You should be able to use the PWM signal on pin 1 provided your happy with a 3.3V output and changing the scale factor from 0-255 to 0-1023.

byte MoonPhase()
{
int m,d,y;
int yy,mm;
long K1,K2,K3,J,V;
byte PWMvalue;
m = month();
d = day();
y = year();
yy = y-((12-m)/10);
mm = m+9;
if (mm>=12) mm -= 12;
K1 = 365.25*(yy+4712);
K2 = 30.6*mm+.5;
K3 = int(int((yy/100)+49)*.75)-38;
J = K1+K2+d+59-K3;
V = (J-2451550.1)/0.29530588853;
V -= int(V/100)*100;
V = abs(V-50);
PWMvalue = 4*abs(50-V); // 5.12=100% 4=~80%
//pinMode(lowATOPin,OUTPUT);
return (PWMvalue*100)/255;
}

It returns 0-100.
You can comment the kast line and return PWMvalue for 0-255.

I will try and build a non development version of the wxwidgets to use with the stable builds.

Are you able to post your configuration for cmake when building kicad on you machine? What system we you building on? Where did you get the 10.7 SDK from?

Any information you can give will help me improve the build scripts.

Cheers,

Nick.

I believe OSX is under active development but not as higher priority as linux and windows.

Try https://launchpad.net/~jerryjacobs for more info on OSX Development of KiCAD.

I have been following the mailing list for a couple of days, but it is hard to get an overview of which developments are going on. Do you know if KiCad for OSX is under active development? As I said before, it would be great if there was a viable OSX EDA alternative to Eagle.

Try version 2806 and see how that works. I still find it is much faster to use the windows version under wine than the native mac version.

Let me know how you get on.

Thanks man! I’ve got it working using the libraries from the Ubuntu stable release. Can you recommend a certain build for OS X? I’m using 2893 at the moment, but especially pcbnew is very buggy. I’m going to first try out the stable release on Ubuntu in VMware.

I’ve joined the mailing list and will keep an eye on it in the next few days/weeks. It would be nice if there is anything I can do to help make KiCad a viable alternative to Eagle. There are not that many options for EDA suites for Mac users.

Cheers,

David

I also get that message but have some libraries under /usr/local/kicad/ but these will have been installed some time previously. There are no libraries distributed with the nightly builds. The easiest way to get the libraries will be to download the latest official Linux distribution file from http://iut-tice.ujf-grenoble.fr/cao/kicad-2010-05-05-BZR2356-stable-UBUNTU_9.10.tgzand extract the libraries from that (look in /kicad/share/library).

You will then want to add the corresponding path to your libraries path in kicad (look in the Preferences Menu).

More libraries can be found at http://www.kicadlib.org/index.php

Hope this helps.

Nick.

1. When starting eeschema I get the following error(s)

../include/wx/strvararg.h(449): assert “(argtype & (wxFormatStringSpecifier::value)) == argtype” failed in wxArgNormalizer(): format specifier doesn’t match argument type
Do you want to stop the program?
You can also choose [Cancel] to suppress further warnings.

After pressing cancel, the following comes up:
2. The following libraries could not be found:
[then a list of all libraries]

I haven’t been able to find the location of the libraries. Actually, I haven’t been able to find the libraries at all. Are they included in your binaries? Or can I get them from somewhere?

i´m very interested in kits! Is there anyone outside with a working code which implements basic functionality like, Play, Pause, Fwd, Rew etc? Because my “C” skills are very lousy

Nick, keep this project alive

Thanks
bye
Christian

Extruder has an insulator 16mm diameter 40mm long Peek 4.2mm hole through with 6mm long 6mm thread for 6mm diameter 25mm long 0.6mm Welding tip.
The Welding tip is drilled out to 3.2mm for 24.2mm of its depth, leaving a 0.8mm long 0.6mm diameter hole to extrude thru.

6.8 ohms of Nichrome is wound onto the 25mm welding tip with a thermcouple and Kapton tape.

I get actual 0.9mm extruded.
So the calculated figure was 45.930300393787200
When Calculated for a 0.6 extruder tip it would be 19.117710881909400

So Giles is right Die Swell has to be included (it might be more noticeable with a worm drive extruder though as it makes a significant number of steps difference)

Now I have a consistent extruder I hope to revise my maths against what I actually produce.

Have you managed to get good results from using your python script above?

#!/usr/bin/python
import math

def getValue(prompt,default):
try:
print prompt + ” (” + str(default) + “): “,
val = float(raw_input())
return val
except ValueError:
return default

fd = getValue(“Filament Diameter”,3.0)
ed = getValue(“Extruded Diameter”,0.7)
pwd = getValue(“Pinch Wheel Diameter”,7.0)
sr = getValue(“Steps Per Revolution”,400)
gr = getValue(“Gear Ratio”,1)

fl = (math.pi * pwd)
r1 = (fd/2)
r2 = (ed/2)
el = ( (fl*r1*r1)/(r2*r2) )
e = (sr * gr / el)

print “E_STEPS_PER_MM = ” + str(e)

looking forward to seeing a “revised” formula, if you still intend on doing one.

Including postage VAT and Delivery in the UK. If your in London you can pick up yourself and save about USD$10-$20 off the price.

Your formula is the way I do it though, but I always use layer height = 0.8 * diameter and width = 1.2 * diameter. That gives a squashed ellipse with an area of about 0.82 times the rectangle.

Another thing, your Real world examples are based on driving the stepper using half steps (400 steps / rev). Unless I am mistaken the empirical value of 0.706 in the default configuration.h is based on full steps (200 steps / rev) How do you explain that discrepancy?

Anyway, I had much the same trouble as you coming up with the correct E_STEPS_PER_MM value and like you I also had a conrad electronics brass insert. In the end I set E_STEPS_PER_MM to 1.0 and adjust the extrusion amount using the extrusion multiplier in my host software (Which is Repsnapper) This saves me having to re-flash the firmware each time I decide to change it. Its also easier to switch from ABS to PLA that way, since PLA seems to need a larger value.

Renoir

Let me know how you get on as this problem does crop up from time to time.

Cheers.

Thanks for sharing the code on mp3 for Arduino. I have somewhat of a sucess adopting it for an Arduino mega.

I have encountered some problems. The mp3 file I tried to play sounds like it is in slow motion. This holds true even for files with bitrate as low as 32kbps.

I am wondering if you encountered similar issue?

This should not interfere in anyway with the igs file itself.

cp ${file}.d/*.igs ${file}.igs

does this copy all the resulting .igs files into one .igs file? i mean does it merge/append several igs files into one and if so will it work when imported?

I noticed the work you did with brl-cad for converting sketchup files to iges.

I have tried to do similar without success including brl-cad.

First, none of this will work with the windows version of brl-cad.

On linux the converters do run, with no apparent problem in going from binary stl to brl-cad .g dbase files

As you noted trying to run g-iges without putting the pieces into a directory does not work at all.

But even using g-iges with the -m option fails after maybe 10 minutes of executing something. (-v for verbose apparently does nothing)

the file i was experimenting with was a sketchup file from google's own database of community .skp files of a car. Sketchup can read it, and meshlab can see the stl file.

I see this kind of output:

header data:
solid vcg
facet normal -5.169539e-001 -1.819919e-001 8.364434e-001
outer l

Using solid name: s.stl
175140719 facets
98815 faces were degenerate
Making region (r.stl)

Then it goes away for while and then writes a ***bomb***.log file which is almost impossible to read.

I also tried setting tolerances with the -a etc options but no success.

btw i also tried a windows code named cad-exchanger which simply fails out of the box trying to convert the stl file to iges with no explanation either.

Is it possible the 'degenerate faces' are the culprit. and perhaps meshlab is simply more forgiving

thx

mike in utah

You could probably write a windows version easily enough. I don't have any experience with BRL-CAD on windows.

the important commands are
stl-g FILE FILE.G
and
g-iges -m -o DIRECTORY FILE all

We at http://www.arduino.nl ; http://www.floris.cc are planning to make a big load of empty pcb's of the Adafruit Waveshield, Motorshield and Protoshield together with some of our own designs. Maybe you are interested to join. ?

Michel,

GRTX

Contact me (see kicad about of course!)
Jerry Jacobs
- http://www.xor-gate.org

If you *do* need AssertMacros.h, then maybe you can try including boost before Carbon.

As a Brand new C student and now a getting gray beard, this is a great project for me to learn on! I got a version to compile this week with WinAVR 20090313. I changed a few things to get it to build, and referenced the mega169p to match the chip on my B'fly. It is logging away for the night right now, and I am absolutely thrilled! I just HAD to thank you!

Many happy projects to you sir!

Are you planing to sell boards for the project or make the files available?

/temp/kicad-sources/3d-viewer/3d_canvas.cpp:66: error: call of overloaded 'wxGLCanvas(WinEDA3D_DrawFrame*&, int, NULL, const wxPoint&, const wxSize&, int)' is ambiguous

No new SVN version today so hopefully someone is working on a fix for it.

I will check the build logs.

http://www.brokentoaster.com/kicad/kicad_errors-1877.txt

http://www.brokentoaster.com/kicad/kicad_errors-1878.txt

If you have problems of this kind post it in the developers ML, next time

I have uncovered a problem in the eeschema, materials list export tool.

I have images and an error report for you.

I've considered this a secondary issue but i'll take a look on it soon.

Importing a module in PCB edit works now as well.

Nick keep these builds coming the OSX version dare I say could almost be a release candidate.

Thank you for your efforts to trying to resolve this. I downloaded kicad_osx_v1823 tonight, expanded and moved to the /appplications folder. When I tried to run kicad.app, I received the error:

"Link (dyld) error:

Library not loaded: /usr/lib/libiconv.2.dylib
Referenced from: /Applications/kicad_osx_v1823/bin/kicad.app/Contents/MacOS/kicad
Reason: Incompatible library version: kicad requires version 7.0.0 or later, but libiconv.2.dylib provides version 5.0.0"

So not much has changed from my earlier report.

This weekend I plan to update Mac OX X 10.4.6 on the PPC to the latest 10.4.x to see if that improves anything.

I've recompiled the wxMac libraies. I haven't had a chance to test with tiger myself yet.

/Developer/usr/bin/packagemaker –doc ./build/release/osx-package.pmdoc –title 'Kicad' -o kicad_osx_v1811.mpkg

If you can figure out what stupid thing I've done I'd be grateful.

I was not certain where to place all the files, so I decided to place the whole folder in the Applications directory. When I start kicad.app, it crashes with error:

"dyld: Library not loaded: /usr/lib/libiconv.2.dylib
Referenced from: /Applications/kicad_osx_v1817/bin/kicad.app/Contents/MacOS/kicad
Reason: Incompatible library version: kicad requires version 7.0.0 or later, but libiconv.2.dylib provides version 5.0.0
Jun 16 21:23:22 x177 crashdump[228]: kicad crashed"

A google search on "mac libiconv.2.dylib" suggests that this is related to an application built on OSX 10.5 with OSX 10.4 compatibility not having the correct set of options. A google search on "Building on Leopard with Tiger compatibility" yields some ideas for a possible fix: http://forum.filezilla-project.org/viewtopic.php?f=3&t=10220 .

Last login: Wed Jun 10 16:50:27 on ttys000
You have mail.
% ssh mini
% cd /temp/kicad-sources/
% cd build/release/
% /Developer/usr/bin/packagemaker –doc osx-package.pmdoc –title 'Kicad' -o kicad_osx_v1811.mpkg
2009-06-10 17:03:03.284 packagemaker[37763:10b] Setting to : 0 (null)
2009-06-10 17:03:03.337 packagemaker[37763:10b] Setting to : 1 net.sourceforge.kicad.cvpcb
2009-06-10 17:03:03.352 packagemaker[37763:10b] Setting to : 0 (null)
2009-06-10 17:03:03.361 packagemaker[37763:10b] Setting to : 1 net.sourceforge.kicad.eeschema
2009-06-10 17:03:03.370 packagemaker[37763:10b] Setting to : 0 (null)
2009-06-10 17:03:03.379 packagemaker[37763:10b] Setting to : 1 net.sourceforge.kicad.pcbnew
2009-06-10 17:03:03.385 packagemaker[37763:10b] Setting to : 1 net.sourceforge.kicad.pcbnew
2009-06-10 17:03:03.401 packagemaker[37763:10b] Setting to : 0 (null)
2009-06-10 17:03:03.410 packagemaker[37763:10b] Setting to : 1 net.sourceforge.kicad.kicad
2009-06-10 17:03:03.414 packagemaker[37763:10b] Setting to : 1 net.sourceforge.kicad.kicad
2009-06-10 17:03:03.422 packagemaker[37763:10b] Setting to : 0 (null)
2009-06-10 17:03:03.433 packagemaker[37763:10b] Setting to : 1 net.sourceforge.kicad.gerbview
2009-06-10 17:03:03.437 packagemaker[37763:10b] Setting to : 1 net.sourceforge.kicad.gerbview
2009-06-10 17:03:03.446 packagemaker[37763:10b] Setting to : 0 (null)
2009-06-10 17:03:03.454 packagemaker[37763:10b] Setting to : 0 net.sourceforge.kicad.demos
2009-06-10 17:03:03.461 packagemaker[37763:10b] Setting to : 0 net.sourceforge.kicad.demos
ERROR: Could not copy resources to "en.lproj".
%

Thanks a lot for your notes. However im having the following:

Undefined symbols for architecture ppc:
“_main”, referenced from:
start in crt1.10.5.o
ld: symbol(s) not found for architecture ppc
collect2: ld returned 1 exit status
ld warning: option -s is obsolete and being ignored

Do you know how to solve this? I can’t find a a way of telling cmake to change architecture.

Best regards,

Nuno

Is there a forum where we can post any issues we come across, like crashes etc?

Some soldering, Some coding, mostly other things

http://www.timeanddate.com/date/durationresult.html?m1=8&d1=7&y1=2007&m2=4&d2=12&y2=2009&ti=on

–Phil.

Finally the crash-on-exit bug appears to be gone!

Unfortunately the displaying of “rubber banding” lines still doesn’t work.

Thanks for fixing the builds.

–Phil.

After some unescaping, I found a few terms which led me here.

Did you ultimately discover how the culprits were able to get into your site and drop the code bits?