Update was fairly straightforward. After reading other comments about the new jerk settings, I battened down the hatches. It was definitely noticable, especially when using the touch probe, but not as bad when running the program. Here is the result of Carbide Create program using build 514:
I did remove the monitor from the magnetic mount. I think the jerk might be strong enough to knock it off.
Next steps will be playing around with those settings and running some more tests. The update of Carbide Create and the Onefinity Firmware will hopefully resolve my Y-axis jumping issue. So far, so good.
@OnefinityCNC I have been communicating with the Buildbotics folks and they just fixed the issue where units default is not reflected on the Control screen. It is in their github as issue #206 and available as a release candidate for testing. Hopefully you will be able to pull this fix into your codebase.
I reduced my new max jerk from 15000 to 10000 and I like it much better. It tones down the jerk while maintaining a much improved speed for 3D carving.
Just installed firmware ver 1.0.6 BETA… wow! Works great. I figured I would leave the max jerk setting at 15000 for the first couple of test cuts. I’m glad I did. The machine has no problem handling the increased speeds. I will leave it at 15000 for now. I do have it bolted to a 4’x8’ table. I guess the weight of the table helps.
After a couple of v bit signs, I loaded a large file I cut last week. This was a 26 inch wide Last Supper 3D carve. Before the firmware update, this file ran for a little over 19 hours. Tonight, after the update, the projected time was down to 13 hours. Exact same gcode… This was using a 1/16th tapered ball nose with a 10% step-over.
For reference, I use the the metric vectric post processor on vcarve pro 10.5.
See photo below, the installed fence on the left is what I produced once I reverted back to 1.0.5. It behaved flawlessly. But if you look at the two fences on the right and the middle, it has done a weird thing where the toolpaths are coming out the right size but not in the right spot relative to each other. I’m not sure if I just stuffed up some setting in 1.0.6 and then went back to defaults under 1.0.5 but the only setting I thought I had played with was the jerk rate which i had lowered to around 7500 just to make it a little less crazy than being at 15000.
I updated my firmware this afternoon. Holy Buckets, those jerk and accel rates are scary fast! I buried the bit during the x,y move on the first carve with the new settings. Do I need to raise my safe z height with these changes? I’ve backed them off to 5000 for now until I figure this out. Thanks
Is there any plan to have the repo be more active publicly? I know there are a ton of moving parts and that adds a lot more complexity just asking for clarity and expectation leveling for us all.
Updated to 1.0.6 Beta and my probing no longer works correctly. Probed Z and X axis with no issue, but when it went to Probe Y it was not moving far enough back to clear the probe block. All of my settings for probing were set the same as 1.0.5 Firmware. Tried rebooting and re-homing and had same issue. Anyone else experience this?
I believe posts require a minimum of 20 written characters, and when answers are short - a few words and therefore less than the 20 characters minimum - people often add the line “20 characters” to allow the post to be sent
Yep, posts must be at least 20 characters. Maybe, perhaps, that is to encourage slightly more verbose responses with salutations or a signature or some basic form of personalization and ettiequte. Maybe something simple like “hello”, or “thanks for asking”, or any other opening that warms the reader up and makes them feel like the response was crafted with attention, care, and enthusiasm. But I could be wrong - it is just be about spam…
From onefinitycnc… Updated 1/15/2021 : We’ve identified a bug that was introduced with 1.0.6 Beta. Apparently, when the jerk is raised beyond a certain number (in this case up to 15000) it affects the z probe height, it adds 1.5mm depth when probing for z in xyz probe button. Reducing it back down to 1000 (but staying on 1.0.6 beta) will correct the z height (or manually zero z). We’re working on a fix.