CNC Just Does What It Wants

[dusty]

I am hoping to find a solution to my problem. The Axiom runs the rough cuts just fine. The finish cuts are a gamble. Five times now the cutter will pick up, move on the Y axis, plunge into the material until the collet stops the Z axis plunge and then it tries to move on the X axis, burning the collet into the material and tearing up my spoil board AFTER it has successfully ran starting out on the finish cut for a while.

I have sent the files to Vectric and they have found nothing wrong with the code and nothing that would create this problem.

I have e-mailed Axiom two weeks ago and have heard nothing. Now it is becoming expensive in lost jobs, ruined materials, burnt collets, destroyed ball end mills, lost time and delays in production.

Any ideas?


This is the ruined material showing where it just did it's own thing.


As you can see from the cuts in the spoil board, this is not the first time.

Thank you in advance.

[pieman]

Try running the program from the controller internal memory. I've seen similar actions when running long programs from the flash drive.

[dusty]

Thanks pieman. I will do that in the morning and will be crossing my fingers. I'll let you know. Other than this the machine is great.

[stevem]

I have run cut files for almost 2 hours from the flash drive and never had a problem!
Are you changing bits after the rough cut? If so, are you resetting your Z with the bit with the puck.

[chuck26287]

Pieman has it right if you're running a 3D finish toolpath from the usb. It's far to data-intensive to run large/complex finish toolpaths from usb... they have to run from the faster access of controller memory. The reads from the USB eventually can't keep up, and you'll see exactly what you're experiencing. No real way to know when a problem will show up, or how it will manifest itself. Just live by the rule of running anything 3D from internal memory only.

[chuck26287]

Nov 8, 2017 22:26:01 GMT -5 stevem said:

I have run cut files for almost 2 hours from the flash drive and never had a problem!
Are you changing bits after the rough cut? If so, are you resetting your Z with the bit with the puck.

If I'm not mistaken, both complexity of the files, and even the robustness of the USB stick can have a say in if/when things go haywire.  Some people can indeed run from USB for some time.  However, as soon as you get feeling confident, the next file runs just a little longer, and that was too long.  Or, you inadvertently use an older/slower USB stick.  Nothing absolute about it... just running large/long files from USB will probably get you sooner or later.  The other thing is the pieces that require longer or larger files are usually an investment in both time and materials.  It's really frustrating to loose them just to not have to take the handful of seconds to transfer the file to internal memory.

[Axiom Tool Group]

Doug,

We received the email a couple weeks back. We apologize, but at the time it appeared that vectric had mentioned that the size of the file may be a problem, and we did not realize that you had asked for further advice.

The gentlemen here are correct, we would recommend running the files from the USB to avoid communication loss.

There are a couple other things that can cause a loss of position, such as speeds and internal controller parameters. If you'd like to email us with the exact model of your machine, we can send you the restore files if needed to fix any parameters that need it.

[garylyb]

Nov 8, 2017 20:01:04 GMT -5 dusty said:

This just happened to me for the first time. was cutting out a piece, and the spindle took a dive.



It's just cutting a 10 inch circle at that point, not a very large file. I was running from the usb.



I'll copy files to the controller from now on.





Attachment Deleted

[dusty]

This time I ran the finish cut from the controller instead of the USB drive. Same result. Tried a different program and again, same result.

 There has got to be something-stupid that is causing this problem. I already called the customer and told them that I cannot complete their order so I am under no pressure there. Didn't want to do that but I had too. Still open for ideas/solutions.

Help!

Doug

[dusty]

Nov 9, 2017 9:56:44 GMT -5 Axiom Tool Group said:

Doug,

We received the email a couple weeks back. We apologize, but at the time it appeared that vectric had mentioned that the size of the file may be a problem, and we did not realize that you had asked for further advice.

The gentlemen here are correct, we would recommend running the files from the USB to avoid communication loss.

There are a couple other things that can cause a loss of position, such as speeds and internal controller parameters. If you'd like to email us with the exact model of your machine, we can send you the restore files if needed to fix any parameters that need it.

The model is the AUTO ROUTE 4 PRO. I tried again and saw that the LED on the controller stated "oL2". Hope this helps.

Doug

[chuck26287]

Nov 9, 2017 13:44:51 GMT -5 dusty said:

Nov 9, 2017 9:56:44 GMT -5 Axiom Tool Group said:

Doug,

We received the email a couple weeks back. We apologize, but at the time it appeared that vectric had mentioned that the size of the file may be a problem, and we did not realize that you had asked for further advice.

The gentlemen here are correct, we would recommend running the files from the USB to avoid communication loss.

There are a couple other things that can cause a loss of position, such as speeds and internal controller parameters. If you'd like to email us with the exact model of your machine, we can send you the restore files if needed to fix any parameters that need it.

The model is the AUTO ROUTE 4 PRO. I tried again and saw that the LED on the controller stated "oL2". Hope this helps.

Doug

Oh... that's a slightly different (and better, as we have more info) picture... looks like you're getting an overload error.  Here is Axiom's comments on the ol2 error indication from another thread...

"The OL error is indicating an overload.



Its not an error that occurs often and may be due to low voltage that is not allowing the spindle the full torque required for that depth of cut.



With a 3HP spindle, at full power you shouldn't have much trouble with that bit unless you are trying to plunge full thickness too quickly.



Read more: axiomprecision.proboards.com/thread/475/vfd-errors#ixzz4xyKQFmmA"

[dusty]

Chuck,

I believe what caused the code was that the collet was trying to push through the material. That would naturally slow the spindle down however the spindle did not turn off/shut down. It is on it's own dedicated breaker and there is a built-in surge p[protector on our meter.

[chuck26287]

Ahhh... the EC was the RESULT of the problem, not the CAUSE of it. Would have been nice if you'd have just been cutting too deep and/or fast, got the EC, then things went haywire. If you didn't get the EC until after the collet hit, you're right... EC is probably not significant to the cause at all. Sorry... just thinking out loud. Sometimes, after you figure out a few things that DIDN'T cause your problem, you see past that to what did.

You'll figure this out.

Out of curiosity... when it does this, is it in the same place? I had a weird issue once where it did some weird moves (nothing like the plunges you're getting). It happened at different points in the run so I eliminated the file. I suspected communications from the HHC to the control box. I disconnected and reconnected both ends of the HHC cable and never had the problem again. Might be worth a try. Might even open the control box and reseat connectors in there too.

[dusty]

It has done this with different files. When I run the same program it happens at random and does not duplicate the error in the same spot. It's getting very frustrating now. I have two signs I have to deliver next week and cannot run them. Very frustrating.

[chuck26287]

I would definitely try disconnecting/reconnecting the cable from the HHC to the control box. Random positions in random files really sounds like it could be random data/communication loss. A CNC definitely produces vibration capable of loosening connections. Also check the twist lock connectors at the Y-motor end of the table (if they're there on the AR4 PRO... I don't have that model and don't know it's configuration for sure).

[dusty]

Well... took connectors off. Blew out with dry air (no dust present), re-attached all connectors, started same program since I left the material in tact. What happened? I did it again, only sooner than previously. Since it starts somewhere near the finish tool path it cannot be the number of code lines.

Thanks Chuck for the help but now it is Axiom's turn to fix this. Seems to be getting worse and more frequent. Axiom had mentioned the possibility of corrupt files so I am waiting for their response to my e-mails and voice message I left.

[chuck26287]

Yep... forums are great, but there's a time to see the professionals. Good luck, and let us know what it ends up being.

[traindriver]

I was surprised to see the comments on large finish files. The longest I've run my machine on a finish file was 28 hours making this for a friend (18" x 18" model).  I've made two of them so far with no problems reading from the flash drive (I hope I haven't just cursed myself) yet.  The flash drive I use is about 10 years old, so it is the USB 1.0 version, which has a slower data transfer rate than new ones.  When you sent the files to Vectric, did you send the cut files from the post processor, or just the model?  I've had mine do some squirrelly things the first time I cut a large finish file (>8 hours) such as decide to move across the work gouging 1/4" paths across the scene, despite the preview showing no problems.  I wound up starting with a fresh file and smaller model to see if I could figure it out.  Apparently, I had corrupted the drawing file while playing around with it to get it to look like I wanted, because after recreating the drawing file from scratch, I had no issues.  My AR8 is 2 years old, so there may be differences there.  Dusty, how big is your sign?  Do you have this issue doing a simple V-carve, or is it with cutting a 3d model?

[redwood]

I also have had no problems running large files off the USB 2.0 drive (knock on wood). I've had my AR6 for over 1 1/2 years.

[dusty]

Well, here is the end result of my problem.

It was not the size of the code or whether I was running it off of the flash drive or direct from the controller. With the help and guidance from Chad McCarty it was resolved. Bottom line... I was running the feed rate too fast in my finish cut. I initially had it set up for a rate of 200. Chad suggested 100 maximum. I changed the feed and the CNC has been performing great ever since.

Chad had a lot of patience with a very frustrated end user (me) and was a great help. My hat is off to Chad and Axiom. I have a mew saying, "My machine runs good, just like an Axiom should".

[jwardy]

Dec 15, 2017 7:22:12 GMT -5 dusty said:

Well, here is the end result of my problem.

It was not the size of the code or whether I was running it off of the flash drive or direct from the controller. With the help and guidance from Chad McCarty it was resolved. Bottom line... I was running the feed rate too fast in my finish cut. I initially had it set up for a rate of 200. Chad suggested 100 maximum. I changed the feed and the CNC has been performing great ever since.

Chad had a lot of patience with a very frustrated end user (me) and was a great help. My hat is off to Chad and Axiom. I have a mew saying, "My machine runs good, just like an Axiom should".

Dusty, for those of us still dealing with these issues, can you answer a few more questions? Can you tell us the size of bit you have been using, # of flutes, depth of cut per pass, total cut depth, and material being cut?

Thanks!

[stevem]

Chad is great at solving problems. Axiom seems to always be very helpful when I call.
Glad everything worked out.

[dusty]

Dec 15, 2017 19:48:58 GMT -5 jwardy said:

Dec 15, 2017 7:22:12 GMT -5 dusty said:

Well, here is the end result of my problem.

It was not the size of the code or whether I was running it off of the flash drive or direct from the controller. With the help and guidance from Chad McCarty it was resolved. Bottom line... I was running the feed rate too fast in my finish cut. I initially had it set up for a rate of 200. Chad suggested 100 maximum. I changed the feed and the CNC has been performing great ever since.

Chad had a lot of patience with a very frustrated end user (me) and was a great help. My hat is off to Chad and Axiom. I have a mew saying, "My machine runs good, just like an Axiom should".

Dusty, for those of us still dealing with these issues, can you answer a few more questions? Can you tell us the size of bit you have been using, # of flutes, depth of cut per pass, total cut depth, and material being cut?

Thanks!Yes. I was using the tapered ball end, 2 flute. Depth of cut was .02 for the final finish pass with the step-over set at 10%. Being a very detailed carving, the Z axis was constantly in movement. 200 ipm was just too much too fast. I never would have thought of slowing it down so I am grateful for Chad's input. Simply slowing down to 100 ipm maximum obviously produces a better finish but it completely eliminated the problem of the "Z Slip".