PDA

View Full Version : Problem Carving Anomoly



HDTHEATER
06-02-2015, 02:37 PM
I have discovered an odd occurrence when I was carving today. I am carving a Hexagonal based shape and the corners have different outcomes as seen here. The top picture is correct and the bottom is not. Apologize that I had not sanded first before taking the shots.

The initial job is set to 400ipm at a depth of .0313 and then a second pas at .0625 using an 1/8" end mill. Material is MDF.

http://i278.photobucket.com/albums/kk106/ebeeks/CNC%20Build/D27C1442-53B5-4D2C-AA01-35C3AC5FD619_zps0jt56u5e.jpg (http://s278.photobucket.com/user/ebeeks/media/CNC%20Build/D27C1442-53B5-4D2C-AA01-35C3AC5FD619_zps0jt56u5e.jpg.html)

http://i278.photobucket.com/albums/kk106/ebeeks/CNC%20Build/B48BB6E7-9011-4286-95B7-E9D3AE1D1C79_zps0ibtu6ss.jpg (http://s278.photobucket.com/user/ebeeks/media/CNC%20Build/B48BB6E7-9011-4286-95B7-E9D3AE1D1C79_zps0ibtu6ss.jpg.html)

I have reviewed the toolpath file in both Aspire and Mach 3 and the anomaly is not there.

I re-ran the file with a single pass at 200ipm at a depth of .07 and the anomaly disappeared. While slowing the job down fixed the issue, shouldn't the machine run at that speed and produce the desired output?

In addition, There is a slight bump in the pocket that is about 1/4" from where the bit stops and goes back as well. This occurred on both runs of the toolpath.

Anybody have some alternate ideas? Is it possible something has worked it's way loose and is causing this? Being able to pick up speed on my jobs was one of the key reasons I upgraded. Like to find a way out of this one.

Thanks,

-Eric

hub
06-02-2015, 02:46 PM
Maybe try in mach3 settings to go from constant velocity mode to exact stop mode.

And remove G64 (constant vel.) code(s) from your g-code if there is. I'm not sure, but I think the G64 over-rides the mach3 setting..

ger21
06-02-2015, 02:57 PM
How fast your machine can go around a corner is dictated by your acceleration settings, as it has to slow down and speed back up to follow the specified path.
When Mach3 is in CV mode, you're telling it to try to maintain a constant velocity. If your acceleration isn't quick enough to follow the path, then it needs to start decelerating earlier, resulting in a rounded corner.

If you want your machine to be as fast as possible, then it needs to be able to accelerate quickly.

Generally, the best option for dealing with this issue is to play with the "Stop CV on Angles >" setting in General Config.You may need to set it as low as 30°. I'd start at 60° and decrease it until you get the results that you're looking for.

CRP
06-02-2015, 02:59 PM
Hi Eric,

I think you need to change the constant velocity setting in Mach3, because the shape is a Hexagon you might want to adjust the Stop CV on angles >

You can find this under Config } General Config282184

HDTHEATER
06-02-2015, 05:11 PM
Thanks for the responses. CV was already set. I need to go back and play with the angles. I did try this job twice and while the same issues occurred on both jobs, they did not result in the same anomaly. I will adjust the angles setting and try again. Will that fix the hump that occurs as well?

As I inspected it closer the hump occurs when the bit is going back and forth in small places (about an inch across), it does not occur when going a distance of 6" across.

-Eric

ger21
06-02-2015, 05:54 PM
What does the hump look like? It sounds like you may be seeing some deflection when plunging?

HDTHEATER
06-02-2015, 07:13 PM
It's just a "halo" of the edge of the pocket. It is smooth. No rough edge on it.

I'll snap a pic when I get back out there.

HDTHEATER
06-03-2015, 12:23 AM
Here is a pic of the "halo" bump I was referring to.

http://i278.photobucket.com/albums/kk106/ebeeks/CNC%20Build/805e67da26290e1895f0472584723853_zpswmiksnul.jpg

I ran another job this evening and while I dropped the constant velocity down to 60, there is stil some evidence of the first anomaly. I'll keep going down until it's right.

As for the hump, it occurred on this job as well. It shows the same "halo" effect. When I ran the job it appeared that the gantry and the motor for the z axis jitters a little when the x axis changes direction. I could not find anything loose On the system.

It could very well be deflection. There is no rough edge and from what I can tell, it looks like the motor makes it too top heavy and causes the mount to flex. I am sure I need to tighten something. I hate to think that 1/32" deep at 400ipm is enough to cause this.

-Eric

victorofga
06-03-2015, 12:37 AM
you can set ""stop cv at angles"" to 120 what actually the hexagon corner..

the corner cv distance set to 180 you can turn it on and set a lower number.. but a lower number combined with higher accel, might causing jerk motion..

so it need some test to find the right settings for your machine..

between passes the line occur in any machine.. it is because of the rotary tool..
as passes are next each other, the fibers are ""bending""
as tool spin on left side it going forward, right side backward... so it makes a visible difference.. what is not a measurable difference..

ger21
06-03-2015, 07:55 AM
you can set ""stop cv at angles"" to 120 what actually the hexagon corner..

It doesn't work that way. It needs to be set to 180-120 = 60, but should be a little less to be safe.

Personally, I'd make sure that all other CV options are turned off.

HDTHEATER
06-03-2015, 12:34 PM
Here is my current screen for Config in Mach 3. CV is turned on and I set the angle down to 60. What other options should I turn off? Disclaimer: I am new to Mach 3 and don't know all the features yet.

http://i278.photobucket.com/albums/kk106/ebeeks/CNC%20Build/CV%206-3_zps0jxnqrv9.jpg (http://s278.photobucket.com/user/ebeeks/media/CNC%20Build/CV%206-3_zps0jxnqrv9.jpg.html)

victorofga
06-03-2015, 01:43 PM
thank you Gerry.. I really didn't know that..

the other cv settings, I think he has to make his own experiment to see and learn how it works.. turning on one and making tests..
otherwise he going to be confused with any advice..

HDTHEATER
06-03-2015, 08:05 PM
I ran another job today and it looks the first anomaly has gone away. However, the "hump" is still there. I ran two passes on it. One at 400ipm and the other at 200ipm. I ran it a third time at 100ipm and it disappeared.

What I observed is that the top of the Z axis has a jitter to it when the y motor changes direction. The centripetal force of the stop and start causes the bit to lift. I checked all my mount points and all is tight.

Is it possible to decelerate as it approaches a stop and then speed back up?

Thanks,

-Eric

ger21
06-03-2015, 11:06 PM
I can't really see the "hump" in the pics, so I don't really know what it is.


Is it possible to decelerate as it approaches a stop and then speed back up?

It always does. Physics won't allow it to be done differently.

One issue with Mach3, especially at higher speeds, is that the trapezoidal acceleration curve can get rather violent, and if there's any flex in the machine at all, you'll see it.

I think on the Settings page, you'll see a CV feedrate. I think it's on by default. I'd turn it off.

HDTHEATER
06-03-2015, 11:56 PM
I can't really see the "hump" in the pics, so I don't really know what it is.



It always does. Physics won't allow it to be done differently.

One issue with Mach3, especially at higher speeds, is that the trapezoidal acceleration curve can get rather violent, and if there's any flex in the machine at all, you'll see it.

I think on the Settings page, you'll see a CV feedrate. I think it's on by default. I'd turn it off.

I'll check the settings page and try again tomorrow.

The pic that illustrates the "hump" is on post #8. The one with the arrows.

If you follow the curve in the pic, it is about a 1/4" away from the curve right where the arrows are pointing too.

I cal it a "hump" there may be another term for it. It essentially a smooth rise and fall about a 1/4" from the edges of a pocket. The hump is bigger the faster I go and it disappears around 100ipm.

-Eric

ahren
06-04-2015, 08:04 AM
It is possible to tighten the v-bearings on the z axis by rotating the eccentric bushings underneath the v-bearings on one side of the axis. If the issue is in fact deflection, this is where I would suspect it would occur. If you'd like to discuss feel free to call us and Cory or I can look into this with you.

Best regards,

Ahren
CNCRouterParts (http://www.cncrouterparts.com)

ger21
06-04-2015, 08:31 AM
I see the pic, but I can't really make out the hump.

victorofga
06-04-2015, 09:14 AM
if the 100 ipm eliminate it, and it is important to looking nice, then use the 100 ipm...
saving 19.5 second and sanding 15 min is not an advantage really..

let machine working a min longer and save a lot on sanding..

I know its very impressive when machine runs with 300 ipm.. but you have to take into account routers are not machining centers..

a good vmc can run even 2000 ipm and making smooth surfaces..
routers, no matter what made they always behind machining centers..

HDTHEATER
06-15-2015, 11:55 AM
My schedule got really busy and was finally able to get back to this. I shot a video of the motor on top of the Z axis. When it stops and changes direction you can see the motor stutter a bit. That stutter translates into the "hump" on the workpiece.

https://youtu.be/WUxHRIiOlro