g2 in Win7 64 about g2 HOT 10 CLOSED

Update, I fired up a linux vm on the AMD box, passed through the usb connection, and am talking to my compiled TinyG2 code via cutecom.

from g2.

So far you have done everything right. The defines should have been
provided by the PLATFORM=due, which is default is PLATFORM isn't provided
on the command line, but adding them (in this case) doesn't hurt.

The yagarto switch was only in the last few commits and hasn't been
documented yet.

You are showing the com port, so that's close. Have you tried CoolTerm to
connect?

I'll try again with my latest commit (Rob branch) and see if I get the same
results. I'll report back here.

-Rob

On Thursday, January 23, 2014, dicepd [email protected] wrote:

Update, I fired up a linux vm on the AMD box, passed through the usb
connection, and am talking to my compiled TinyG2 code via cutecom.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/5#issuecomment-33133328
.

from g2.

I have just downloaded CoolTerm for windows, it will not even show COM10 in its connection list, even though it appears in device manager.

from g2.

As for the defines the only non ifdef SAM3X8E I can find are in,
project.pbxproj (2 matches) // Atmel Studio ?
.cproject (3 matches) // ??
TinyG2.cppproj (3 matches) // Visual Studio
Maybe putting them in the makefile would then not require these to be added, though maybe they are required by these IDEs for debugging identification of the processor.

from g2.

To be clear, I normally don’t use Atmel Studio, myself. I build from the makefile on OS X.

The define is in platform/atmel_sam.mk, which is conditionally included in the main makefile. It uses the CHIP variable passed in from the main Makefile, which is set based not he PLATFORM, which defaults to due. (I apologize that this isn’t all documented yet.)

The “breathing” TX light is simply a matter of heartbeat, and simply means that the board is still working and running through the runloop correctly. I hijacked the TX light for that, and haven’t moved it back to the L light and hooked up the TX/RX lights to communications yet.

I just pulled the latest alden branch and built it (from the Makefile), uploaded it and was able to connect to it from Win7 64-bit in a VM.

I am building using gcc-arm-none-eabi-4_7-2013q3, so there may be some differences. I’ll try 4.8 later today.

-Rob

On Jan 23, 2014, at 10:14 AM, dicepd [email protected] wrote:

As for the defines the only non ifdef SAM3X8E I can find are in,
project.pbxproj (2 matches) // Atmel Studio ?
.cproject (3 matches) // ??
TinyG2.cppproj (3 matches) // Visual Studio
Maybe putting them in the makefile would then not require these to be added, though maybe they are required by these IDEs for debugging identification of the processor.

—
Reply to this email directly or view it on GitHub.

from g2.

I will try my win7 vm then, maybe the vm fudges the usb connection so win7 in a vm sees it correctly.

from g2.

Ok Win7 in a vm sees it and CoolTerm connects! I can also get a response from the port doing a mode command from the command line. Looks like I am going to have to get one of my mini-itx motherboards hooked up to XP sooner than I thought.

from g2.

Hi,

I now have it working on all boxes. The main difference between the VM and native was, the VM had never had the arduino software stack installed. Hunting through the inf file from git I tried a few things but the one that made the difference was changing

[DefaultInstall]
CopyINF=arduino.inf

To:

[DefaultInstall]
CopyINF=TinyGv2.inf

in the TinyGv2.inf file. It would appear as if arduino was installed then the TingGv2 never got copied to the windows system dirs, but it would copy if there was no arduino.inf already on the system.

from g2.

Excellent! Thank you for diagnosing this.

If you feel comfortable with git, you can make the change and issue a pull
request, which gives you credit in the commit. Otherwise, I'll mention you
in the commit message. (I only have "dicepd" to mention as a name, however.)

-Rob

On Saturday, January 25, 2014, dicepd [email protected] wrote:

Hi,

I now have it working on all boxes. The main difference between the VM and
native was the VM had never had the arduino software stack installed.
Hunting through the inf file from git I tried a few things but the one that
made the difference was changing

[DefaultInstall]
CopyINF=arduino.inf

To:

[DefaultInstall]
CopyINF=TinyG2.inf

in the TinyG2.inf file. It would appear as if arduino was installed then
the TingG2 never got copied to the windows system dirs, but it would copy
if there was no arduino.inf already on the system.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/5#issuecomment-33283139
.

from g2.

I'll let you make the change, after playing around with it I find that it is unfortunately not suitable due to lack of backlash compensation. So I am concentrating my efforts on deciding between EMC and Mach. Shame, as I could see various ways that using a separate motion controller would be beneficial. Will track this project though and see if backlash ever gets implemented.

from g2.