What's new
What's new

CNC Lathe Conversational--which do you prefer

Mtndew

Diamond
Joined
Jun 7, 2012
Location
Michigan
Are there different variations of the Manual Guide i? Or just the same shitty version they've had for the last few years?

Also, if you had to choose a conversational control for a cnc lathe what would be your preference EXCLUDING the Okuma control? (edit: and also excluding the Mazak control).

We're looking at a few new cnc lathe mfg's and trying to stay under the price point of an Okuma.
 
We have a 2010? Samsung with guide i and a 2006 hurco. We use conversational on both.

Not sure how well either system has been updated from what we have, but I prefer the hurco programming over the guide i. Hurco is more intuitive. Guide i is ok, but usually takes a few more steps to get there and seems more clunky.
 
Basically, Fanuc Manual Guide i hasn't changed that much. There's a few versions with modifications for speedier programming, but generally, it hasn't changed much. I've always preferred Okuma's IGF. However, since that really isn't a consideration for you, Siemens and Hurco both offer conversational style programming that is considerably more powerful than Fanuc.
 
DON'T buy a Siemens in the USA!
The support aint here!

(Unless you are getting some $1/2 mil + 5x Gee Whiz from a Kraut company)


-------------------------

Think Snow Eh!
Ox
 
We have a bunch of Okuma lathes, a newer Samsung (built like a tank) a really old shitty Haas, and 3 older Mori Seiki lathes. 2 of them being SL300 with Yasnac controls.
We're pondering replacing one of the SL300's with something newer and the boss is trying to stay under the Okuma price tag.

That being said the Yasnac conversational is decent, and the Samsung has the Fanuc 0-T control with the Manual Guide-i that nobody likes.
BUT they really haven't taken the time to learn it IMO. The flow of the IGF is nice and smooth and they're used to that.

My guess is that they will eventually steer toward a new Okuma but that's only a guess.


Side Note:
Anyone know if the new Fanuc SMART control is different? What is the SMART control anyway?
We're getting a new VMC that has the Fanuc Smart control and I can't find anything on it.
 
We have a bunch of Okuma lathes, a newer Samsung (built like a tank) a really old shitty Haas, and 3 older Mori Seiki lathes. 2 of them being SL300 with Yasnac controls.
We're pondering replacing one of the SL300's with something newer and the boss is trying to stay under the Okuma price tag.

That being said the Yasnac conversational is decent, and the Samsung has the Fanuc 0-T control with the Manual Guide-i that nobody likes.
BUT they really haven't taken the time to learn it IMO. The flow of the IGF is nice and smooth and they're used to that.

My guess is that they will eventually steer toward a new Okuma but that's only a guess.


Side Note:
Anyone know if the new Fanuc SMART control is different? What is the SMART control anyway?
We're getting a new VMC that has the Fanuc Smart control and I can't find anything on it.

Smart , or plus?

Mastercam lathe is only like $2000 btw
 
The new Fanuc 0i-Plus is an enhanced 0i-F. More memory, larger display with better graphics, optional iHMI, and the Manual Guide i is a little better integrated. Not bad and works well enough, it's no OSP tho.
 
Not sure why you are excluding the Mazatrol control, but for 2- and 3-axis conversational lathe programming, it's hard to beat.....

Sure, a good CAD/CAM man can whip out lathe programs pretty darn quick. But for editing on the fly at the machine, tweaking surface finishes, roughing passes, finish doc's, alternate tools, chamfers and fillets, etc., a good Mazatrol man makes child's play out of creating and optimizing a program, and getting parts tweaked, finessed, to-print, and done.

ToolCat
 
... a good Mazatrol man makes child's play out of creating and optimizing a program, and getting parts tweaked, finessed, to-print, and done.
But not optimized. If you are doing many parts, none of this shit cuts the mustard.

Lathe stuff is so straight up, in this day of cut and paste and easy geometry, I can't see why people are so fixated on cam for 2 axis. I mean sure, for a quicky, but people now seem to struggle with writing simple threading code. And a chamfer, bfd, it's a couple of triangles and a radius or two. Scary.
 
But not optimized. If you are doing many parts, none of this shit cuts the mustard.

And a chamfer, bfd, it's a couple of triangles and a radius or two. Scary.

ON EDIT: I'm not talking high-volume, automotive-automated-balls-out here, but typical job-shop-type work:

Chamfers and radiuses can be a big deal on lathe parts. Especially when between tapered surfaces. Takes some pretty good trigonometry skills (and time) to manually solve the points for a corner radius between a tapered line blending into another tapered line, or a curved line...(Not talking CAD here...)

With Mazatrol is a R.xxx or C.xxx prefix (or suffix for end-point R or C) on the program line. No math, no CAD needed.

With g-code/CAD, it's back to the drawing board. With Mazatrol you tweak the C or R value on that line (which is easy to find in the program because it's abbreviated English words, blueprint dimensions, and no minus Z values) , and run the next part. No fudging, no saying "to hell with it, it's just a chamfer or radius"....

(And I realize Fanuc has radii and chamfer functions, but what a pain-in-the-ass to use.)

Tool nose radius compensation is another pain-in-the-ass for G-code driven lathes. You actually have to tell each tool in the program to do it (often fraught with ticky-tacky alarms), and by not using TNR the more-complex machined shapes just aren't as accurate --- especially the chamfers and radiuses.

Doesn't matter for roughing, but for finishing, proper TNR usage is important on a LOT of lathe work.

In Mazatrol you describe the tool nose radius in the Tool Data page (actually all tool geometries), and never have to think about it again. All toolpaths are calculated using this specific tool data. Radii and tapers and C's and R's are easy, they are accurate, easy to tweak, and by doing so we make a better looking part.

And all this mumbo-jumbo about Mazatrol not creating efficient lathe programs, it's just hogwash. The machine uses said-Tool Data info to automatically establish the closest turret indexing position that clears all tools from the boundaries of your stock. No time wasted there.

Again, you don't have to think about it.

Mazatrol makes efficient cutting paths in turning as well. After all, turning toolpaths aren't rocket science, it's just feed, speed, and DOC. You can easily change anything with the roughing and finishing paths. You have control. And you're not cutting air.

Because it's conversational, doesn't mean it's inefficient.

ToolCat
 
ON EDIT: I'm not talking high-volume, automotive-automated-balls-out here, but typical job-shop-type work:
What the heck, haven't had my daily argument yet (girlfriend is busy this weekend)...

For short runs, sure. But if you're doing hundreds or it's a part that's going to repeat for years, I totally disagree. No conversational control can approach a hand-written program, and for turning, it's just not that hard.

All the geometry problems were solved by the time of NCHQ on a Trash-80. No more sine-cosine buttoning on the HP required. My favorite since is Bobcad 14 in a dosbox, very fast. For some reason all the buttons happen to be in a convenient place, your mileage may vary. But I can do all the correct geometry in a matter of minutes. It takes longer to transfer and type the numbers than it does to figure them.

And I'd never consider using tool nose comp ever anywhere, imo it sucks and no, the correct geometry is not difficult. There's lots more involved than "where the turret indexes" in creating a really efficient and smooth-running program.

With milling, I give up, there's too much to deal with to get a really good program by hand. But 2 axis turning is simple --- and I've never seen a conversational program that did a really good job of even the basics.

Depends on your definition of "efficient" or "good" I guess :)
 
Well, considering that I have never ran a Yamazaki in my life, I wouldn't know - other than what I hear.
... and ... myabe the fella's that I talk to don't know the work-arounds, but:

A simple example was once when a fella asked me about a cut-off routine for a certain app.

I told him that it was no issue, just run in to say .250" and then slow your S and F down 'till it drops.
(or sumpthin like that?)

Or maybe it was more like:
Run your cut-off .020 beyond normal down to .400 or less, and then come back on target and skim that off, and then finish the cut.
(for improved surface finish)

He thought for a few seconds and then said:

"Yeah ... I got Mazatrol. I kan't doo that."


Now I understand that the newer units allow you to add G-code in there as well, but by it'self, canned cycles have limitations.
The only canned cycles that I ever use is threading cycles.

Not sure that helps the OP any tho?


----------------------

Think Snow Eh!
Ox
 
if you had to choose a conversational control for a cnc lathe what would be your preference EXCLUDING the Okuma control? (edit: and also excluding the Mazak control).
Manual guide i to program and have it spit straight G code.
I work with sand castings the majority of time and I can easily edit out the code to what I want it to do.
 
Fanuc's FAPT, which I suspect is the basis of Guide i, was about the best Conversation Lathe Programming System going about. The program generated was all G Code, very rapid to use and could make use of Roughing Cycles or output longhand code for roughing applications.

Where concave features were concerned, the back angle of the cutting tool was taken into account and as much material as was possible was roughed out, leaving an image of the material left by the trailing edge of the tool. Another tool would then be called to rough the remaining material using the absolute minimum amount of air cutting; less than most would program when programming the part manually. Accordingly, whether the program was for a single jobbing part, or high production application, the program generated by FAPT was very efficient.

The only criticism I had of the system, was when multiple concave features existed on a long work-piece. In this case FAPT would decide the order in which they were fully roughed and in some instances, a concave feature closest to the chuck was completed first, resulting in what effectively was a smaller diameter shaft when the roughing of the detail close to the Z+ end of the work-piece was continued. However, it didn't take much time using the system to work out a workaround to force it to rough the part in a logical order.

Regards,

Bill
 
I don't have a preference unless you count that I prefer not to use conversational. Back in the day was efficiency or optimal performance even much of a concern of the folks building these conversational controls?

Wasn't the main focus to make a visual interface so people who knew little to nothing about programming could use and operate the machine? Kinda like the Windows point and click visual interface let everyone be able to use the computer?

What is the "A" word?

Brent
 
Last edited:
.......What is the "A" word?

Brent

Hi Brent. APT. An early programming software. Text based. Very powerful, but quite arcane. A good APT programmer could do some amazing things with it.

FAPT was Fanuc’s version of it. It incorporated graphical prompts for data input. Pretty decent system and useful in some manufacturing environments.

When I was at Fanuc in Japan for some training in the mid-80s, some engineers I spoke with wanted my opinion on why many US customers were interested in conversational programming at the machine. To them in made no sense so they were interested to hear what others thought.
 
He thought for a few seconds and then said:

"Yeah ... I got Mazatrol. I kan't doo that."


Now I understand that the newer units allow you to add G-code in there as well, but by it'self, canned cycles have limitations.
The only canned cycles that I ever use is threading cycles.

Not sure that helps the OP any tho?


----------------------

Think Snow Eh!
Ox

I think since the T2, or at least the M32, so at least 30 years, you've been able to cap your spindle revs just for doing a part off.

So, yeah... He could do that.

I also hear the newer stuff allows short bits of G-Code. Then again so does
my 1984 QT20.
 








 
Back
Top