Solidworks for $10 a month or $99 for the year

[Bob La Londe]

Well, if businesses were more honest, you would still have rapids The primary reason for a lot of the changes were businesses were violating the license terms and using the free version.

And if AutoDesk had been honest I might be a subscriber. I quit using it years ago when I saw the first instance of their normal business practices popping back out again.

And for comparison I've bought almost every upgrade of ViaCAD Pro since I started using it with version 5.

I am a big believer in businesses telling me the damn price and letting me make a decision from the beginning. When they use drug dealer practices you know they're not going to be honest.

[djc]

The major problem with the free version CAM is that they neutered the rapid moves..

Can you post a snippet of the actual G-code? What does it do - just just put G01 everywhere you would expect a G00? Does F360 allow you to change feedrates at all? If it is possible to have a load of G01s with differing F's associated with them, it should be easy enough to do some search and replace post-post processing on the code.

[Sparky_NY]

Can you post a snippet of the actual G-code? What does it do - just just put G01 everywhere you would expect a G00? Does F360 allow you to change feedrates at all? If it is possible to have a load of G01s with differing F's associated with them, it should be easy enough to do some search and replace post-post processing on the code.

Yes, it replaces the G00's with the most recent G01 feedrate. Yes you can spec the feedrate for cutting operations. Find and replace would also replace the desired feedrates used for cutting not just the offending ones. Find and replace is not as easy as it would seem, you would also screw up the feedrates for the cutting operations causing crashes, broken tools etc. Autodesk considered people attempting to work around it no doubt when the implemented it.

No snippets of code, I dumped Fusion 360 when they did the last batch of "revisions" on the free hobby version.

[Sparky_NY]

If you're running CNC and do ALL of your CAM with Fusion, well... You should be paying the subscription. You pay more in a year for your daily coffee than the piece of software that controls all of your machines...

Didn't they also remove 4 axis CAM from both the free and the paid version you have? How do plan on implementing the 4th axis you are building?

[djc]

Find and replace would also replace the desired feedrates used for cutting not just the offending ones.

After I posted, I went searching and it is possible to find workarounds. For one thing, you are never going to do your rapids at the same Z- setting as your cutting moves, so there will be a Z-move immediately preceeding and immediately succeeding what you want to be a rapid. G-code is a language in the same way as C-code and you just need the right parser to make it dance.

[RB211]

Didn't they also remove 4 axis CAM from both the free and the paid version you have? How do plan on implementing the 4th axis you are building?

They removed simultaneous 4th axis, really one two strategies, but they gave you additional 3+1. 3+2 strategies that used to be paid extra.
My personal reality is that 99% of what I have done so far has been 2.5D, barely touched true 3D. LOL.

[Bob La Londe]

Most of the average 4th axis work I have seen is either just 3 axis work substituting a rotary axis for one of the linear axis, or 3+1 in a format I consider just indexing. Before setting up a real 4th axis I often did 4th axis work with a spindexer. (Actually a pair of spindexers face to face.) I'd insert simple manual code to shut down spindle, turn off coolant, position the head away from the work, and pause the machine in the operation header in my simple 3 axis CAM software. While the machine was "parked" I'd manually index the work piece and then press start. While I am not and never will be fan of drug dealers (no offensive to any licensed pharmacists in the crowd) 3+1 and 3+2 is more than adequate for most things I would need to do.

One might argue that "turning" spindles is true 4 axis, but most times it really isn't. You may often choose a different 3rd axis location before using the 4th axis, but usually the actual operation is still only 3 axis using 2 linear axis and 1 rotary axis. Again 3+1 or 3+2 is more than adequate.

I have to be honest. I have a hard time wrapping my head around true simultaneous motion control of more than 3 axis. I understand spatially what is happening, but struggle with how to program it and locate the work piece continuously if the other two axis are rotary work piece axis which is common.

[Dan Dubeau]

Most of the average 4th axis work I have seen is either just 3 axis work substituting a rotary axis for one of the linear axis, or 3+1 in a format I consider just indexing. Before setting up a real 4th axis I often did 4th axis work with a spindexer. (Actually a pair of spindexers face to face.) I'd insert simple manual code to shut down spindle, turn off coolant, position the head away from the work, and pause the machine in the operation header in my simple 3 axis CAM software. While the machine was "parked" I'd manually index the work piece and then press start. While I am not and never will be fan of drug dealers (no offensive to any licensed pharmacists in the crowd) 3+1 and 3+2 is more than adequate for most things I would need to do.

One might argue that "turning" spindles is true 4 axis, but most times it really isn't. You may often choose a different 3rd axis location before using the 4th axis, but usually the actual operation is still only 3 axis using 2 linear axis and 1 rotary axis. Again 3+1 or 3+2 is more than adequate.

I have to be honest. I have a hard time wrapping my head around true simultaneous motion control of more than 3 axis. I understand spatially what is happening, but struggle with how to program it and locate the work piece continuously if the other two axis are rotary work piece axis which is common.

When you get to that point, you get off your wallet and let the software handle simultaneous motion control for you. Having done a fair bit of true 5 axis work early on in my career, I now do pretty much the same work, but with 3axis/sometimes a 4th. Programming and running the 5 axis work is dead simple compared to doing it with 3axis. Seems counter intuitive, but you can get yourself out of some tough situations when you can just rotate the head to get a small ball in places you normally other couldn't etc. With only 3 axis you have to get really creative sometimes.

TBH most of the time we could have got away with simple 3+2 indexing, and sometimes we did, but when you have a fancy toy, you use a fancy toy, so there was a lot of true 5 axis toolpaths used when we really didn't need too. Pretty cool to watch them run. Was fun to program too, I really miss it. Software (mastercam) and machine control (heidenhain) handled all the thinking.

[Rich Carlstedt]

When you get to that point, you get off your wallet and let the software handle simultaneous motion control for you. Having done a fair bit of true 5 axis work early on in my career, I now do pretty much the same work, but with 3axis/sometimes a 4th. Programming and running the 5 axis work is dead simple compared to doing it with 3axis. Seems counter intuitive, but you can get yourself out of some tough situations when you can just rotate the head to get a small ball in places you normally other couldn't etc. With only 3 axis you have to get really creative sometimes.

TBH most of the time we could have got away with simple 3+2 indexing, and sometimes we did, but when you have a fancy toy, you use a fancy toy, so there was a lot of true 5 axis toolpaths used when we really didn't need too. Pretty cool to watch them run. Was fun to program too, I really miss it. Software (mastercam) and machine control (heidenhain) handled all the thinking.

Dan , you nailed my comments as well and I am in total agreement with you- most work can be done with 3 axis and some creativity
I did full 5 axis programming back in the late 1980's using a GE 2000 Control/Software , a true Parametric programing package. In 1988 (~) GE went/joined with FANUC and dropped a magnificent program . And we had no CAM or post processor, it was all math equations.
I don't think MasterCam was even around then. What software were you using back then ?

Rich

[Dan Dubeau]

Dan , you nailed my comments as well and I am in total agreement with you- most work can be done with 3 axis and some creativity
I did full 5 axis programming back in the late 1980's using a GE 2000 Control/Software , a true Parametric programing package. In 1988 (~) GE went/joined with FANUC and dropped a magnificent program . And we had no CAM or post processor, it was all math equations.
I don't think MasterCam was even around then. What software were you using back then ?

Rich

This was in 2004/5 and we were using Mastercam V9. X had just came out, but we all switched back to V9 as it was clunky and unreliable.

My hats off to you guys that figured that stuff out before CAM. That kind of thing would be completely out of my wheelhouse.

Were you guys doing production parts, or 1 off stuff? I could see production parts (impellers, screws, repeating features, etc) being easier to program via math equations, but for mold and die work/3d surfacing My brain would explode just thinking about it.

Most of our need for 5 axis was for compound angle pocketing/holes, 3d surfacing, rest finishing and undercuts. It was really nice to tilt the head on the side and cut a freeform surface off the side of a ball instead of the tip also. This was checking fixtures and mold/die work.

Edit: As for the fixed rapids in fusion, it should be an easy text editor fix with find/replace. I used to do some universal head undercuting programs by programming in the xz plane in cam and doing a find/replace swapping the z/y axis in the .nc file. Sounds sketchy, and it is, but it works. More 3axis limitation problems.

[polaraligned]

The major problem with the free version CAM is that they neutered the rapid moves, they happen at the feedrate rather than rapid speed. This can make a 3d part take 10X longer or more to make. That was the change that got me to abandon Fusion, I liked it otherwise.

Most guys search G1 in their Gcode editors and replace with G0 where appropriate. I know, it is a pain in the ass, but what do you want for free. I doubt Solidworks even has CAM for their $99 package.

[RB211]

Most guys search G1 in their Gcode editors and replace with G0 where appropriate. I know, it is a pain in the ass, but what do you want for free. I doubt Solidworks even has CAM for their $99 package.

Haven't looked at it yet but it DOES have cam, 3 axis only.

[polaraligned]

Haven't looked at it yet but it DOES have cam, 3 axis only.

That is very good....Solidworks isn't known for giving anything more than they have to. So it would be good to have a side by side comparison of what you really get with Solidworks vs Fusion.

[RB211]

So I have Solidworks Connected behaving as I would like now, click on the taskbar icon and it loads. As said before, getting to that point was just stupid!

[Rich Carlstedt]

This was in 2004/5 and we were using Mastercam V9. .........................................Were you guys doing production parts, or 1 off stuff? I could see production parts (impellers, screws, repeating features, etc) being easier to program via math equations, but for mold and die work/3d surfacing My brain would explode just thinking about it.

Most of our need for 5 axis was for compound angle pocketing/holes, 3d surfacing, rest finishing and undercuts. It was really nice to tilt the head on the side and cut a freeform surface off the side of a ball instead of the tip also. This was checking fixtures and mold/die work.........................

Dan, we were doing Die Parts and most of the work , like 90 % was with Ball Endmills. and it was almost all surfacing- One off -
The value of the GE 2000 control cannot be understood or valued by most CNC guys because they have never seen it . It was a Parametric program with full math abilities ( you could use any equations !) and your G Code entries used parameters if you wanted (99 allowed) and get this, you could change a parameter WHILE the program was running. Say you want a Z depth of -1.000 and you are doing a peck cycle , Well instead of writing Z -1.000 you say Z Par 32 instead....and that means the control looks at what Par 32 says and in the parameter box 32 you write -1.000 , now as the endmill is pecking to that depth , you can go in and change Par 32 to - 1.625 and the spindle will update and go to -1.625 for finished depth.....got it ?
So the control constantly scanned all the parameters and updated itself during production
Now here is the neat part .Put a Math formula in the box, like ( Sine Y/ 2) and now when the spindle bores, it has to look at where is "Y" at that moment and then calculate the Sine and divide it by 2 to know how deep to go. ..so what you get is a curved depth based on where Y is at .

Now what if Y is a function of X ..? Yes you now have multiple calculations done at the same instant . The 4 axis result is rotating a round part while modifying the cutter depth while moving in the X axis and the Y axis in a curve at the same time ... 1985 fashion ..and get this. it was done on 1 or 2 sheets of paper, not 2,000 like G code generators do today

And GE dropped the project because they thought after they brought it out that the CNC machinists could not handle the math and going to a CAM maker and FANUC was the future.
Now, you could manually code the 2000 like any other G machine , and many did that , but the real value was using the Parametric package for complicated work
Our boring mill had 6 axis's but the control was limited to 5, so you picked which ones to use

Rich