[ashes-man]

Is the Z-50. to the top of the part

You can't use a ( 0 ) for the ( I ) value

I - Distance along X Axis to center of circle

K - Distance along Z Axis to center of circle

I haven't tried without the I0. in the arc. Doesn't not specifying the I value risk using an unknown I value from a previous command, or are I J K not modal? I could easily try without the I0. to see what happens. I have seen other examples that have I0. The fusion generated code had an I0 too. However, I will go give it a try...

[ashes-man]

...Does your control have F91 Parameters?

No, the M32 manual I have keeps referring to bits in F92 parameters. My machine has about six screens of user parameters but are all I000, I001, I002 K..., J,..., H... etc. I have worked out what a few of them are as they are used in the tool change macro, some others are obvious by their value. But no idea which ones control axis behavior. I am working on trying to get a manual, but have to register the machine and declare that I am not making weapons with it etc before I can get any support or parts from Mazak...

[ashes-man]

So, tonight I had another go and tried each of the following things one at a time, but none of them helped at all in the arc center being correct:

- Put a G1 Z-100. after the G03 arc move
- Tried G03 instead of G3
- Put a G92 X-100. Y-100. Z-100. command at the start of the program (I dont really understand G92 yet, but it had no effect)
- Removed the I0. from the G3 command
- Changed G43 Z10. H1 to G43 X0. Z10. H1
- Moved the arc start and end point 10 units in X so wasn't on the axis

I then tried G43 H1 (instead of G43 Z10. H1. It caused a drop in Z position exactly by the tool offset amount (-211.404). Whereas G43 Z10.H1. only drops Z by the amount needed to make the tool tip 10mm above the origin. I don't really understand why, I will have to study that one some more.

One of the manuals I have for the M2 controller (which is still not the same as mine) has the following. The examples in the M32 manual also often show a G92 X0. Y0. at the start of their programs. Is this relevant to my arc problem?

2-3-12. · Coordinate system setting (G92)
Before transfering the tool using absolute instruction, specify a
coordinate system using instruction:
G92 X ~ Y z Z z A a
Accordingly such a coordinate system is established that the tip
of the tool installed on the spindle locate at (x, y, z, a).
It is called work coordinate system.
Note 1)
Note 2)
If G92 is used in offset, such coordinate system is
established that a position designated in accordance with
G92 is taken up before hand.

[machinehop5]

I then tried G43 H1 (instead of G43 Z10. H1. It caused a drop in Z position exactly by the tool offset amount (-211.404). Whereas G43 Z10.H1. only drops Z by the amount needed to make the tool tip 10mm above the origin. I don't really understand why, I will have to study that one some more.

...delete the G43 H1 completely in the program and see what happens.

the G92 presetting is like G54 presetting...its a Tool Comp problem we are looking for IMO

[mactec54]

I haven't tried without the I0. in the arc. Doesn't not specifying the I value risk using an unknown I value from a previous command, or are I J K not modal? I could easily try without the I0. to see what happens. I have seen other examples that have I0. The fusion generated code had an I0 too. However, I will go give it a try...

Try a Z0. and a F100. value at the end of that line

The I0 value depends where the X axis is when you start the arc so if the X axis moves say 20 the the ( I )start point would be 20 if the X axes is zero when you start the arc then the ( I ) would be 0

Stick with the basics G54 a G92 can be another problem that you don't need to deal with

Your program is doing what you are telling it to do the Z axes is at 10 above the part, you would have to have a Z0. for it to be on the top of the part, and if you want to go below that then you would need to have a Z-.5. this would move the Z axes below the top of the part by that amount

[Heavy_Metal]

It seems that you don't need the G43 H_ on a Mazak, it's a parameter setting when you want to program like Fanuc/Haas ( G43 Z___ H__ ).
Are there no programs stored in the LIB that you can use as an example that shows G43 H is not necessary.
I think there is a parameter active that conflicts with the use of G43.

http://www.cnctrainingcentre.com/tag/mazak/
They don't tell which parameter but maybe it depends on the Mazak model.
.... scroll down on the page.

[ashes-man]

It seems that you don't need the G43 H_ on a Mazak, it's a parameter setting when you want to program like Fanuc/Haas ( G43 Z___ H__ ).
Are there no programs stored in the LIB that you can use as an example that shows G43 H is not necessary.
I think there is a parameter active that conflicts with the use of G43.

http://www.cnctrainingcentre.com/tag/mazak/
They don't tell which parameter but maybe it depends on the Mazak model.
.... scroll down on the page.

Thanks for digging that info out. The older manuals I have say that you should not use G43 if you are using a Mazatrol based machine. Apparently Mazatrol takes care of the tool length offset for you and you don't need G43. My controller does not support Mazatrol as far as I know.

I did some more experiments this morning. I tried making sure tool 1 was loaded as I was using H1 with tool 0, but as I suspected that did not help. The machine does not associate tool numbers with offsets, there is just a table of offsets, up to you to choose which offset should be used.. I also tried swapping to a positive tool length offset that represents the actual tool length from the spindle as per the manual (rather than negative tool length method). I corrected the work offset to suit. Interestingly, this did change the arc shape. It was still not correct but definitely changed it. I will plot out the differences and also try without G43 at all tonight and maybe I can come up with some useful information.

[machinehop5]

...Mazaktrol has the power to make so, cant buy a manual...due to possible weapons being made....?

sad country you live in

[mactec54]

It seems that you don't need the G43 H_ on a Mazak, it's a parameter setting when you want to program like Fanuc/Haas ( G43 Z___ H__ ).
Are there no programs stored in the LIB that you can use as an example that shows G43 H is not necessary.
I think there is a parameter active that conflicts with the use of G43.

http://www.cnctrainingcentre.com/tag/mazak/
They don't tell which parameter but maybe it depends on the Mazak model.
.... scroll down on the page.

Yes it will depend on what control he has as they can do both

Depending on his control vintage is to how the control is setup if it is using Mazatrol or Fanuc they can do both, depending on the vintage, it seem that he is not having a problem with that part of it though

[ashes-man]

I have my registration sorted now so can buy spare parts. The local dealer has been really good and helped me out with that.

In case anyone has any patience left for this problem, I have prepared another example, this one to hopefully get Mazak support with. It shows a complete example of the code, what the machine thinks it should do and what it actually does. Another video of it doing a big bang. This one made me jump while filming it, even though I knew it was coming!

I am pretty sure I have confirmed now that with all the combinations I have tried, the arc Z center is always at work offset Z machine coordinate - radius. Like it tool length offset is not applied to the radius center.

See attached pdf.

https://youtu.be/ZPdCrzxeH8k

[ashes-man]

So, I stumbled upon this service manual that seems to match my machine. Woo hoo! https://www.cnczone.com/forums/machi...t-imac-tp.html

It has all the parameters and how to get to the screens to set them. However, I cannot find any parameters that would relate to tool offset or Z axis behavior.

[mcode]

It seems to me the arc centers are absolute. Try specifying I,J,K's in relation to the work offset.

[mactec54]

It seems to me the arc centers are absolute. Try specifying I,J,K's in relation to the work offset.

With a G18 there is no J used only I, and K or an R can be used

I'm no sure but I think he may have tried having the offset in the ( I ) which I asked him to try and by the looks of this video he needs to try it again

G3 X50. Z-50. I50. K-50.

[mactec54]

I have my registration sorted now so can buy spare parts. The local dealer has been really good and helped me out with that.

In case anyone has any patience left for this problem, I have prepared another example, this one to hopefully get Mazak support with. It shows a complete example of the code, what the machine thinks it should do and what it actually does. Another video of it doing a big bang. This one made me jump while filming it, even though I knew it was coming!

I am pretty sure I have confirmed now that with all the combinations I have tried, the arc Z center is always at work offset Z machine coordinate - radius. Like it tool length offset is not applied to the radius center.

See attached pdf.

https://youtu.be/ZPdCrzxeH8k

G3 X50. Z-50. I50. K-50. try this ( I )=50.

[mcode]

Sorry I missed the x-y plane test. The numbers for the Z positions look correct, but it looks like if you use K-261.404 that would work which would seem that in G18 mode the arc center K is added to the length offset -(211.404+50). Must be a Parameter.

Edit: I'm referring to the .PDF example

[ashes-man]

Sorry I missed the x-y plane test. The numbers for the Z positions look correct, but it looks like if you use K-261.404 that would work which would seem that in G18 mode the arc center K is added to the length offset -(211.404+50). Must be a Parameter.

Edit: I'm referring to the .PDF example

Thanks for having a look at my pdf. I have actually tried doing K-261.404 and 211.404 - 50 and combinations. It seems that any K value (or I value) that doesn't make a legal arc as you would expect under normal conditions is rejected and an error is generated. It seems to be checking the arc in the correct incremental I K coordinates, then calculating the tool path with the wrong arc center!

[machinehop5]

...I notice you are still trying to use H codes....try no H codes and using T code at the beginning of operation to see if tool offset goes active automatically

(Do an arc in XZ)
T01M06
G90 G17 G49
G53 Z0.
G54
G0 X0. Y0.
G43 Z10.
G1 Z0. F175
G18
G3 X50. Z-50. I0. K-50.
M30

Work Offset (G54)
X = -260.897
Y = -53.354
Z = 114.994
Tool Offset
#1 = -211.404

[ashes-man]

...I notice you are still trying to use H codes....try no H codes and using T code at the beginning of operation to see if tool offset goes active automatically

I am pretty sure I have tried that. I definitely tried using T1M6 as well as H1. I have tried so many combinations I really cant remember. I will give it a go tomorrow and see what happens. From my understanding this machine does not have any relationship between tool selection and tool offset but I will try as only takes a few minutes.

[ashes-man]

Thanks for having a look at my pdf. I have actually tried doing K-261.404 and 211.404 - 50 and combinations. It seems that any K value (or I value) that doesn't make a legal arc as you would expect under normal conditions is rejected and an error is generated. It seems to be checking the arc in the correct incremental I K coordinates, then calculating the tool path with the wrong arc center!

I tried again tonight, and the only values it is satisfied with are those that make a valid arc eg I0 K-50, I-10 K-60, I-50 K-100 etc. Setting K to any combination of work ofset plus or minus tool length and/or radius did not work.

[machinehop5]

I am pretty sure I have tried that. I definitely tried using T1M6 as well as H1. I have tried so many combinations I really cant remember. I will give it a go tomorrow and see what happens. From my understanding this machine does not have any relationship between tool selection and tool offset but I will try as only takes a few minutes.

...I found an example of someones Hobbing program from cncmanuals. If you have a B axis you can make gears with it....if, not it's example of Mazatol format