Drill toolpath problem

[Michael Rytter]

Hi Folks....
I have had my laguna Swift for 4 years now, made probably about 500 crib boards without a glitch, and now suddenly in the last 2 days, on the hole drilling toolpath, it will be humming merrily along....AND SUDDENLY DIVE straight down.....this is random....I can be running multiple boards, and it will have nicely completed the drill path on number 1, running along nicely on the second....and suddenly, randomly dive!

It also does it when I am just running single boards.....sometimes halfway thru, or 1/3.....just random.

All the other paths seem fine...cutout (profile), 3d roughing and finishing....all seem to work well.

I checked for lose wires going to the Z motor, all fine.... And I haven't changed any settings in Aspire....plunge is 30 mm/sec, using peck drilling, retract is 0.02, pass depth is 0.25, cut depth 0.4......

This is so strange....one would think that if Aspires path output files were messed up, the whole drill path would not work as it should at all. And if the controller is getting flakey, why does intensive stuff like 3d cut fine?...(It would really be a mess....but it seems fine)! It is just the drill path (so far)!

Here are a couple pics of what it does....


Any help would be so appreciated!
Tks,
Mike

[mtylerfl]

Hi Mike,

Can you upload the file you showed in the photo? It's always helpful to have the actual file to inspect before we embark on making guesses!

Good to eliminate the simple stuff first by looking at the file when possible. (I realize you have described this as a random occurrence, so the file might be just fine...but, won't hurt to make double doggone sure.)

[4DThinker]

Any chance you are following the router around with a shop vac? I ask because as winter approaches and the air dries out there is a greater chance of static discharges. When I first got my CNC I was doing just that and at a random time a discharge threw off my X axis motor and caused it to ignore a series of commands which ended up ruining the current project.

4D

[mtylerfl]

That was one of my guesses IF the file checks out alright. Still want to see the file to minimize guesses, if possible. (I do have more guesses to offer after eliminating any potential Toolpath issues.)

[Will Williamson]

From the picture it Looks like, you bent the spindle mount Also

[Michael Rytter]

Thanks everyone....Will, that is just my dust brush holder that you are seeing! (Thank goodness)

I pulled the head cover off to see if the drive pulleys on the z motor and the z screw are maybe slipping once in a while, and they're not (they have keyways as well as tap screws into the pulleys), but in doing so, I discovered my Z screw bearings are rough and grindy...with metal shavings in one (pls forgive my terminology...self taught...I'm guessing as to what everything is called! LOL!)

So I have pulled and am replacing them.

As for the file...how do I do that in Aspire? Do you want the saved .crv3d file, or the output file for the machine? (in this case for the Laguna, a .mmg file) or is there some way to generate the plain G code and save it as txt. I also have a desktop shark and I think it's control panel lets you do that....(save the Gcode as txt..or at least look at it line by line before you run it)

The thing is, the reason I think it may be mechanical....this is happening randomly on all drilling paths in multiple different projects...and if I do run the same project twice, it happens in entirely different places each time. If a line of code was bad, wouldn't this happen in the same place every time the project is run? (Again, just guessing here.... self taught and maybe completely wrong! )

What would make Aspire suddenly produce bad code just in one toolpath type? (I'm asking cause I really don't know....) If this isn't mechanical, I am worried it may be the controller (RichAuto A11E), or maybe the Z driver, all of which are easy to replace, but probably hard to get! (And expensive). Thing is, I am not getting any errors on the controller. Thanks for any ideas you guys have....I'll keep you updated as I go...hopefully it is the bearings...

[Adrian]

The only time I've seen results like that and that random on my machines is when there's been interference with the communication to the motors. On my first machine it was cross interference across the wires where insulation had worn. On my current machine it happened with static until I grounded the dust hose and when something has interfered with USB comms on the controller PC. In my case a wireless keyboard going to sleep.

[Will Williamson]

When you say bad code, did you read the NC file and see that the software, wrote the Z move to bury the head?

[Rcnewcomb]

What would make Aspire suddenly produce bad code just in one toolpath type?

Nothing.
Software doesn't break like this. It's easy enough to verify this by looking at the output code.

The more likely culprit is an electrical or mechanical issue causing the system to lose position. Long ago I had issues like this due to grounding issues. The system may be losing Z position due to lost steps, a loose or binding connection, or an electrical issue (static, ground loops, slight break in a cable, control cables parallel with power cables) causing a communication interruption

[mtylerfl]

I agree the problem is most likely mechanical or electrical.

As far as the file request, you need not bother with uploading it now. You have plenty of good educated guesses to troubleshoot from. And yes, the actual Aspire file was meant. If too large to upload to the forum, you would use DropBox or a similar free service to provide a large file for inspection/troubleshooting.

I hope you'll be able to locate and fix the problem very soon and get back on track with your projects!

[IslaWW]

The most likely culprit for this issue is that steps are being lost on the rapid retract. This is usually due to mechanical binds in the Z motion system or acceleration and/or rapid rate set too high for the machine. In many cases its a combo platter of all 3

[CosmosK]

My z axis tried to commit suicide like that one time. I believe the culprit was a not so tightly connected com cable. Hasn't happened since fixing the connection.

[Leo]

I will go with Gary and look at acel / decel settings

[CosmosK]

steps lost doesn't seem to equate to sudden nosedive. Lost steps would be hole is a little deeper or less so. My money is on electrical connection somewhere. If he ran 500 boards just fine, I don't think it's a acc/dec setting.

[rig gap]

What control software are you using...a friend who has a Piranha had a similar problem and he had to get a fix from the maker of his machine because of a bug in the software...maybe even delete and reinstall your control software could be the fix...maybe.