1/8th inch drift recently on Y-axis, past issues with random diving into material

[antaesthetic]

I've been searching through the forums recently and finding similar issues that others have experienced, but frequently they end with no final solution.

Sorry this is so long, but I wanted to cover all bases.

I have the Shopbot Desktop with Spindle, and have been having a few intermittent issues with new and old SBP files. Some were created last year, and I've run them plenty of times this year with no problems. But the past two months I started having issues with two particular things:

1. The first was a newly created file 2 months ago. A simple light v-carve across a board, tracing the outlines of text for an event. The letters ranged from 1-3 inches tall. The machine would occasionally stop, hover there for a few seconds, then drop at full speed and dive into the material/work surface. It happened at least 6 times throughout the project, and I found it happening particularly around edges of small things like colons and around a lot of the numbers of a particular font. I ended up deleting a bunch of similar characters from the project and replacing them with a different font, and this fixed some of my issues, but then had it happen a few more times on a V a capital S and an E. I haven't had this happen again, but I've also never run anything like it since. It makes me nervous, it's like a little evil dive-bomber.

2. The second issue happened yesterday, and it was similar to another post I read earlier. I was running an old file for some candle holders I make out of cheap pine, and everything generally seemed to be going well. But on my third file for the profiles which has 6 steps, the second step down seemed to shift in the Y-axis by about 1/8th of an inch. When it did so, the machine also made a terrible grinding noise, but then kept going for a bit. I thought maybe it had hit a knot at first. When it tried to make it around the next corner (while still following the coordinates in the X axis appropriately), it ground again and came to a sad juddery stop before I could even react and hit the button.

I went back into my VCarve file, deleted the profile that was offending things, and had it start at a second copy of the profile next to where the other one carved. It did almost the exact same thing in the same places (just maybe in the opposite direction + or - on the Y-axis). The top left edge where it's going left around a curve, and the bottom right where it goes into another curve.

Strangely, I was able to run another profile to the right of this one, and it worked out fine... And then immediately after, I got overconfident, and the exact same juttery jamming thing as the first two happened again.


From other posts, people have suggested the following:

Check Firmware Update for Control Box/SB3
Check USB Speed and Connections
Check RAM
Grounding to prevent static
Disable internet and all unnecessary programs while running SB3
Check corner speed and ips/ipm


So, I started from the Firmware Update & Control Box update. I was behind a few, so I upgraded to the most recent (3.8.42).

Then I ran Memtest86 to test my RAM, because my laptop is a bit slower than it was when I first got it, but it didn't find any errors there.

Then I checked my USB Speed and Connections, made sure I was connected to my USB 2.0 ports, and did the tests in Diagnostics to check the connection speed after using an MX,4 command. I got a variable rate between 17.5-18.1, close but still below the 20...

Feeling somewhat confident that the firmware update helped, I decided to leave my spindle off, zero my z-height higher than usual and run an air cut. Aaand, well, sure enough... it happened in the exact same places again. Same juttery noise around the top corner, probably the same offset, though it didn't come to a sad halt because it wasn't running offset into the side of the material like before.

Remembering what someone had said about testing USB speed, I decided to halve the ipm and go from 3 in/sec to 1.5 in/sec. I did the aircut again and it ran smoothly. Wonderful little ATARI-sounding noises, but nothing scary. So I have a little confidence again that the machine can still cut things fine, but only if I run at 1.5 in/sec...

The thing that gets me, is that I haven't changed those files, and as far as I know I've run them before at that speed, unless my controller or SB3 is interpreting them differently now for some reason. The 3.0 ips is the default setting in the Tool Database, so I haven't changed anything to my knowledge...

I have not yet tried to do any of the other steps that people have recommended-- the grounding, disabling programs, internet, etc. But that's mostly because this has all happened without much change to those things. I will try running a test tomorrow at full speed running Windows in a more bare-bones program state.

But because of the consistency of error at 3.0ips, that weird offset, and the scary dive into the workboard, does anyone have any more clear direction on what to check next?

--I've attached a few pictures showing where the noises/stalls and offsets occurred on the sbp file, as well as the result showing the step-like effect that happened on the second step of each pass of 6.--

[Tim Lucas]

Good morning

Its winter time and static is stronger in cool dry air so you need to make sure dust collection is grounded.

How long has it been since you checked / replaced the pinion gears

[jerry_stanek]

Can you post the CRV file something doesn't look right in the preview. that you posted.

[antaesthetic]

Thank you for your reply. I recently added a ShopVac as dust collection in the early Summer, but when I did the air cuts it wasn't connected and it still had the same problem in the same places. When I was running the sign a few months ago, I don't think I had it running at all since it was barely grazing the surface.

As for the pinion gears, the answer is "never"... I've only had the machine for a little over a year or so though with relatively light use. Would the pinion gears slip at 3.0ips even without any resistance, but still be okay at 1.5ips?

[antaesthetic]

I will tonight—I'm not at home at the moment. But if it helps, I've run that exact sbp file at least 20 times before over the course of the year. Only now has it begun to act strange, even with a new export of the profiles from the crv.

[Tim Lucas]

Thank you for your reply. I recently added a ShopVac as dust collection in the early Summer, but when I did the air cuts it wasn't connected and it still had the same problem in the same places. When I was running the sign a few months ago, I don't think I had it running at all since it was barely grazing the surface.

As for the pinion gears, the answer is "never"... I've only had the machine for a little over a year or so though with relatively light use. Would the pinion gears slip at 3.0ips even without any resistance, but still be okay at 1.5ips?

Yes they can slip and faster ips will slip worse, the set screws can back out and cause slippage or the gear may have trash (dust) built up and cause slippage - I would loosen the motors clean and service the pinions and rack, lube and then try again at 3 ips air cut

[antaesthetic]

I apologize for not replying sooner about what worked. A pet-peeve of mine is when questions like this don't get followed-up on forums!

I spoke with Tom at Shopbot and he gave me many things to try, and I DID end up finding a solution (seemingly, at least) for my problem. TL;DR — For me, it turned out to be the fact that my computer has become increasingly bogged down, and my system resources were actually competing with the communication to the Shopbot.

But here's what was recommended, and what I tried:

- Because I have a Shopbot Desktop, I was politely informed that I do not have pinion gears, let alone a rack and pinion system. :p

So from there I was asked to check my bearing blocks, and was given information on how to test if they were worn, slipping, or dirty.

Communications tests:

- I ran an MX,20 test, my Packet_ET showed 20.0, which is within acceptable range.

- I did a quick air cut with the spindle off to test the issue again as a control, with all of my typical programs running, including Adobe Creative Cloud which is a memory hog.
It had the error in the second iteration of the cutouts rather than the first. The first seemed to "cut" properly at 3.0ips, so I let it keep going, but got startled by the slamming noises in the second step down of the second cutout in the series. So I moved on to more tests.

- I then went in and closed every possible running program and process that could be using active memory including virus protection, User Account Settings to "Never Notify," changed my Power Settings and Sleep Options to not put various things to sleep after a period of time, set ‘USB selective suspend setting to 'Disabled,' disabled my Wi-Fi and Updates by putting the computer into Airplane Mode, and ran everything again at full speed. It worked!!!

So, for me it was a communications error, somewhere along the line. Something in settings fixed things, and it could've been anything from the USB suspend, to the active memory (Adobe CC recently updated and is a bit more of a memory hog than usual.)

I hope this helps someone else out there who might be having similar issues. Let me know if you have any questions about specific settings, I can go into more detail if you need it.