Strange error mid-cut

[Arichards]

Hi. Thanks for your help. I am new to the Maslow scene, and have spent a number of days getting everything prepped for my first project. Things were going great when about 1 hour in to a sign design, the cut decided to erratically change direction. I was watching on the screen as the Black location indicator and red location indicators went out of alignment and decided to go in different directions.

Here was the result. Any ideas? The machine worked so well with incredible accuracy unattended for a couple hours. I happened to be watching it when the error occurred. It didn’t lose precision. Just went the wrong direction.

Thanks so much for your help.

[bar]

That sounds like it might be a bug on our end with the way the machine is interpenetrating the gcode.

Would you be willing share the file that was running?

[Gero]

Welcome,
The log.txt from the Ground Control folder could also help finding a clue, but I guess you are restricted as a new user to post a link or upload files. A few likes and comments will get you over that.

[Arichards]

Here is where the file cut out. I’ll try to upload or link the file itself below.

[Arichards]

Here is the gcode file: https://www.dropbox.com/s/sceu8ej6oitf0ac/combo2.nc?dl=0

[Keith]

@Arichards

It didn’t lose precision. Just went the wrong

Did you stop the cut when it went off course, or did you let it go and find that it eventually got back on course? I only ask because I am wondering if there was some z-move that was missed in the g-code sent to the motor controller.

[Arichards]

I thought the same, but waited long enough to determine to my best knowledge that this hadn’t happened.

What was particularly strange was that the black target (I’m guessing position in g-code) and red target (estimated position of cutting head) were going in different directions on my computer screen. Nearly perpendicular. I let it go fro about 60 seconds before stopping.

[Arichards]

You can see in the shape of the bad cuts, it looks like it is still cutting sun-rays as it should have, but they were just rotated about 90 degrees.

[Arichards]

Here is the log file. Just found it.

log.txt

[madgrizzle]

Is the connection timeout significant? Did you disconnect the arduino?

[Arichards]

I unplugged it once the error occurred.

[bar]

The log is super helpful, thank you!

I can see where things go wrong at around line 350,633.

The machine gets the command G1 X-33.443 Y9.7261

And after that the reported position error starts climbing [PE:2.38,-0.49,127]

There’s no warnings from the software that something has gone wrong, which is making me think it could be a hardware thing.

Will you run the hardware test by clicking Actions -> Test Motors/Encoders?

[Arichards]

Well… that’s a bummer. First test had a motor fail followed by two where all pass. Sounds like that might be indicative.

I pulled the chain on the second test which may have engaged something that was out of place.

Have you seen anything like this before?

[bar]

That sounds like a loose connector to me! Do they all look fully seated?

[Arichards]

Okay. Connectors are fully seated. Noticed two things.

1. When returning to center, left motor made a strange sound. Took apart gearbox. All gears In good order. Resituated everything and now no sound.

2. Left motor controller on board has a slight bump on it. It also keeps losing the cooling fins. I reapplied the cooling fins and secured them with some tape. Maybe overheating is causing the adhesive to fail. Maybe failed adhesive is causing overheating?

Lots of possibilities. In any case, just started the cut again. We’ll see what happens.

[madgrizzle]

@Arichards

Left motor controller on board has a slight bump on it.

Can you describe the “bump” or post a picture?

[dlang]

if there is a bump on the chip, it’s probably blown

a picture would let us see for sure.

[Gero]

A chip with a bump does not sound good.
I have the file running on a simulator for over 4 hours now and are at 12%
So far no sign of buffer overflows or any other issues with the file. Will check where I am in the morning after breakfast.

[madgrizzle]

He can’t post anymore today (I guess forum limit for new people), but here’s the picture he messaged me.

[Gero]

That chip is gone, for sure