V23 drags tool across work face in visual check?

Hi folks, Pulled out and dusted off V23 for use with a machine here. Received and setup a simple nested SW file from a customer to have parts ripped off and when I run the verification through BobCad the virtualization show the cutter being dragged across the face in certain locations and I'm not entirely sure what is going on. It most notably occurs during profiling operations, but I've also had it happen during engraving, drilling, and sometimes pocketing operations but not nearly to this extent. This has been a recurring issue here in the past, but I've always worked around it by breaking up the operations into separate posts and just loading the individually into the machine driver software. Not an issue on a small single part but clearly a huge time disadvantage here where I have to cut out ~60 little internal pockets.

It's almost as though the program is not moving up to the proper rapid position before moving to the next profile to be cut. There are some major drag points but if you look closely at the jpegs(screen shots of the actual test run) you can see it happens in more than just those places(red)

Just wondering if someone could look over this and see if there's something I've got messed up here, or maybe this is a post processor issue? I have attached an .rar file that contains the V23 file, the original SW file, and two screenshots that show my current issue. The machine is a Torchmate Routermate and I'm using the torchmate_router_rev1 mill post processor that was available for download off their site(CNC Machine Post Processors | CAD/CAM Software | BobCAD-CAM)

Thanks!

If you have not actually run the program yet to verify this, I would READ your program and see if the tool is programmed to go to the proper clearance height. Normally in our machining that is .100".

It may be just a simulator error caused by your video processing card.

Mike

I do not know why you have top of job at -8.9925,,but that is not it.The simulation shows bad here too.But Preditor Editor Back Plot shows good.So I don't know,I can say PE backplot is very reliable.Other than looking at all the code (ughh)I would say it is good to go ? ? ?BTW,I used my Post Processor as I have no yours.
Here is back plot photo

Good idea. I will do that in a moment here. Looks like i better brush up in my gcode skills here. Crash course anyone :)

The rapid clearance for this particular file is set at .1". Many times im no worried about machine runtime or speed so ill often increse this to .25" or more to ensure adequate clearance. I have noted in the past though with this issue that no amount of rapid clearance remedies the problem, even if i set it to 2" or more. While nobody is particularly fond of tool crashes in this case im running 1/8" CVD endmills to cut carbon plate and while not the most expensive cutters out there at over $100 each theyre not something i want to risk snapping off on a whim that its just a graphics error.

Just pulled up my pc specs here for reference:

3ghx amd phenom
8gb ddr3
Win7 64 bit
Graphics card is an ATI Radeon HD 4850 X2(2gb mem card)

Nothing crazy but gets the job done for the most part. Ill double check to make sure the card driver is up to date here too

jR the reason for the negative numbers is because of how this particular machine is setup hardware wise, and where it references Z0.0. Machine homes to the proximity sensor in the z axis at the top of its travel and sets home there, all subsequent movement of that axis is down from that point(negative). Its a pain in the arse to setup and orient files into the CAM for the machine but ive gotten used to it by now. I just havent had the time nor real desire/need to setup the home switches to reverse it all to "normal"

Interesting,I took and translated your stock to zero.Changed all your top of stock to zero.Re-selected all geometry and computed tool path and it simulated right.
Here is simulation photo.


I am strongly thinking everything will run fine.I believe it is a simulation error,I get it too.Preditor Editor back plot is really good about showing what really is going to get cut and how.However,I give no Warranty,lol

Quote:

---

Originally Posted by Rennfab

jR the reason for the negative numbers is because of how this particular machine is setup hardware wise, and where it references Z0.0. Machine homes to the proximity sensor in the z axis at the top of its travel and sets home there, all subsequent movement of that axis is down from that point(negative). Its a pain in the arse to setup and orient files into the CAM for the machine but ive gotten used to it by now. I just havent had the time nor real desire/need to setup the home switches to reverse it all to "normal"

---

That would never Fly in my Universe,lol
Maybe someone more familiar with your set-up can offer a better way.It just seems wrong.

Interesting indeed. I wonder if this is just the integrated BC simulator not computing for the negative numbers properly? I mean ive never really had any issue with any machine cuts in this fashion when the simulation does check out. Ive never run code that didnt check out 100% in the simulator. I used to have a demo version of predator but that ran out long ago and i never purchased a license for it. Just been making do with the integral system.

I cant imagine it being the fact that its in the negative range actually causing the issue as i dont think bobcad would have designed theinterface to allowit if the program couldnt comput it, but then again stranger things happen.


Hmmmm. Let me check some hardware and drivers here and report back..

Thanks for the help sofar guys!!

Quote:

---

Originally Posted by jrmach

That would never Fly in my Universe,lol
Maybe someone more familiar with your set-up can offer a better way.It just seems wrong.

---

Hah, i like how you think. Yeah i do need to set it up different. Need the downtime on the machine to do so though. Not in the cards just yet

One thing to remember this machine was initially designed as a plasma cutter, which always travelled down until the torch tip contacted the metal to be cut, then debounced a preset amount to creat the necessary torch airgap. Their routermate variant worked good for what i do with it mostly(cutting foam molds and plugs) but ive since beefed it up to do much more. Just never got around to converting over the home and limit system to work like a normal mill.......yet

I checked the drivers. Theyre up to date

I also looked at the code. My gcode skills are very rusty but i think its doing what it should. Lines 486 to 507 look to bewhere the first and second profile cuts occur. Its showing a rapid movement to z-8.8925" which is correct being set for .1" rapid clearance. Lines 501-504 are where i beleive the machine transitions betweem profiles at this location. It completes the arc, rapids to -8.8925", then moves down the x axis before plunging into the material again.

Correct me if im wrong above....

The simulation made it appear that lines 502 and 503 were happening simultaneously(amongst other places in the simulation) but this doesnt appear to be the case according to the code itself when i open it in an external editor.

I guess ill take a shot and defer to your judgement that this is a simulation error. Ill load up a super cheap 1/8" endmill and turn it loose on some tooling wax and see what happens. Worst case i break a $5 mill and some tooling wax instead of a $106 CVD and some carbon plate :/

I give my thanks guys, really appreciate the insight here. This reminded me of another issue i have been having but ill post about that at a later time.

Cheers!

-Adam