Why do I get these when cuttting 3d?

[knight_toolworks]

I always get these ridges when cutting 3d. this is the first time I had so much flat on 3d so it really shows. the ripples are going across the cut direction. I am wondering if it is a ramping issue. this is on a prt alpha using the default ramping. I cut it at 6,6 ips and 4,6 today with the same effect.
here is a link since I don't want to reduce the pic quality
http://i154.photobucket.com/albums/s266/knighttoolworks/cnc%203d%20work/P1010310.jpg

[myxpykalix]

I have the same problems with some things also. It was explained to me as "resolution". We did a comparison carving with a PRS and it did not have the ridges like you see here.

7504

[cabnet636]

looks like a 12% stepover or above to me, try reducing stepover. i use 10% some use 8

[knight_toolworks]

I used 11% but the ripples are on a flat surface across 90 degrees to the cut.

[jerry_stanek]

your saying that in the picture your bit is running up and down

[eaglesplsh]

Steve - I took a look at your picture... are you saying that your cutter is moving across the grain and leaving those ripples, which run parallel to the grain? I'm having trouble understanding what you mean by "ripples going across the cut direction" and "ripples... 90 degrees to the cut".

I've had to change the resolution on models that I'm importing into PW 3-D. Instead of curved surfaces, I was getting flat ribbons welded together to approximate curved surfaces. The standard resolution when exporting STL files from Alibre was way to rough. When I discovered this problem, I had initially assumed that it was a step-over/machining issue, as Jack has suggested.

[dlcw]

I cut at 8% stepover at 4ips with an Onsrud 1/8" ballnose and get very good results. A little sanding but not the ridges seen in the pictures above. When I started 3D model cutting I used the default setting of 12% and got the results pictured above.

I recently purchased a couple of Centurian 1/8" ballnose bits but haven't used them yet. Anyone have good/bad/indifferent results with this bit?

In Cut3D you can specify a 90 degree second pass. I've been experimenting with creating 2 passes in Aspire 90 degrees apart to see what the results will be. More this coming week after I make a couple of new fireplace mantle faces.

Thanks,

Don
www.dlwoodworks.com (http://www.dlwoodworks.com)

[knight_toolworks]

the ripples are running across the grain. I am not talking abut the tool marks going with the grain that's a no brainier but see the ripples that go across? the first pic shows a little and the second show them a lot. the second I am cutting again the grain top the bottom but the ripples go across. this section is actually flat. so it is not even in 3d.
this was a time issue with a 1/8" bit the estimate time was 15 hours with a 11% stepover and 6 with a 1/4" I used my machine to preview it and I got 6 hours and it took 4
they had to sand anyway to crisp up the edges so the extra machine time would not have save any time really.


7505

7506

[dlcw]

The next two mantles I'm doing will be the first time I've used Aspire to do the models. In the past its been Partworks 2D and Partworks 3D.

At Tim Merrill's suggestion instead of 2 passes at 90 degrees apart I'm going to reduce my stepover to 6.5% and do a single pass. I'll post the results when I'm done. I know that according to Aspires estimate the new plan will reduce carving time by over 4 hours on a 48"x6" mantle carving.

[dlcw]

Steve,

That bottom picture looks like a chattering problem. Dull bit? Cut speed too fast? Dirty rails? That is weird. Does the machine sound likes it's chattering when it's cutting this?

[gc3]

Maybe the wood species and the MC...I achieved similar results from carving Makore(African Cherry)with .25/.125 bits. Nature of the beast?


Gene crain
www.plantasymaderas.com (http://www.plantasymaderas.com)

[navigator7]

Hi Don,
Changing cutter bits has got to be a bugaboo but sanding ridges is too.
My question is: Is it worth the down time to machine the broad open expanses with a bigger ball mill and change bits after the roughing?
As I recall, the program won't let you machine into your finish work if the tool doesn't fit the area requiring carving.
I've only experimented with this with stl models I created in other programs and it worked brilliantly.
Or...am I off base?

Chuck

[dlcw]

Chuck,

I rough with a 1/4" end mill and finish with a 1/8" ballnose. No big deal to change bits. I've done it hundreds and hundreds of times and am getting pretty good (and quick) at it. Maybe someday I'll have the money to get an ATC. That would be the cats meow! :-)

[tparr]

ATC??

[dlcw]

ATC = Automatic Tool Changer.

[gabriel_mckeagney]

I recently had this problem and needed to change the pinions on my prs.
Gabriel

[myxpykalix]

Don the centurion bits are good, you will like them.

[rcnewcomb]

We found that going with the wood grain greatly reduced the visibility of the lines. Try to avoid machining cross grain, especially on open grain woods like oak.

[magic]

I also vote for chatter.

[bcondon]

Steve, are you using a ball nose bit?

I have not started cutting in 3D yet, but I have been wondering is you were cutting a surfaced object, say the inside of a cereal bowl, to the top rim and then back down the outside with a 1/4" straight cutter, (so really only the leading or trailing bottom/side edge would do the cutting) is the software smart enough to know that when it is cutter needs to switch from leading to trailing (or reverse edge)

A better question is: Do you only use a ball node cutter for cutting 3d? (or at least the final pass)

===
I would expect to see those flats if you used a larger cutter that cuts Bottom and Sides even with a small stepover

[knight_toolworks]

yes this was a ballnose. you can do simple slopes with a straight ended bit but it is limited.
it could be chatter but why don't I get it when I did the roughing path or when I pocket? maybe it just does not show up?

[beacon14]

Pinions - that was what I was going to to suggest. When were the pinions last replaced and how much cutting has been done since?

[myxpykalix]

My interpretation of "chatter" is random vibration of the bit and you can see based on the picture that it is uniform pattern of cutting and as James McGrew points out it is more a problem of stepover being too large a percentage causing the grooves.

[knight_toolworks]

the pinions are pretty fresh maybe 100 hours oat most.

[Mayo]

You mentioned that this only happens when you run a 3-D file and not when you pocket. You could run a test, pocketing something with the same stepover used on the job with the problem. Make sure you pocket in the same direction that the problem occured. If it doesn't happen in the pocketing test, then I think it would be software/file related rather than mechanical.

Does this only become visible in the Y direction when the tool has cut in the X direction? Or only visible in the X direction when the tool has cut in the Y direction?

[knight_toolworks]

I always get ripples in 3d like on a dish shape they would be even around the sides all the way around. this has happened since I have had the machine.
I will check it with the same bit when I have a chance.

[robtown]

do you do your roughing and pocketing with a ball nose bit as well?

Maybe I'm missing something, but the striping look pretty uniform to be chatter. I'd say it's stepover based on what I see...

[ken_rychlik]

Steve, Have you upgraded your Y car hold down yet? On my old prt that is what corrected the same ripples that you have.

Pull the four inner rollers riding on the inside of the y rails that keep it in line and leave them off. Then build a bearing block for the Y car on the opposite side of the Y motor to simulate the say way the motor hold the car down on the other side.

Basically the motors and springs pull down the X car to the rails and that works fine. The Y car has it on one side only. If you change it you will see drastic 3d improvement.

My setup may have been a little crude, but it worked great.
7507

[dmidkiff]

This looks like a problem that I saw in one of my projects. I cut with the grain and the grooves were against the grain. Used a 10% step over. I thought it was unusual but decided that it was the wood grain and showed it to some Shopbot and Vectric folks at the Dallas Vectric Convention. Others there agreed that it was in the wood. The wood was german beech. The german beech also left more fussies than other woods. I did another just like it and had the same effect. I did a couple of smaller ones in oak and did not notice the grooves. It should be noted that my pinions have never been changed and the machine is 2001. Hope this helps.

[ed_lang]

Steve Knight,

What you are seeing is a lack of resolution in the Z. Change your Z motor to the 7.2:1 gear head motor like is used on the PRS. I had the same problem on my PRT until I worked with ShopBot to resolve the problem. Another way would be to build or buy a belt reduction box to add to the existing motor. Jay Wyant went this route to solve the same problem.

I heard that one of the ShopBot Camps folks shared the configuration of the OM Driver that helps to reduce, but not fix this resolution problem. By cutting a dish with nothing else in it you can test for this condition. Direction of cut (with or cross grain) does not change this condition.

Hope this helps.

[knight_toolworks]

that makes sense. though so far I have only had two paying 3d jobs so I can't afford to invest in an upgrade.
yes cutting a dish I get the rings all the way around it.
I would like to invest in aspire and now the new motor and or the new z upgrade.

[ed_lang]

You are in good hands with ShopBot. They will take care of you.

[jim_rutske]

Steve, I cut 3d every day and have this same frustrating problem. I use 5 wood different woods and they all do it. I have used onsrud cutters but have switched to centurion and still have the grooves. I'm going to try Kenneth's idea for Y car hold down this weekend. It makes sense and couldn't hurt. Considering i often have to spend just as much time sanding as the bot takes to cut a file...anything has got to help. I have a PRT with a welded y frame and am also replacing the x rails and rollers in the next week. It seems the weight of the welded y car has taken its toll on the x rails.

[cnc_works]

Given the last post by Jim, I wonder if it has something to do with the SB controller. I have a standard 2003 PRT that has seen a fair amount of use, but I have the Ascension gecko based controller and have no issues like those described despite doing a fair amount of 3D including mold work.

Those marks just look too regular to be anything loose or out of adjustment on the machine. Jacks referring to it as resolution problems would seem to buttress my observation.

Anybody with a gecko controller have this problem? Or not?

Donn

[jim_rutske]

I do have the 4G upgrade which greatly increased speed and resolution but didn't do anything for these ripples. It is a pattern and not random. I have watched the z numbers when it runs a flat path and it isn't raising or lowering. My production is really hurting because of the excess sanding time needed. Is this inherant to a PRT? Or am I missing an easy solution?

[cnc_works]

As described above, Jim, I have the PRT and no issues as are described. I would have a hard time considering that the difference is the Ascension box compared to the G4 box, but when push comes to shove you have to ask what the difference is that I don't have the problem and you guys do. Could it have something to do with my running Mach3 instead of SB software?

In a way that would be good news because it would mean there may be an adjustment in settings to improve the issue.

What does Shopbot support say about the crosswise pattern?

I don't think you would see much change in Z since from examining the photos I would think that those ridges couldn't be more than .001" or so high.

Are you saying that you get this even if you are doing a flat raster pocket or cleanout?

I'm leaning toward controller/steppers/software relationship.

Donn

[ken_rychlik]

The bearing on the other side opposite the y motor will help. It won't make it a prs, but it will help.

My prs has the same motors and controller that the upgraded prt had, so I don't think it would be in that area.

The Z is bouncing up and down. Adjusting the inner roller bearings on the y car would help some but the bearing and spring helped more.

Kenneth

[jim_rutske]

Kenneth I think you are on it. I just did a rebuild on my Porter Cable tonight and will build your tension arm idea hopefully tomorrow. I will report back with what i discover...hopefully a beautiful cut!

[applik]

So, did anyone come up with anything new on this? I don't know whether to do the 7.2 upgrade or if it will truly fix the problem. Since all I do is 3D, I'm going to have to find a solution and pretty quick.
chow,
Shari

[harryball]

You may also be getting a carpeting effect while cutting across grain. I had carpet installed once and they had to take it back out. They ran the carpet in different directions coming from the roll and seamed it in the middle. You could see the nap difference on each side. Or like mowed grass. Perhaps the end cuts of the grain are compressed in different directions.

/RB

[jim_rutske]

Shari, I didn't get it totally elimanated but I did significantly reduce the stripes. I changed both bearings in my Porter Cable router and that seemed to really make a change. I don't know what machine you have but we have a PRT and it vibrates pretty bad when moving on the X. We have troubleshot it for hours and can't get the vibration down. Needless to say that vibration translates to a messy carving. I feel that now I have it as good as a PRT can get. We are trying to come up wit the money for a new Alpha. I just can't keep spending hours sanding my carvings like i am now.

[dray]

If you are getting ripples with an end mill this means that your router or spindle is not exactly 90 deg with your work surface causing a slight angle on each pass.

[applik]

I am working with Shopbot to try and get this resolved. We have the Z squared up. We tried carving a file on another bot and the ridges were there. Even checked some of his older carvings, and there they were! Those of you who do a lot of 3D and have the 7.2 upgrade....Did it totally eliminate the ridges? Is there anyone close who has the 7.2's?
thanks for the info,
Shari
Show Low, AZ

[bleeth]

This very much reminds me of all the discussion some year back about ridges along the cut edges in basic parts cutting which was worse with the first alphas than the standard PRT's. The first real improvements were a combination of changing to gecko drives (Many more steps per inch than the old drivers), stiffening the heck out of the old style gantry, and then came Mike Richards manual gearbox.

Assuming tight and properly aligned axes the geared engines will help this immensely. On the often suggested resolution issue the first thing anyone learning when working with a 3-d program is the larger the resolution number the slower the computer but smoother the product. I would imagine that those who have this issue and are doing 3-d professionally already know that.

As far as eliminating it entirely, I know that companies such as Enkebol use a crew of fine sanders to hand finish all of their parts and they are using some pretty serious big iron servo machines. Think about the jewelry makers whose files take many hours to run producing a small wax and then still have cleanup to do. Typical wood parts are pretty darn large so the whole problem seems magnified.

Although 3-d is something I only do occaisonally for "fun" I can tell you that to get the least cleanup I do the reliefs in a very large resolution (over 2 million in Artcam) and will set my stepover for final cleanup at 5%. Makes for some very slow cutting and still has cleanup. For those reading this who don't already know I have a PRT, Columbo Spindle, and Gecko control board upgrade.

Jim: You should not be vibrating in the x the way you say. Something is out of whack.

[cnc_works]

You know, I don't have a lot more to contribute other than an observation. It just seems to me that these ridges are entirely too regular to be anything loose, or even out of adjustment. If something were loose, by definition, it would be varying in pattern. While I recognize that more than one person experiences this problem, I can say it isn't inherent to the machine. I have a PRT96, spindle and gecko drivers run by Mach3. I do a lot of 3D and do not experience any of this problem. Maybe its from control software or controller. I'm still curious what SB support has to say about the issue.

I do get the edge chatter on arcs, worse where X and Y hand off. Now that I would like to solve without buying a PRS alpha.

[applik]

Dave, I appreciate your perspective but we still have a problem. I'm not trying to point fingers at the machine, just feeling a little frustrated at the moment. I also use Artcam, I do hi res and tight tolerances. Agreed it makes for huge files, but makes for nice carving details. I have done tests to the enth degree using different stepovers, etc. Absolutely nothing changes the ridge pattern. They are present regardless. Never random, always consistent in size and direction. The strange thing is they only run two directions. If you carve in the Y direction, the ridges will run in the X direction. If you carve in the X direction, they run in the Y direction. Never at an angle; always X or Y. I've tried carving at different angles, they are always X or Y. That just seems really weird to me. If it were just a matter of sanding, I would have no problems doing the work. My dilemma, you cannot sand them out of a fine texture like hair or grass. The textures would be gone.
That's why I'm trying to figure out if this is something we just have to live with or can it be eliminated by mechanical alterations or adjustments to the machine. Shopbot is trying to help. I think lots of great minds suggesting new and untried possibilities is a good thing. Thanks for all the input.
Shari

[tmerrill]

Shari,

Here is a suggestion for a simple test. Create a simple 3D rectangular component with a perfectly flat top. Toolpath it using a raster technique and look at the finish pass .sbp file in a text editor. I would expect to see a consistent Z cut depth.

Now run the file using some scrap wood and watch the Z readout on your computer screen.

If it is steady and you are getting the ripples, it would point to something mechanical with your machine or an electrical interference problem between the control box and Z axis motor. If you could then attach a dial indicator to monitor ONLY the actual Z axis movement, it may provide a clue of where the problem is.

If the Z readout is fluctuating, then it would point to an issue with your control box or possibly some type of electrical interference within the control box. I agree with others that the consistency of the pattern indicates something other than a random problem, but haven't seen the mention of the possibility it is a frequency based interference.

One final thought. If the above test produces ripples, try a pocket toolpath with a raster in the same direction, using the same bit, cut depth and stepover. It would be interesting to see if that is smooth or rippled.

Tim

[bill.young]

If you have the direct-drive motor on your Z...not the gearbox Z upgrade...I would be willing to bet that this is the same thing that Ed Lang posted about earlier in this thread....that it's basically a resolution issue that appears with shallow cuts. He had upgraded the X and Y motors to 7.2 gearboxes but not the Z, and I believe he resolved it by upgrading the Zaxis to a gearbox motor as well. Hopefully he'll jump back in with more info.

http://www.talkshopbot.com/forum/showpost.php?p=70898&postcount=30

[rb99]

It would be interesting to see if a machine with the ripples got hooked up to a Mach 3 control box if the ripples are still there.

Is there anyone with both controllers and a good 3d file to test?

RIB

[ed_lang]

So, did anyone come up with anything new on this? I don't know whether to do the 7.2 upgrade or if it will truly fix the problem. Since all I do is 3D, I'm going to have to find a solution and pretty quick.
chow,
Shari

Shari-

I had this condition on my PRT alpha and installed the 7.2:1 gearmotors and the problem did go away. It is the resolution of the not gear motor. This can be tested with a dial indicator if you like. Making a gear box or belt reduction unit will also fix the condition.

[applik]

Thanks Ed. It was a pleasure talkin to ya. So my question now is, anyone out there that has upgraded ONLY the Z to the 7.2 motor? Anyone?
Shari

[Brady Watson]

Shari,
Upgrading to a 7.2:1 Z would work very well for you. I used to get those lines in my 3D reliefs when I ran my original 3.6:1 PRT motors and original 3G control board with 1/4 stepping drivers.

I had one of the 1st 7.2:1 Alphas in production, and it has 5 times the resolution as my original PRT 3G. It made a HUGE difference in ALL cutting conditions, especially in 3D.

These lines have NOTHING to do with the ShopBot control system. Running another control system will not make an iota of difference.

-Brady

[applik]

Brady,
Thank you for the insight. Just wanted to know if it had been tried and fixed the problem. Sounds like you've been there...done that. I do have a questions for ya. A botter friend has a prt alpha (alum side rails, two spindles) with a 10/1 gear on his Z (ASM98AA-T10), regular 1/1 on x & y and his makes the little ridges as well. Any thoughts?

thanks again,
Shari

[Brady Watson]

Shari,
The 10:1 motors were found on specialized PRT Alphas for a short time. They were typically installed on tools that had a big 5hp Colombo spindle and/or a 12" Z axis, to overcome the increasingly stronger pull of the springs as the tool went deep in it's travel. I think that the 7.2s are perfect for ALL axes on the machine, as they seem to provide all the torque, resolution and speed that one would ever need out of a ShopBot & then some...

-B

[jshoes51]

Could your z axis be off a little?

[applik]

I want to thank everyone for all their ideas and input.

Shari