Marlin, gen7, pronterface, issues

I have a Prusa with gen7 1.3. It ran Teacup and was working good. I wanted to try Marlin for the look ahead feature so after some messing around i managed to upload it. Pronterface connects but says something about an error and try fixing the error and restarting. Then says something about temperature. (sorry on my phone at the moment, don't remember exact error message). Here's a strange part, once i've manually started the psu using m80 and try to move the motors nothing works but the homing buttons for x and y. And those work wrong, they move the axes about 10mm away from the endstops. I picked the gen7 1.3 board in the configuration.h file and set every thing else to what i thought it should be. I'm a bit puzzled about that movement thing.. Anyone have any suggestions?

Errors from Pronterface >
"Printer stopped deu to errors. Fix the error and use M999 to restart!. (Temperature is reset. Set it before restarting)"

Once M999 is issued >
"Error:0
: Extruder switched off. MINTEMP triggered !"

So my min temp setting is a problem. I've tried this setting at 0 and -10 and still the same issue. My temp sensors don't measure properly until the PSU is on as this Gen7 1.3 board doesn't have that feature...

Another factor in it may be that my heatbed thermistor doesn't start reading above 0 until it hits 40 degrees actual temp. But once at 80 degrees reads pretty accurate up until at least 110 degrees.

Edited 1 time(s). Last edit at 02/03/2013 02:11AM by Phizinza.

You can try Cyberwizzard's Teacup which does feature look-ahead: [github.com]

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

That would be even better (easier)... I'm having an issue compiling it in Arduino though

I get this:
dda_lookahead.c: In function 'dda_join_moves':
dda_lookahead.c:219: error: 'for' loop initial declaration used outside C99 mode
dda_lookahead.c:461: error: redefinition of 'sreg_save'
dda_lookahead.c:219: error: previous definition of 'sreg_save' was here
dda_lookahead.c:461: error: redefinition of '__ToDo'
dda_lookahead.c:219: error: previous definition of '__ToDo' was here
dda_lookahead.c:461: error: 'for' loop initial declaration used outside C99 mode

I've tried in Arduino 1.0.1, then I downloaded 1.0.3 and the 2.1 support package for gen7. Same error. All I can seem to find out about it is Arduino isn't supporting some C99 (somewhat like C++?) code which is in the lookahead feature file. I don't know C, and have limited programming experience so I'm at a loss...

Thanks for the bug report :-} I've copied it to a Github issue: [github.com]

Looks like ATOMIC_BLOCK() isn't compatible with Arduino IDE. Certainly solvable, but right now I have no quick fix.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Is there an alternative to Arduino IDE to compile this firmware?

Traumflug Wrote:
-------------------------------------------------------
> Thanks for the bug report :-} I've copied it to a
> Github issue:
> [github.com]
> 40
>
> Looks like ATOMIC_BLOCK() isn't compatible with
> Arduino IDE. Certainly solvable, but right now I
> have no quick fix.

I concluded the same thing in the bug report as I missed this thread :-)

The quickest way I can think of is renaming the file to cpp which will trick the Arduino suite to run g++ instead of gcc which should compile it fine as well (after adjusting the headers to properly mix C and C++ of course).

@Phizinza: Yes there is: use the makefiles. If you edit Makefile-AVR and set the correct port and CPU speed etc the firmware can be built using:
make -f Makefile-AVR clean all

If that worked, upload it by running:
make -f Makefile-AVR program

So I managed to get Cyberwizzard's Teacup uploaded using WinAVR installed with the Programmer's Notepad and using the default WinAVR make and program tools in Programmer's Notepad. I had to change the /etc/ttyUSB0 part in the Makefile to COM4 but other than that it worked.

I am however having a problem that now the prints are even worse around corners, the bed seems to be jerking a lot. I have tried jerk on 5, 10 and 30 and it seems the same on all settings. Infill also is suffering a massive quality drop.

Any suggestions? I'm going to continue trying changing settings...

I managed to get Marlin on now. I had to do the standby power mod for the thermistors to my Gen7 electronics and also include in my tables *OVERSPAMPLR after the first set of numbers just how all the other tables have it.

For some reason I have to manually start and stop the PSU using M80 and M81. It also does a strange thing where the first 5% of the move is really slow, than it accelerates to full speed, back to very slow for the last 5% of the move. Is this normal?

Phizinza Wrote:
-------------------------------------------------------
> So I managed to get Cyberwizzard's Teacup uploaded
> using WinAVR installed with the Programmer's
> Notepad and using the default WinAVR make and
> program tools in Programmer's Notepad. I had to
> change the /etc/ttyUSB0 part in the Makefile to
> COM4 but other than that it worked.

Glad that it worked: I haven't had time to rewrite the atomic barriers to make it work again in the Arduino IDE.

> I am however having a problem that now the prints
> are even worse around corners, the bed seems to be
> jerking a lot. I have tried jerk on 5, 10 and 30
> and it seems the same on all settings. Infill also
> is suffering a massive quality drop.

It seems like you are hoping the lookahead would somehow improve your print quality: it will not. Lookahead is meant to speed up prints using settings which give you an excellent print result already.

The only time I can imagine you improve print quality if you have heat issues and moving the system rapidly prevents overheating your prints.

Lookahead is a method to keep the toolhead moving around curved paths (lots of small straights), speeding up the printing of such segments. To move from one straight segment into another without doing a full stop means you jump from one motion vector into another at an acceptable speed. Effectively you assume the mechanics of the machine will handle the instantaneous change in direction. A value of 30 is quite high: when set too high you tell the printer toolhead/bed to barely break before heading in another direction which will give you violent vibrations. I currently print with this set to 6 or 7, which gives me awesome speedups while printing circles without rattling the printer to bits and pieces.

If you have quality issues, I would suggest trying to fix those bit by bit. Perhaps Marlin gives you a better starting point but I reckon every firmware will require you to do manual tweaking, not to mention that the settings of Skeinforge/Slic3r will make or break your print.

The printing quality issue I am having is with curves. It gives a jagged edge. From what it looks like and what I read this is because the printer is stopping on every facet end and starting again; this gives the print head time to ooze just a little. I was told that look ahead removes the stops from happening and will therefor prevent the little ooze.

I finally managed to get Marlin working and it is doing a superb job with my curves just using all the old Teacup settings that I spent a few hours honing in. I still think I need to tune the look ahead jerk and also the acceleration a bit more. But it is looking much more hopeful now!

Thanks for the help.