[solved] BADCHECKSUM 0

[Anonymous]

Hi,

I set up an installation using the AMS2 and the control software communicates with the board and works fine in manual mode. But as soon as I convert a file and start drawing, the only message I get is something like:

> Opening file /Users/user1/Desktop/bead-chain-gear.jpg…

>
> 4183 points,
Finding greedy tour solution…
Running Lin/Kerighan optimization…

> 0s : 5431,82mm

> 1s : 5278,16mm
Converting to gcode and saving /Users/user1/Desktop/Makelangelo-master/last_official_release/OSX/temp.ngc
Completed.
4191 line segments.
527.6517cm of line.
Estimated 5m38s s to draw.

>
>
>
>
>
>
> M110 N0 M110 N0
> N0 CONFIG T45.5 B-45.5 L-45.5 R45.5 I1 J-1;
> BADCHECKSUM 0
>
>
>
>

Also the > lines tick away without a pause. I am quite sure that I installed the newest versions of firmware and OSX app. Any ideas why this checksum thing occurs?

Thanks.

[Anonymous]

When the PC sends a message to the arduino there is a very small chance that one character doesn’t get transmitted, or gets sent garbled. The software adds up all the characters in a message and the lowest 8 bits form a checksum. On the receiving end the arduino also adds up all the characters and does the same checksum. If they match, the message is processed. If they don’t, BADCHECKSUM [line number]. the PC then knows to resend the failed line. If somehow BADCHECKSUM *also* gets garbled, the PC would then try to send [line number+1], and arduino would send BADLINENUM [line number].

Does the message match what’s in the /temp.ngc file?
Have you tried another USB cable?

I suspect it’s the AMS firmware that’s angry about something but I want to rule out the most obvious solutions first.

[Anonymous]

Hold on…

M110 N0 M110 N0

?

That doesn’t look right. it should be

M110 N0

Hrm…

[Anonymous]

Had the same error. Tried switching out the USB but no luck. Also verified the gcode and tried eliminating the config info but still no luck.

[Anonymous]

I have been running tests to try and reproduce the problem. I will get you a solution ASAP.

[Anonymous]

Great….thanks! Wanted to let you know that I also get the double M110 N0 before hitting BAD CHECKSUM.

[Anonymous]

I’ve just tested release #80 with AMS1 and the latest firmware. I see that I can load a picture and start a drawing.
If I try jog motors or pen up/down before that, the board appears to hang. Probably something got broken when I merged the code for AMS1 and AMS2.

I will be looking into this immediately.

HOWEVER

I will put your name in the credits if you spot the problem in the code first.

[Anonymous]

I am running the latest AMS firmware with the v1 board. I see a problem on one machine and no problem on the other. I suspect the servo is somehow causing a problem, but the issue isn’t BADCHECKSUM. While I’m still looking into this, I’m low on ideas. Anyone?

[Anonymous]

Same problem here.

just got the Makelangelo3 monday, so i thought i was doing something wrong.
in my case the motors are connected with a few extenders.

[Anonymous]

Running latest firmware with AFMS v2. Will try with servo disconnected.

[Anonymous]

Any luck with this?

[Anonymous]

Anyone have luck resolving this issue?

[Anonymous]

I have not been able to reproduce this problem, tho I’m still trying.

[Anonymous]

Some interesting feedback from people who don’t post here –

>:(

– has revealed that the UNO board is apparently generating checksums that don’t match the RUMBA board, even though both are running the same code. This might be some kind of compiler error.

Anyone willing to take a crack at it before I do? search the arduino code for “checksum” and you’ll see the section that generates and compares values. I’ve tried a few ways already that haven’t succeeded.

[Anonymous]

Jenelle might have found the problem.

near line ~833

// is there a checksum?
if(strchr(buffer,’*’)!=0) {
// Yes. Is it valid?
unsigned char checksum=0;
// CHANGE BEGINS HERE
int c=0;
while(buffer[c]!=’*’ && c<MAX_BUF) checksum ^= buffer[c++];
// CHANGE ENDS HERE
c++; // skip *
unsigned char against = (unsigned char)strtod(buffer+c,NULL);

Seems in the arduino ‘c’ was never initialized and produced random checksum values. Wow! So small, so obvious.

[Anonymous]

Updated the FW with the new changes and still get the BADCHECKSUM error. Should I just give up on the AFMS2 and upgrade to a RUMBA? Can someone post a verified working FW file for AFMS2?

[Anonymous]

Got it working by reinstalling Java and running the github version of the Makelangelo app. Now the bot is drawing but dimensions are weird (stretching along Y axis). But that can probably be fixed with some configuration changes. Thanks!

[Author]

Posts