Toolpath not cutting the same model twice

[TimSchubach]

I'm having trouble getting Aspire to consistently cut multiples of the same model on consecutive runs. Sometimes the model cuts just fine, and then the next time I'll get ridges or valleys in the model, and I've had to throw the finished piece away. I bought Aspire specifically so I could cut this model, and I have several more that I need to cut over the next couple of weeks, so trying to figure this out is kind of high on my priority list. I've tried to attach the Aspire file, but can't for some reason. Twice, the bit cut too deep, and the bit broke using a Freud 18-100 1/8" ball nose bit for the finish cut. I'm using a 9% step over a fairly slow feed rate of about 50 in/min. Anything faster seems to be too fast for the bit, and as it is, it takes more than 4h for the finish cut to complete. And it doesn't seem to matter what I set the Area Machine Strategy to, I can't seem to get consistent results.

Absent being able to attach the file I'm using, is there any reason anyone can give me for why this might be happening? Words of wisdom welcome!

Thanks,
Tim

[AZRoger]

Tim,

I have a Shark HD 2.0. The stock machine had a plastic collar around a full size router that was then bolted to the Z-Axis platform.
I could not get the clamp tight enough to hold the router steady. Over some runs the router would slide down a little causing it to cut too deep. I'm looking at Rockler's site at the CNC Shark HD3 SlimLine. The mounting collar on that machine appears to be twice as tall and has two clamping bolts. That may be enough. I don't know the configuration of your clamp but it's worth checking.

I my opinion, plastic is just very hard grease. When you heat it up a little it expands and gets slippery.
I bought an aluminum clamp that fits my router from these guys - business@dixiebillet.com .
With aluminum supporting aluminum, the expansion coefficients match and all is good.
I have had no trouble with a sliding router since then.

Hope this helps.
Roger

[Xxray]

Your problem is almost certainly machine related, as AZ says. Many people here myself included routinely cut identical multiples without issue.

As for the bit breaking, that happens when they are pushed beyond their limits, sometimes wood wins out over metal.
You may need to consider doing a roughing pass first, not enough info in your post to make a determination about that. There is no reason why the bit should cut too deep other than an operator error or machine/table flaw. 4 hr cuts are not uncommon, depends upon size of blank, model, desired level of detail, bit used and stepover. Feed rate usually doesn't have a large effect on cutting time as lots of Z movement in 3D files will eat up time no matter what feed rate is set at.

[martin54]

If your having problems with some but not others then it's not an aspire problem but a machine or machine control software problem, I have read some other posts where people have had problems with the original plastic router mount that Roger has mentioned.

[TimSchubach]

Thanks for your replies. Please let me give you more detail. Mine is a Shark HD with a bed that's a little better than 2' x 4'. My version of Aspire is 8.025, but I am upgrading to 8.5 in the morning. My control program is v2.1 build 22, firmware v2083. As far as I can tell, that is the most current version.

The router slipping is definitely something I considered, but it is definitely secure and is not moving.

Definitely understand where wood wins over metal, especially with a bit that small. However, I do use a roughing cut, but I haven't seen any issues with that cut. Generally, I can see the path made by the bit, in this case a 1/2" end mill, but there are big ridges or valleys.

Operator error was also considered, but I'm not sure what I could be doing wrong given that I get good results one time but not the next. In my mind, that leaves machine/control software as a likely culprit. But again, why does it look good once, and not so the next time? Interesting, though, Xxray mentions a possible table flaw - I had not considered that. Can you please let me know what kind of a flaw you're referring to? When I do the setup, I always load the files with Virtual Zero turned on. Speaking of Virtual Zero, why might the control program sometimes ask if I want to load a .tap file using existing preferences, and sometimes not? In my case, for this project, I can't run virtual zero with new setting once the rough cut is complete since the original center point is gone.

Thanks again for all your feedback.

Tim

[Rcnewcomb]

I can see the path made by the bit, in this case a 1/2" end mill, but there are big ridges or valleys.

Posting photos would help us diagnose the issue(s).

why does it look good once, and not so the next time?
...that leaves machine/control software...

If it is a software issue then it would behave the same each time the toolpath files are run.

That means you are dealing with a mechanical issues somewhere.
Check that the material isn't moving while it is being cut.
Ensure the bit is properly tightened in the collet.
Check for binding or play in each axis.
Does the material vary in thickness? Have you measured it?
Does the router return to exact same home location at the end of each run, whether good or bad?

[Xxray]

Well, the table must be perfectly flat and square to the gantry, but then again any flaw is that regards would repeat itself until repaired, so likely is not your issue - But material hold down might well be an issue.
Not all that familiar with Sharks and less with its control system, but if you made 1 good project then the fault must lie in the machine. [What material are you using btw ?] Not uncommon to see ridges in roughing cuts, with a 40% stepover there will be ridges by default and present no problems. If you are breaking a finish bit after a roughing cut, something is definitely not right. I would suggest using tapered ballnoses for 3D work but tapered or not, should not be happening unless you are cutting into something not meant to be cut.

Might want to try something real simple, like say a cutout of a 3" letter A in .5 mdf and see if you can run off a few of them without errors, them maybe try a couple simple 3D hearts or a star in any wood and see if you can run off a few without flaws.

Still would need troubleshooting obviously if there were problems, but sometimes little trial & error tests like this can present solutions or at least eliminate possibilities.

[martin54]

Tim have you asked this question on the shark forum? You gave some more detail but is the computer running the CNC used solely for the CNC or is it doing other things as well? Have you got speep mode, screen savers, power saving etc etc turned off or set to never shut down? Have you got anti virus software either removed or at least disabled? There are lots of things that could either slow down data transfer between the control software & the cnc machine.
What size finishing bit are you using? Have you tried using a smaller roughing endmill, a 1/2" endmill may be leaving a lot of material for the finishing bit if there are a lot of areas it can't get into because of its size.

I do remember a post some time ago where the usb cable on a shark machine was causing an intermittent problem & had to be replaced but not great at searching forums myself

[TimSchubach]

I'll post some photos tomorrow when I cut another model. I'm also going to change the Area Machine Strategy back to raster so I'll have two version to look at.

BTW - my project is 10" x 10" and I use four hold-downs - the wood is not moving while the model is being cut. And yes, the bit does return to the same 0,0,0 position at the end of each run.

No, I have not posted this on the Shark forum, but I'll do that later tonight.

Thanks,
Tim

[Ms Wolffie]

Tim,
definitely a machine problem.
The Shark is notorious for having problems.
You have a plastic router holder that is inconsistent in holding the router tight. Not only does the router slip in the bracket, you only need a minute slip before the cutting changes.
You also have a gantry that is made of plastic and the aluminium reinforcement looks pretty but not do a thing.
When I took mine apart, the sides of the gantry were bent sideways,making the screw move.
You also have a problem with the screws that keep the joiner between the screw and the motor fixed, they have a habit of coming loose.
Check all the screws, tighten them and, as soon as possible, change them with grub screws. I don't know what you call them over there but they are little screws with a round head that are tightened with an allen key.
If more problem, post and I can give you more trouble shooting resolutions.