Using Mach3 with UC300eth

[cncdrive]

No motion controllers work with Mach3 R2.xx, because those versions not supporting motion controllers at all, they only work with printer ports.

[NikolayUA24]

Got it, Thanks for the answer.

[NikolayUA24]

I have been eyeing your UC300eth controller for a long time. Installed UCCNC in demo mode.
3d pack on 2 million lines. What I see immediately - when loading, the 4-core processor loaded on the full and partially video card for a minute and a half, and took up more gig operational. For comparison, the mach3 processor almost does not notice, and in the operational area around 100mb - that is, a dozen times less. no matter how scary, but it says that with the optimization of the problem in UCCNC.
Further, with absolutely identical settings, mach3 launched this pack in parallel, and after an hour looked what was happening there - mach3 percent is 25 percent ahead of the lines.
Again, maybe it’s just that the speed is considered differently, I would not want to have a controller that works with brakes and jerking on heavy units.
According to the simulation, it’s not really clear how it really works ... especially with heavy code, arcs, etc. ... If at least at the level of mach3, then it’s definitely a worthy controller for its price. It’s interesting to see real users in real work. What do UC300eth users say?

[cncdrive]

Well, you can see real users in the forum, in the what I've made with my machine ... area.
E.g.: viewtopic.php?f=5&t=2046
7 millions lines of code ran for 3.5days long.

And it does not really matters for the UCCNC how long your code is, it will not load the CPU more if the code is longer, however ofcourse your videocard will be loaded more with the larger toolpath to display.
If your video card is weak then it does not matter how strong your CPU is, it will overload your CPU, because then the video card cannot do the job and then the CPU has to do it.
For computation other than displaying the UCCNC requires very small computing power.

UCCNC has path deviation control similar to how industrial controllers work, Mach3 does not.
With mach3 you can only guess about how much error it will make on the paths, you cannot set the workpiece tolerances.
Yes, mach3 is fast, but it is only fast, because:

1.) it has no path deviation control, so you cannot define the tolerances for the jobs. It will round corners and remove details from your work as it wants to.
With UCCNC you can define the tolerances exactly, for example you can tell the controller to optimise the path with a maximum path error of 0.05mm and then it will only optimise to that level.

2.) it's trajectory planner is broken, it overaccelerates/deccelerates your settings causing jerks. 2x overacceleration is very usual while we even measured out 8x overaccelerations with some extreme axis settings.
For this reason ofcourse mach3 is fast, but in the same time you have to realise that you have to downtune the acceleration parameter in real world compared to what you can set in UCCNC, otherwise your motors will loose steps.
In other words in the UCCNC you can set your axes accelerations at least 2 times than what you can setup in mach3 to make it work the same safe!
And this means that the UCCNC will be faster with the higher accelerations and it will still keep more workpiece details with lower tolerances than mach3.

[ger21]

What do UC300eth users say?

UCCNC is far superior to Mach3.

[Tweakie.CNC]

Hi NikolayUA24,

I am a long time Mach3 user and I have recently started trials with UCCNC and the UC400ETH controller (I am a newbie here).

With Mach3 (and Mach4 using the ESS motion controller) my maximum (100% reliable) X and Y axis rapids (G00) were 4,500mm/min. With UCCNC that is now 8,000mm/min with a corresponding doubling of Acceleration. Certainly for laser work and rotary engraving work UCCNC has noticeably reduced my job runtimes.

Tweakie.

[NikolayUA24]

Sounds tempting.
I tried working on controllers.NCStudio,XC609MD-4axis,PLCM-E1b,and I returned to mach3, because I was not happy with their work.

[cncdrive]

That is because Mach3 has a "cheat bug" with the broken acceleration/decceleration profiler which makes users thinks it is fast, which is true, but in comparion you can make any controllers faster than mach3 if you higher the acceleration with what mach3 would fail...just users do not know that and so they doing exactly what you did, comparing with same acceleration values...