XHC Pendent

[Robertspark]

Don't know about that... Some is just logic and math

[ger21]

It's all logic and math.

[Greolt]

I am just a hobby user.

G93 for me is certainly not complicated and have found it to work veery well for what I do.

G93 is easy to produce and has the advantage of running on most hobby level control software.

Think Mach3, LinuxCNC among others.

I am hopeful that UCCNC will not be far behind.

[Greolt]

Unfortunately I cannot talk about the CAM software that I use as yet. Soon I hope.

What I did before I had access to that CAM was use the converter that Shawn Gano wrote.

https://www.ganotechnologies.com/cnc/rapidrotary/

This converts G94 code to G93. The underlying math is not that complicated.

As I said it works well for what I do. Maybe there are instances where G93 falls down. But I have not discovered them.

[Greolt]

Suggest you use the converter on some rotary axis code of your own.

Shawn's program is very small and does not require installation.

However because you asked...

Z origin at surface of 40mm dia cylinder

Code: Select all

( Diameter = 40.000 mm  )
G0 G21 G90 G40 G49 G80 G94
S12000 M03
X0.000 A0.000 F60.0
G0 X-20.000 A57.296 Z5.000
G1 X-20.000 A57.296 Z-1.000 F60.0
G1 X-20.000 A-57.296 Z-1.000
G1 X20.000 A-57.296 Z-1.000
G1 X20.000 A57.296 Z-1.000
G1 X-20.000 A57.296 Z-1.000
G0 X-20.000 A57.296 Z5.000
G0 Z20.000
G0 X0.000 A0.000
M05
M30
%

Code: Select all

( Diameter = 40.000 mm  )
G0 G21 G90 G40 G49 G80 G93 (Inverse Time Mode, Converted from G94)
S12000 M03
X0.000 A0.000
G0 X-20.000 A57.296 Z5.000
G1 X-20.000 A57.296 Z-1.000 F10.0000
G1 X-20.000 A-57.296 Z-1.000 F1.5789
G1 X20.000 A-57.296 Z-1.000 F1.5000
G1 X20.000 A57.296 Z-1.000 F1.5789
G1 X-20.000 A57.296 Z-1.000 F1.5000
G0 X-20.000 A57.296 Z5.000
G0 Z20.000
G0 X0.000 A0.000
M05
G94
M30
%

[ger21]

I would think that UCCNC would respect the 4 decimal feedrate, even though it doesn't display it.
But is it really that critical, even if it didn't? I guess worst case would be 0.6 seconds per move, if it was actually rounded off.

[Greolt]

I remember Art saying that Mach3 used ten decimal places for all the calculations in the background.

Only the DRO displays were rounded for visual convenience.

I assume UCCNC is similar.

And I would argue that the math is fairly simple when we use computers to crunch the numbers.

Maybe I am just one of the lucky few that use CAM to do that for me.

[cncdrive]

The floating point numbers are represented as 64bits double precision type.
https://en.wikipedia.org/wiki/IEEE_754
https://docs.microsoft.com/en-us/dotnet ... rds/double

For the feedrate value only the visualisation contains 2 decimal digits, but ofcourse the number calculated with is still handled as a 64bits double type, so there is no digits loss for the calculations, only for displaying...

[Greolt]

For the feedrate value only the visualisation contains 2 decimal digits, but of course the number calculated with is still handled as a 64bits double type, so there is no digits loss for the calculations, only for displaying...

That is what I expected.

Thanks

[Greolt]

Interesting do you hand program or use Cam ??

The updated version of Vectric software (Aspire and Vcarve Pro) have now been released.

They will now produce G93 code for rotary axis work.

They also have much better integration of rotary axis CAM capabilities.

I believe that with their very large user base, this will encourage even more users to look at doing this sort of work.

Would be great to see UCCNC support G93. It would be a shame to have the lack of this feature deter new users taking up UCCNC.