The Sequel - Oh Blimey I bought a CNC Lathe (Beaver TC 20)

[awemawson]

I'm sure I'm telling you how to suck eggs, but I've re read your post and you didn't explicitly say that you would normally put tools in a rear turret 'facing down', ie the chip curls on the floor side of the insert, not the ceiling side. M3 is then clockwise from the rear of the chuck  etc....

PK I chose to use mostly 'facing up' tools as a/ the lathe came with many of this style VDI 40 holder and only a few of the facing down type, and b/ the cutting forces are correctly down into the ways rather than tending to lift the turret.

From an operational point of view the only disadvantage is the way the chips fly, and it has the huge advantage of easy access to the tip for changing, and visibility whilst cutting is in progress.

But you might well be right that this is the root cause of my confusion !

The fact remains that M03 makes the chuck turn clockwise looking from the tailstock on this lathe

[awemawson]

Surely Beaver didn't convince Siemens to reverse the Beaver lathe control's spindle rotation away from the world standard M3/M4/M5 conventions just for their machine?

Milton, Siemens needn't have been involved - all it would take is to reverse the wires to the spindle motor armature, only two wires as it's DC. (The two from the tacho would also need reversing of course)

[awemawson]

Just to convince any doubters, myself included, here is a little video showing an M03 starting the spindle !

[russ57]

No there isn't...

Russ

[awemawson]

Strange - it's marked as Public 

Maybe it need time to cook 

I came across this paragraph in the  Generation 2 commissioning notes - (NB mine is a Gen1) Of course they don't say the view point, but if the convention is to look from behind the chuck  then it would confirm that as mine is set is intentional


later edit: I've just tried the YouTube video on the Wife's iPad (so that I'm not logged in) and it shows fine

[seadog]

Love the spelling 

[awemawson]

Yes it is a bit hit and miss isn't it. I've noticed that on several of the Beaver documents

[russ57]

Strange - it's marked as Public 

Maybe it need time to cook 



later edit: I've just tried the YouTube video on the Wife's iPad (so that I'm not logged in) and it shows fine

I usually use tapatalk, no worries usually. But that video doesn't even appear as a link.

But it is there from the website...

Russ

[awemawson]

Although when I set the machine to cut a thread 'in the air' it appeared to be cutting correctly, now I've been brave enough to cut metal (and actually mount a threading tip, and measure it's offsets, it is obvious that all is NOT well. It looks to me as though the spindle rotation is not synchronised to the Z travel. Presumably the controller should wait for a reference point in the rotation before firing off the Z movement whereas the two look to be asynchronous to me.

I'd not be a bit surprised to eventually find that this is tied to the M19 spindle orientation problem that I'm having. Diagnostically it might just be a blessing in disguise, but then it might just be a pain in the neck 

This is supposed to be an M12 x 1.75 mm pitch thread !

[charadam]

Is it supposed to be a  LH thread?

[awemawson]

Nope !

On the simulation it is distinctly Right Handed 

If you notice, the Z travel must be accelerating cutting from the chuck outwards as revealed by the increasing 'thread' spacing

[charadam]

So there is a dichotomy beween simulation and the actualisation.

I suspect that beetroot juice has got in the gigglilator sprocket and is confolunding the fundamole.

[PK]

Just to convince any doubters, myself included, here is a little video showing an M03 starting the spindle !

Maybe a previous owner bought a whole bunch of upside down tooling by mistake and then swapped the wires to the spindle motor to make their code work....
 

[awemawson]

OK I finally think that I've gone MAD 

Someone please check my logic here:

A/ The program is cutting a normal right handed thread of 1.75 mm pitch, using an INVERTED tool located BEHIND the work and the spindle running
     CLOCKWISE viewed from BEHIND the chuck

B/ This is exactly akin to a normal manual lathe configuration if you imagine the tool swinging round the work 180 degrees to be located in the normal
     manual position in front of the work. Of course this movement has also cancelled the inversion of the tool.

C/ So as with a manual lathe, a right handed thread is cut TOWARDS the chuck, the tool withdrawing on the return cycle

So why, dear forum members, please tell me (to preserve my sanity!) is the tool engaged in the work on the travel AWAY from the chuck, and disengaged on the travel TOWARDS the chuck.

I closely monitored the threading progress, watching the controllers display of 'feed rate' - now when threading it is using 'G95 feed per rev' and the display shows 1.75R ie 1.75 mm per rev BUT it's showing this on the travel TOWARDS the chuck when the tool is DISENGAGED, and about 1000 mm per rev on the travel AWAY from the chuck when the tool is ENGAGED 

So basically the tool X movement is reversed in this situation - but travelling to the beginning of the threading cycle it gets to the right place so is not reversed.


. . . argh . . .I need a coffee 

[PK]

I've seen this..... oh bugger, where have I seen this?.... I vaguely remember a discussion about threading with rear toolposts in the days of turbocnc... 

If I'm right then it's a CAM thing....Is there an option in your CAM package for front or rear toolposts?

How does X work on your machine? not inverted?....

Maybe a different canned cycle or a parameter for rear posts in a machine controller able to cope with different machine configurations?

[cnc-it]

M3 spindle anti clockwise looking from the tailstock end..and the tip facing down to the floor..basically a manual lathe with the tool post rotated 180 degrees around the chuck..a throw back from when early cnc lathes which were modified gap bed lathes..the swarf could then be fired down into the trough so it was easier to leave the machine running without the operator needing to clear the swarf.

The Hitachi Seiki 4nE11 was this same configuration.

[cnc-it]

As the Beaver is running clockwise in M3 upward facing tools all the way.
 

[awemawson]

SOLVED 

The post processor is producing a positive depth of thread not a negative one. I have manually altered the produced program shoving a minus sign in and produced a decent thread.

I've not  worked out yet why the PP is doing this - it might think it's doing an internal thread - I'll investigate after lunch

[russ57]

I'm the furthest from an expert in cnc. But as someone suggested above, in absolute terms, the x direction would need to be reversed for a rear toolpost.
Either that, or you have the southern hemisphere model...



Sent from my SM-G965F using Tapatalk

[awemawson]

I'm the furthest from an expert in cnc. But as someone suggested above, in absolute terms, the x direction would need to be reversed for a rear toolpost.
Either that, or you have the southern hemisphere model...

No Russ I don't think that is the case - remember that your frame of reference has rotated. Minus X travel is into the work piece in each case. In all cases the work piece diameter decreases as the tool approaches the axis of rotation.

So the test piece passes the 'will a nut fit' test with flying colours, and also the repeatability test.

As for altering the post processor, when the cycle for threading is generated, the PP doesn't know if it is threading internally (so positive thread depth) or externally (so negative thread depth) and I can't logically derive which case is in operation from the parameters passed from the program. You can't even enter a negative depth of thread where you define the feature, as it slaps your wrist!

One comfort from all this is I'm more than ever convinced that the M19 spindle Orientation issue has to be a software / enabling bit, as this threading is using spindle orientation at a great rate of knots quite satisfactorily.

All this testing is producing masses of swarf - I need to get the covers back asap not only to contain it, but also to allow me to use coolant.

[mc]

All this testing is producing masses of swarf

Surely a smaller bit bar, or a bigger nut would reduce the masses... 

[awemawson]

I know this sounds silly, but the hydraulic chuck jaws that I have on only close down to about 20 mm so I am using 25 mm (well actually 1") stock for testing. I do have soft jaws to fit but don't want to waste them. The jaws fitted are hard jaws. 

[awemawson]

After much digging about in the inner workings of the FeatureCAM Post Processor I managed to prove that the program was outputting an 'absolute'  depth of thread rather than a 'signed' value as required by the Sinumerik 820T control. But I also realised that actually I could determine if when the threading cycle was entered it was an internal thread (requiring a positive depth), or an external thread (requiring a negative depth) by testing an internal variable called <OP-TYPE> that I discovered.

After finding that, it was a simple matter of multiplying the thread depth <THRD-DEPTH> by -1 for external threads, and the problem was solved. Dead easy in retrospect, but it took a while !

[cnc-it]

That's why i like Featurecam..editing posts is really easy..the whole cam package is built with machinists in mind 

[awemawson]

That's why i like Featurecam..editing posts is really easy..the whole cam package is built with machinists in mind 

Yes but sometimes the syntax gets overly complicated !