[DrNos]

Yes, I can see the image you attached just fine.

What did you use to generate it?

[deadlykitten]

Yes, I can see the image you attached just fine.

What did you use to generate it?

1 method ; there are 3 more

[deadlykitten]

... feed in the Z-direction is constant but the feed in the X-direction changes. - While F changes with each iteration of the loop, the contribution of F to the actual feedrate in the Z-direction is constant. If this wasn't the case, the graph would not look like a pure sine wave

hy doc, you are right, something like that happens ...

each movement has projection among X and Y, so there is dx and dy ( x = Z axis, y = X axis)

check out attached graph for " dx / dy * cos " : it is always ( almost ) 1, except for a zone near 90* and 270*

i have no clue what this means in reality, i can dig it futher, but still, i see no point in using F = 0.1 + ABS [ COS [ ... ] ]

why should "the contribution of F to the actual feedrate in the Z-direction be constant" ? why ? what's the catch ? what's the run dude ? ( azaelia banks - 212 )


If this wasn't the case, the graph would not look like a pure sine wave : sorry, but this is not true ... you don't obtain the sine because of the feed value; feed only tells the machine how it should walk among the sin wave, thus the feed has no conection with the toolpath geometry, but only with the dynamics


well, cnc movement changes with feed values, but this is another story when feed values are low, the cnc movement will be pretty close to the programmed movement; deviation increases when feed increases ... again, this is another thing / kindly

[deadlykitten]

hey doc, you still arround ? i just wanna let you know that i am working on this thread

[kurmay]

Hi DrNos

Okuma LookAhead function isn't working good in the iteration. But i would recommend writing the program at least as much as possible.
try this codes in same line
DIA1=DIA1+10 LEN1=LEN1+0.5
it can work faster.

[deadlykitten]

or maybe NOEX DIA1 = DIA1 + 10 LEN1 = LEN1 + 0.5

[kurmay]

or generating code for fastest speed

this sample is excel file

[DrNos]

why should "the contribution of F to the actual feedrate in the Z-direction be constant" ? why ? what's the catch ? what's the run dude ? ( azaelia banks - 212 )

I don't know how to explain it better... The X and Z components get separated into two parts with respect to F. It works and behaves like I want; it's just too slow.

Hi DrNos

Okuma LookAhead function isn't working good in the iteration. But i would recommend writing the program at least as much as possible.
try this codes in same line
DIA1=DIA1+10 LEN1=LEN1+0.5
it can work faster.

I had a similar idea! I should be able to test it this Friday.

or maybe NOEX DIA1 = DIA1 + 10 LEN1 = LEN1 + 0.5

What does NOEX do?

or generating code for fastest speed

this sample is excel file

I was thinking about something in-between the two strategies you suggested. - Run the loop to populate the registers with 127 values and then have an 'unrolled' loop full of G1s. The loop could repeat every rotation, or every other rotation of the spindle, provided I can convince the machine to give up its threading-cycle secret sauce.

[deadlykitten]

hey doc, i thought you are lost with a patient

look man, what i really do here is delay my answer, pls consider that i have allready delivered such custom threads, and the technique works "as it should"

there is a tiny conflict : " as it should " <> " It works and behaves like I want "

computing those X&Z components within respect to F is not a fundament for machining

pls consider these examples; all of them require same time to complete ( 1 minute ) :

    G01 X100                                            F50 G94 G91
    G01                           Z50                   F50 G94 G91
    G01 X + 2 * SQRT [ 2500 / 2 ] Z + SQRT [ 2500 / 2 ] F50 G94 G91
    G01 X + V1                    Z + V2                F50 G94 G91 ( V1 / 2 ) ^ 2 + V2 ^ 2 = 50 ^ 2

somehow, one way or another, we have to fix this variable F that you use; after that, is possible to move on without any doubts about it


it's just too slow : because you succumb the cnc; there is a rule for minimal segmentation and coresponding F; you can not simply input whatever segmentation and whatever feed, and expect the cnc to perform no ... it does not work like that


NOEX is not required to compute code for custom threads; it may boost your code, but it won't work, because specialized threads are not loop-generated; more precise, the loop should not content only cutting movements; the only way to deliver a specialized thread, by using a loop, is to use at least 1 positioning movement within that loop; the loop creates downtime, and this downtime will affect how the cnc behaves :
... if there are only cutting movements, you will put a break on their execution, and surface rugosity will increas, and tool wear/friction may appear
... if there is at least on positionong movement, than all those negative effects should hit into the positioning movement, not into the cutting movement


about that video that you shared at post 3, and you try to reproduce : you won't be able, because that is a specialized machine : the way it works goes beyond basics, and few people have seen such a lathe in reality ... or maybe you have one ?




ps : you may find a code attached ( it works similar to the videos from post 5 ); truth is that it has "tricks" : i can explain them, but i strongly suggest small steps ... you have rushed small steps Doc, small steps : about that variable feed, i will be back with an example, and i hope that it will make some light

[deadlykitten]

hey doc, me again feed is always tangent to the cutting toolpath

so the controller calculates automatically the feed ratio for each motor / did i pushed the right button ? kindly