Circular retraction Mach3 Turn

[joe]

I cant figure out where this retraction is coming from. We need a Mach3 Turn PP that works. What does it take to create one? I have one from VisualCam that works great. Where did they get thiers??

[Steinwerks]

Is there a "Use radius" selection in the post process dialog and can you turn it to "No"?

Because my guess is the control is misinterpreting the R value.

[joe]

Mach3 setup can either be radius or diameter. I have it set to radius because I couldnt get diameter mode to work. I changed the code to scale 1 instead of 2 and that is when the circles showed up. If I left it as it was at the start the tool backed too far out before it started cutting. With the popularity of Mach3 I am disapointed that we dont have a good Turn PP

[Steinwerks · Accepted solution]

That's not what I mean. In the Fusion 360 Post Process dialog there should be a list of options including one called "Use radius". With that option on Yes you get lines with R values:

N25 G18 G3 X0.7496 Z-1.3265 R0.0312

Turning it to No should generate arcs with IJK values instead.

You can look here for a Mach3 turning PP too: https://camforum.autodesk.com/index.php?board=3.0

[joe]

I tried setting radius to "NO" did not change the tool path. I'll keep looking for a PP that works. Thanks for your help

[Laurens-3DTechDraw]

Easiest way to get a working post is contact your local re-seller and have one made for you.

And I can explain the lack of these post processors. HSMWorks that evolved into Autodesk CAM was originally a milling product. Later turning was added but never really got the attention it deserved so post were not really needed. Besides that HSMWorks was not so much for the hobbyist and woodworking scene so mainly serious metal working machines, so a mach3 turning post wasn't really needed.

With Fusion 360 the software got into a whole different world, where hobby machines are most common and post processors like this are more needed but the development team is busy with making the software better and leaving the post editing mainly to the re-sellers now.

[joe]

Laurens, I emailed Mach3, here is there response

Hello Joe,

I am afraid that postprocessors are solely the realm of the respective CAM program, they are not something that we have or create. But, if your CAM package does not have a Mach3 or Mach4-specific post then a Fanuc 6 or Fanuc 11 post should work fairly well. If you contact Fusion they should be able to help you with getting your post processor sorted.


Thanks,
Brett Price
Tech support staff

Im trying to figure out how to get rid of the strange retraction shown in the first post. Who do you mean when you say "re-seller"?

Is there anybody that is using Mach3 Turn sucessfuly??

[HughesTooling]

What was meant by your reseller was your local Autodesk reseller not Mach3.

Mark

[joe]

I bought my subscription direct from Fusion/Autodesk, So I guess I'm in the right place. No?

[Laurens-3DTechDraw]

---

joe wrote:

I bought my subscription direct from Fusion/Autodesk, So I guess I'm in the right place. No?

---

Included with you fusion 360 subscription is indeed forum support. But not really post processor editting.

That is soemthing you have to do yourself or have a CAM re-seller do. See the list here: http://cam.autodesk.com/buy/

About your problem I'm pretty sure it isn't a mach3 general problem because like the guys from Mach say a fanuc style code should run well.

In the CIMCO Edit backplot the arcs look good and this also shows the orientation of the toolpath. So I think we found the problem. You are turning on the negative side(The side you are standing at.) of the centreline or not? Because if that is true the arcs are different. So G2 and G3 need to be swapped.

[peteymidd]

I get these circles too, did anybody figure it out?

Cheers

[joe]

No, I have tried everything suggested. The work around seems to be editing the code by changing C3 to C1 and deleting the R value after the C3. I am going to run it on the lathe today. Not sure why that line of code is written, that type of retraction is not needed. 

[peteymidd]

Ok, will you let me know how it goes please?

[HughesTooling]

@joe Can you try setting useRadius to yes and change the sign of the R numbers.

So the file from the first thread end up like this.

N11 G20
N12 G50 S6000
N13 G28 U0.

(PROFILE1)
N14 T0101
N15 G54
N16 M8
N17 G98
N18 G97 S500 M3
N19 G0 X0.8 Z0.0562
N20 G0 X0.7869
N21 G1 X0.7066 F10.
N22 X0.65 Z-0.0004
N23 Z-1.3 F5.
N24 X0.7187
N25 G18 G3 X0.7496 Z-1.3265 R-0.0312
N26 G1 X0.8062 Z-1.27 F10.
N27 G0 Z0.0562
N28 X0.6066
N29 G1 X0.55 Z-0.0004 F10.
N30 Z-1.3 F5.
N31 X0.7
N32 X0.7566 Z-1.2434 F10.
N33 G0 Z0.0562
N34 X0.5066
N35 G1 X0.45 Z-0.0004 F10.
N36 Z-1.3 F5.
N37 X0.6
N38 X0.6566 Z-1.2434 F10.
N39 G0 Z0.0562
N40 X0.4066
N41 G1 X0.35 Z-0.0004 F10.
N42 Z-0.1933 F5.
N43 X0.3658 Z-0.2092
N44 G3 X0.375 Z-0.2312 R-0.0313
N45 G1 Z-1.3
N46 X0.5
N47 X0.5566 Z-1.2434 F10.
N48 G0 Z0.0562
N49 X0.3145
N50 G1 X0.2579 Z-0.0004 F10.
N51 Z-0.1012 F5.
N52 X0.3658 Z-0.2092
N53 G3 X0.375 Z-0.2312 R-0.0313
N54 G1 X0.4316 Z-0.1747 F10.
N55 G0 Z0.0562
N56 X0.2224
N57 G1 X0.1658 Z-0.0004 F10.
N58 Z-0.0092 F5.
N59 X0.3079 Z-0.1512
N60 X0.3645 Z-0.0947 F10.
N61 G0 X0.8

N62 M9
N63 G28 U0. W0.
N64 M30
%

Mark

[HughesTooling]

---

@Anonymous wrote:

I cant figure out where this retraction is coming from. We need a Mach3 Turn PP that works. What does it take to create one? I have one from VisualCam that works great. Where did they get thiers??

---

Do you have gcode generated in VisualCam so we can see what working code looks like. 

Mark

[joe]

attached is a pic of the retraction with the neg R value, still have something wierd going on.

Also is the VisualMill code that runs well. with a pic

As i stated eailer if I change the G3 to G1 and delete the R value it seems to be fine, question is why is the R bing put in?

[Laurens-3DTechDraw]

Is that code also from VisualMill?

Because that just linearises the radii.(Not what you want)

And you control understands G2's and G3's so really should use those.(You want true round surface right and not a couple of flat faces trying to be round.)

[HughesTooling]

Replacing the G3 with G1 is not a good idea, on a small arc the flat left is small but for larger arcs you will just get a straight line through the job.

Can you try this code, it's the same as my last post but i've changed the G3s to G2. I guess you tried swapping the G3s for G2s in code in your original post.

N11 G20
N12 G50 S6000
N13 G28 U0.

(PROFILE1)
N14 T0101
N15 G54
N16 M8
N17 G98
N18 G97 S500 M3
N19 G0 X0.8 Z0.0562
N20 G0 X0.7869
N21 G1 X0.7066 F10.
N22 X0.65 Z-0.0004
N23 Z-1.3 F5.
N24 X0.7187
N25 G18 G2 X0.7496 Z-1.3265 R-0.0312
N26 G1 X0.8062 Z-1.27 F10.
N27 G0 Z0.0562
N28 X0.6066
N29 G1 X0.55 Z-0.0004 F10.
N30 Z-1.3 F5.
N31 X0.7
N32 X0.7566 Z-1.2434 F10.
N33 G0 Z0.0562
N34 X0.5066
N35 G1 X0.45 Z-0.0004 F10.
N36 Z-1.3 F5.
N37 X0.6
N38 X0.6566 Z-1.2434 F10.
N39 G0 Z0.0562
N40 X0.4066
N41 G1 X0.35 Z-0.0004 F10.
N42 Z-0.1933 F5.
N43 X0.3658 Z-0.2092
N44 G2 X0.375 Z-0.2312 R-0.0313
N45 G1 Z-1.3
N46 X0.5
N47 X0.5566 Z-1.2434 F10.
N48 G0 Z0.0562
N49 X0.3145
N50 G1 X0.2579 Z-0.0004 F10.
N51 Z-0.1012 F5.
N52 X0.3658 Z-0.2092
N53 G2 X0.375 Z-0.2312 R-0.0313
N54 G1 X0.4316 Z-0.1747 F10.
N55 G0 Z0.0562
N56 X0.2224
N57 G1 X0.1658 Z-0.0004 F10.
N58 Z-0.0092 F5.
N59 X0.3079 Z-0.1512
N60 X0.3645 Z-0.0947 F10.
N61 G0 X0.8

N62 M9
N63 G28 U0. W0.
N64 M30
%

 Mark

[joe]

Guy's lets take a look at the model. It is a Tormach tool holder. all the edges are square where it inserts into the draw bar. (I machine the pocket afterwards) The model tool path shows it trying to chamfer the outside corner, that is my intention. Maybe there is something wrong with the model? Or the way I selected the geometry. I have machined this part with the VisualMill code and it runs as it should, no radius cuts

[Laurens-3DTechDraw]

If there are no radius cuts and your tool has a nose radius your toolpaths are wrong.

You actually need to compensate for them and in your VisualMill code it also doens't use in control compensation so I'm pretty sure they have acrs in there but they are made into small straight linge segments.

Point is that your Mach3 controller does take the code we provide it.

Only the radii are going in the wrong direction. So the normal fix for this would be to write all G3's into G2's and vice versa.

Like said by myself and Mark before, so the question is have you tested that code on the machine? And what happened then?