G76 picking up the thread after you've changed the speed.

[Pauly123321]

Is there any way you can pick up the thread after changing the speed using 2 line G76 fanuc controls.

 

[yardbird]

Hope you can see this?

Brent

 

[Pauly123321]

Unfortunately not yardbird.

 

[yardbird]

Sorry give me a little while and I'll type it out when I get a chance. Supposedly a formula on how much to shift the Z start point due to a given RPM change.

Bernt

 

[yardbird]

RPM Change During Threading

If the RPM must be changed before the completion of the threading cycle you must:

1) Divide the amount of increase/decrease by the number of threads.

2) Modify the "Z" offset as follows:

A) RPM increase; Offset "Z" towards the thread start point.

B) RPM decrease; Offset "Z" away from the thread start point.

EXAMPLE; 3" - 8 Pitch ACME thread at 300rpm increase to 350rpm.

50 ÷ 8 = 6.25 or .0062" towards start point.

In conversation at another site about the same topic three different people posted three different pictures of this procedure from three separate manuals. I'm almost certain we were talking about Fanuc controls but can't be positive. I was going to test this when I had time but haven't yet. Would be easy enough to do. Maybe someone else can add more?

Brent

 

[Pauly123321]

Thank you so much yardbird. (Brent) as soon as I get the chance to test this theory I will.

 

[BSCustoms]

You can also change the speed in any multiple of the RPM without issue. So 250-500-750-1000

 

[sinha]

If the formula works, it is fine. If not, a small experiment can be done to measure the shift in the helix when the rpm is changed:
Run G76 with a very small first DOC (Q in the second block), say, Q10 in mm mode. Abort the cycle after the first pass. A thin helix would be made. Put some color in it to differentiate it from the helix to be made next.
Run the same program with a different rpm, keeping the start Z same. A shifted helix would be made.
Measure the gap accurately and observe whether it lies to the left or to the right of the first helix. Start Z of the second G76 can be shifted by this amount in the opposite direction to make the two helices coincide.

I have not actually tested this method, but logically it should work. I will appreciate any feedback to validate or reject this method.

 

[NAST555]

Measure the gap accurately and observe whether it lies to the left or to the right of the first helix. Start Z of the second G76 can be shifted by this amount in the opposite direction to make the two helices coincide.

I have not actually tested this method, but logically it should work. I will appreciate any feedback to validate or reject this method.

It works, It is similar to how I line up threads to repair them without a repair function. But more with eye balling it.

 

[Pauly123321]

Sinha thanks for your input I shall give this a go when I get a opportunity and I will most definitely give you my outcome. Logically this should work especially if we use the digital read out on the screen to work out how far apart they are

 

[Pauly123321]

Nast555 thank you

 

[Pauly123321]

BSCustom. Have you tried This? Is this on fanuc?

 

[litlerob1]

I would really like to hear how well either of those things work. I already know Sinha's way works, but I've never heard of Formulas working, or Multiples of RPM (it's not a lead screw).

R

 

[Pauly123321]

Sinha I tried the slowing if speed and the working out where the next cut was and that worked out perfect the formula on the other hand was a fail so was changing the speed by 500 to 550 to 600. All picked up in random places.

 

[litlerob1]

Sinha I tried the slowing if speed and the working out where the next cut was and that worked out perfect the formula on the other hand was a fail so was changing the speed by 500 to 550 to 600. All picked up in random places.

Too bad, it's what I figured. But I was hoping there was something to it. I'm sure there is a formula for calculating the theoretical start point. Problem is the Ladder logic varies machine to machine.

R

 

[Bobw]

Any formula would have to take into account the acceleration of the Z axis, since that is where
the error comes in.

 

[Pauly123321]

Yeah I'm a little gutted. Would have loved to know the formula of their is one.

 

[Pauly123321]

I do enjoy reading people different ways and methods as well. Some guy called Bill on here taught me 2/3/4 start threads. Some people have great ways of explaining thing.

 

[gregormarwick]

Any formula would have to take into account the acceleration of the Z axis, since that is where
the error comes in.

Yes, but that is not the only thing.

The response time of the control also has a hand in it. The time period between the spindle encoder hitting the marker pulse and the control telling the axis to go. As the spindle speed changes the angular distance travelled between marker pulse and axis moving changes.

 

[Pauly123321]

Where is Einstein when you need him . He could break the formula for us