BOB CAD

[Bill Shields]

heck, I just looked at the MACH 3 programming manual on line, and see nothing about G96 (CSS) being supported.

OK...maybe the manual is old and out of date....???

[Dave_C]

Bill,

Old and out of date, yup, and it will not be updated as they are working on Mach 4 now and say that CSS and FPR will work together.

As far as making a complete Chess set, not really my cup of tea as I am not all that good at Chess. The pawn I made was just something to play with on the Lathe. It is not to scale but kind of cute...

Even if Mach 3 does not do CSS I will be just fine. I'm going to try Milton's setup and see if I can make it work. I did spot some things I was not doing correctly.

Dave C.

[DICKEYBIRD]

OK...maybe the manual is old and out of date....???

Yep. Many, many fixes & updates were made through the years that were not documented anywhere but in the forums. Originally it was a big source of frustration for me since I already had enough trouble trying to figure this stuff out without having manual problems. I've learned enough now not to sweat the small stuff and appreciate the program for what it does for me. Fortunately my needs are simple enough that a forum search usually turns up a good answer or posting a question quickly gets an answer. The people in the forum community are great!

Oh yeah Dave, make sure to backup your .xml and tools3.dat files (in several places) every time you make a config change or set up a new tool. I learned that one the hard way! Come to think about it, do the same for BobCAD if it will let you. Nothing like meticulously dialing in a bunch of tools and later having a "What the he11 was that!" moment and finding out some or all of your tools mysteriously disappeared. Mine was self-inflicted and fairly early-on but made me zealous about the backing-up thing.

[Dave_C]

Ok, I did some research and Bob-Cad had many G code errors in the POST.

So if anyone is using BOB-CAD to post to Mach 3 you really need to make a lot of changes to make it work.

There are four sections that have the line[n, "G50", smax]! Well G50 in Mach 3 is scale reset. So change it to G48 in sections 2,3,4 and 6. G48 in Mach 3 sets the max spindle speed.

Then lines 763 and 764 have the wrong G code as well. 763 needs to be G94 for FPM and 764 needs to be G95 for FPR.

Then lines 785 and 786 are wrong as well. 785 needs to be G96 for CSS and 786 needs to be G97 for fixed spindle speed.

I think that is all so far.

Dave C.

[Dave_C]

I ran a second Pawn with CSS coded properly:

It works just fine but not with Feed per Rev called at the same time. So I just used 6 IPM for the rough passes and 2 IPM for the finish pass. Had to go light as the little ball is only held by about .200" to the main part. In fact if you watch the video, you can see it flex just a tiny bit on the last finish pass.

https://youtu.be/XZl1ZzXjlxs

So CSS works but not with BOB-Cads original POST. Use the fixes above and it works fine. Just make sure you have you tool at the proper X position before you call the routine or it does not calc properly.

Dave C.

[DICKEYBIRD]

Nice video Dave, I keep saying I'm going to do a little video of the ORAC but my camera isn't very good. Gotta remember to keep an eye out for something better.

It sounds like your spindle speed tops out before getting to the smaller diameters. What's your max spindle speed?

I envy your programming skills & ability to edit your own P/P. There's a P/P editing utility supplied with Dolphin but I couldn't make heads or tails of it. Fortunately, Andre over on CNCZone fixed mine for me.

[Dave_C]

Milton,

I capped the spindle in the video with G48 at 1,000 RPM. The lathe will do 1,400 but that is all. It turned smooth as glass at 1,000 RPM.

I contacted BOB-Cad and had them fix their Post Processor. They did!

Dave C.

[Bill Shields]

Capping spindle speed is important...especially with big lathes and non-balanced parts.

Years ago a customer running a hand-written program, loaded a part, jumped to the mid-section of the program and faced off the end.

Naturally the CSS-MAX cap code was at the top of the program, and he bypassed that with the 'start in the middle', so the lathe was running with a 5000 RPM max with a part that should never see more than maybe 150 RPM.

About the time the tool got to X0 and the spindle hit 5000 RPM, the part came loose from the chuck and went flying.

Fortunately the part didn't hit the door and / or come through the glass

Unfortunately, the part went straight down and clobbered the lathe bed.

Very unfortunately, it literally cracked the casting and totally trashed the lathe....it became spare parts.

After that, the director of programming required a CSS RPM cap ANY TIME CSS was called anywhere in the program (which is something that we already did for safety reasons).

Is that something the hobby people need to worry about?

Stop and think about whatever you have spinning around and having it fly loose.

Since most of your machines do not have crash-proof doors in front of the chuck, very often the first flat thing and airborne part would hit would be your forehead...except of course mine since I am such a pointy-head nerd (just ask my wife).

[DICKEYBIRD]

Ouch; that'd get your attention!

Thanks for the reminder. 90% of my stuff is small & chucked in an ER32 collet chuck so I don't have to worry about it too much but definitely another thing to add to the checklist.

I did have a problem a couple weeks ago with G94/G95 though. I was editing a previously saved file in the CAM program & switched to the "Feed per Rev" option. I somehow changed the feed rate to "2" instead of .002, posted and (stupidly) ran it without doing a dry run first. BANG! Scratch one brand new Micro100 carbide boring bar! Imagine that; it couldn't tolerate a 2" per rev feedrate.

[Bill Shields]

which is exactly where you have to be careful...cost of tools goes up exponentially with speed of part.

some controllers would look at 2 as 0.0002" or 0.002 mm, where 2.0 would be what you experienced, depending on the address used. Some addresses shift virtual decimals as a function of units (linear metric or inches), others do not change (time / angle) with units programmed.

Unfortunately there are variants out there that will take 2 as 2.0 or will take 2.0 as 2.0, but if you try to give it 2. - decimal without the following zero, they will alarm and stop.

it is a good habit to get into always stating real numbers (decimal and all and one following digit even if it is a zero) to avoid unexpected results.


About 20 years ago (maybe longer now..time blurrs) here in the Philadelphia area, a fellow who was well-known for making Italian motorcycle hop-up parts was found dead in front of his CNC lathe.

Same sort of thing...speed too high, HYDRAULIC CHUCK could not hold at high speed, part came loose, out through open door, hit him in head.

Hydraulically clamped chucks have speed limits at which the centripetal force of the spinning jaws will overcome the hydraulic clamping force and the part will literally come loose.

[Harold_V]

Hydraulically clamped chucks have speed limits at which the centripetal force of the spinning jaws will overcome the hydraulic clamping force and the part will literally come loose.

Mechanical chucks, too. A good reason for using aluminum for soft jaws when high spindle speeds can be expected.

I've had parts release in several instances. Luckily, without harm. Usually happens when I'm gripping a thin part, trying to chuck lightly.

Harold