Yeah nut an easy one.
-We need to see your firmware, $ss, in the webui terminal. 3.9.1 is recommended. There was an old one with a weird issue and at some point I increased the stepper power a smidge.
-2 flute 1/4" endmills are not recommended for the material you are using. You need a single flute.
-3rd is verifying you are not hitting either end of travel.
Nope…I stay clear of the ends
Up to this point I’ve been using the Single flute upcut from your site…
Why are the 2 flute endmills not good?
I’ll check the firmware tomorrow…
Hmm. That’s a bit perplexing. I still think it’s skipping steps. Maybe it’s moving in a way causing the vacuum hose to catch on something?
I’m probably not the best person to ask this. I don’t tend to cut out the same things multiple times so I run mine quite conservatively.
That seems in the realm of reasonable but I’m not great at determining the ideal setting. Maybe that’s slightly slow? I’m not sure. What do the chips look like?
I haven’t used one but my understanding is that to get a proper chip load, that requires an even faster feedrate.
I just went to line 9604 for the start of the part that you mentioned and searched for the next lines containing a Z move (literally searched for “Z” in my text editor starting at that line). If you keep going past the start of the next pass, you see this starting on line 11328, which would be a tab. It moves up, cuts across the top of the tab, and moves back down.
G00 Z-8.5000
G01 X749.7139 Y788.4509
G02 X765.3548 Y784.3083 I-11.9115 J-76.5709
G01 X765.5471 Y784.2360 Z-8.6186 F1548
G02 X768.1646 Y783.1720 Z-10.2500 I-30.3776 J-78.4822
G03 X765.5471 Y784.2360 Z-11.8814 I-32.9951 J-77.4182
G01 X765.3548 Y784.3083 Z-12.0000
Exactly. Much harder to dial in, in easy material it is hard to get the chips out at such high RPM as a trim router.
His cutting federate is a bit more reasonable at 1548mm/min. The 2700 mm/min is his rapids. His DOC and federate are both more agressive than I would run my Primo.
I don’t think so. They’re all G1 moves with that 2700 feedrate defined. The 1548 feedrate is just because it’s using a slower feedrate because it includes a Z axis move.
G01 X589.1113 Y519.2688 Z-9.0000 F1548
G01 X584.7803 Y516.3978 Z-12.0000
G01 X602.5911 Y528.2044 F2700
With FluidNC, it completely ignores any feedrates with G0 commands. Rapids move at the max speed defined in the config.
These are the defaults in Ryan’s configs for X and Y.
max_rate_mm_per_min: 9000.000
acceleration_mm_per_sec2: 200.000
I set my Kobalt Router to 2 1/2 for plywood and MDF and 2 for foam. I was running it at too high of an RPM. When I slowed mine down it went from sawdust to actual chips and the endmills lasted longer.
Thanks Britt,
What feeds and speeds are you running for 3/4 Ply? I’ll try slowing the router down and see if i get more chips than dust…
I’ve been watching that closely because i did have a hangup once on my bootstrapping table cuts.
it’s so perplexing! I know there is a learning curve with any new tool, I’m excited by what I’ve been able to run already…I just need to figure this one out so I have more consistent results 
I have an MPCNC Primo. I use 780mm/m and a 3mm doc.
I know the LR4 can cut faster and deeper, so I don’t know how this compares.
I would try doing some test cuts and adjust the speed of the Kobalt, while it’s cutting, until you get chips coming off. That’s what I did.
I did try the LR4 with 4000mm/min and 6mm DOC in plywood with a 6mm 2-flute. It does work but you need a really good and sharp endmill for it not to skip steps. I am now down to 3000 or 2500mm/min and 4.5mm DOC because I didn‘t want to have to wonder each time whether this cut was going to work. 
This fast used to be a gamble. 
So my 2700mm should be fine… I’ll back off the DOC to 4.5 and see if that helps.
What router are you using and what apees/dial setting?
…and this is with a 2 flute bit?
I’m going to stick to a single flute per Ryan’s suggestion at least until I get a better grasp on the effects of the settings
I lied, it was only 3500mm/min.
Well, I can tell you that my config file isn’t the latest (i have a copy of it on my laptop) Its dated 3/15/24. Im guessing when I run the $SS command I’ll find that the firmware is also pre 3.9.1…
I ordered a micro USB data cable (amazon same day) hoping it arrives before I need to head to the HS.
Any tips other than those found here, Setting up FluidNC would be helpful.
my laptop is running windows 11…
That should be everything. If it is not we need to add any missing info.
One thing I’ve done before to test if it’s the machine or the g-code is to rotate my part 180 degrees in ESTLCam and produce new g-code.
The part will still come out the same, but all the arcs and cuts are now in different directions. If you still get the same step over in the same spot, then chances are it’s something with the machine catching. Lose/tight belt, vac hose catching, bearing being caught by something, etc, etc.
When I’ve had this issue, it has always been because of missed steps. Last time was because of an old burned bit.
If you can pause the print when it happens, cancel it, and return to 0,0, then you can check if it’s in the right spot. I do this from a pendant. In my case, after it misses steps, the machine will return to a spot that’s offset from 0,0.
Solution (see note from Ryan below!) was increasing power a bit (avoid doing that) but mainly using less depth of cut/ better bit.
$65 plywood can also have some really hard knots.
Please be careful recommending this. The firmware has been changed, Current firmware and configs should have no issues.
Actually almost all the firmware should be fine. For some reason there were a couple configs that were wrong somehow, but most are right.
I did that David and the spot happened at the 180 deg point (the same spot on the file) which made me think its got to be a corrupt file issue. Up till now I’ve been uploading via wifi, and to the root directory of the SD card.
I’m going to try manually moving the files from laptop to Jackpot and also place the gcode in a folder as opposed to the root directory. (I believe I saw Ryan mention that it had faster file access that way)
I appreciate your suggestion…
Ryan, Is it worth just trying to update to the October 2024 config file, or should i make sure the firmware is 3.9.1 first?