settings to confuse users.
It didn’t cross my mind to add a “custom” machine where every setting can be changed. Your custom machine would have to report D10 V0.

Master branch of the github project now has a Makelangelo (custom) which allows all settings to be adjusted.

[Dan]

Please tell me what steps you are taking to set a custom diameter. I just retested changing the pen settings, they save and load OK.

[Dan]

Our makelangelos use 12v2a power supplies for all versions.
AFAIK they are not doing anything interesting.
I’ve never tried to run them with a 5v supply.

[Dan]

Thanks Dirk. No need to repost the code here. I’ll follow up on Github, which sends me notifications when people post issues there.

[Dan]

There was a bug in Right that was going way too far. It will be fixed in the next release on Monday.

You’re right about the version inconsistency. The v1.4 is the source code version. V6 is the EEPROM version. The difference is really esoteric. I will find a way to make them match so it makes more sense. Probably the next official version of the firmware will jump from 1.4 to 7.0.0 and the EEPROM version will also be 7 so they match.

[Dan]

Some versions of Arduino software do treat the command #include <> the same as the rest.

You’ll have to move the two ./AFMotorDrawbot/AFMotorDrawbot.* files to ./ and change the line that says

#include

to

#include

[Dan]

>My friend change the CLOCK_FREQ the 16000000L to 36000000L, and MAX_COUNTER the 65536 to 955536L and velocity of stepper motor more low and the precision of draw is very better, but i think that it is no good idea.

I have never tried this. I have tried 2x CLOCK_FREQ (32000000L) which made the machine 2x faster.

[Dan]

> the software is superb..

Thanks!

[Dan]

It doesn’t occur here when we upload firmware into new robots. :T

[Dan]

Try like this:

Correct placement of calibration PDF

[Dan]

Yes, you need to synchronize. That is what the “set home” button does – it tells the software that the pen holder is at the location on the screen. you can use the calibration.pdf (https://github.com/MarginallyClever/Makelangelo-software/tree/bb690393628b8442c9e740018b6a7bed16c3ade9/calibration%20pattern) to line up the pen holder with the software. I did not include the paper instructions in the video for the 2.5, I have to edit and re-upload it.

Machine width affects all the math for triangulation. Much more than wire length. Machine height also affects position of paper.

[Dan]

Hrm. Try this:

https://github.com/MarginallyClever/Makelangelo-software/tree/4e0724b717b9698b9f0d27585f407b817ab17031/calibration%20pattern

[Dan]

“m4 calibration.pdf” is included in the download zip file. Print it without scaling.
Position the paper at the top center of the board.
Move the pen holder until it matches the outline of the pen holder on the paper.
Set Home.

Done!

[Dan]

updated. I replaced ‘d’ with ‘step_delay’.

[Dan]

Try this version of the firmware:
https://github.com/MarginallyClever/Makelangelo-firmware/blob/master/firmware_ams/firmware_ams.ino

[Dan]

adjust the feedrate above the manual driving buttons. type in a new number and click “set” next to “engage/disengage”. then moving the pen holder with the buttons should be visibly faster or slower.

[Dan]

you can set the feed rate to be lower than 3500 and you should start to see a difference. The max speed bottlenecks in the AF_stepper library, but anything lower than the max should make a difference. Try 700 and see what happens.

[Dan]

Hi Phil,

In the top of the Makelangelo-firmware you’ll find

// Marginally Clever steppers are 400 steps per turn.
#define STEPPER_STEPS_PER_TURN (400.0)

Change 400 to 200 and that should scale things down.

The top speed of the AMS1 and AMS2 boards is about 2500. In the current version of the software I don’t have a way to tell what board you are using so I leave it at the default for the v3 hardware. makes no difference to the v2 users, makes v3 users happy.

[Dan]

On OSX, if you have JAVA installed, double-clicking the JAR file should start the application. You may run into resistance from security preferences because the app was downloaded off the internet.

I have written a new OSX start in the last few days that is more robust. You can get a copy and try it here:

https://raw.githubusercontent.com/MarginallyClever/Makelangelo-software/dev/start%20OSX.command

[Dan]

You are correct that they end with a semicolon. Does anyone put gcode AFTER the semicolon? Programmer error.

Does Gcode come with lower-case letter or no-space-before letters? No, so no need to cover those cases.

I love that you’re making an effort, that’s rare and special. I want to encourage you. How? The fixes you’re proposing create new problems. By pointing them out I’m sending the “you’re doing it wrong” message instead of the “good try!” message.

[Dan]

The things you have added are not in the specs for GCODE.

Please limit your fix to the actual problem.

What do you bet there’s a compiler option that makes your system process the arduino code without error?

[Dan]

uploading firmware => the hard way (AF_Stepper::onestep)

[Dan]

What about when you have

G00 X10; G00 X-10

?

while(ptr>=buffer)

could be a very long time.

[Author]

Posts