Arc slowdown problem

[Travis]

Hi all, I am fully aware there has been discussions for years about this topic. I also understand that problems don't get solved overnight. But the constant velocity settings in uccnc are still a huge issue for plasma machines. I can say that chances are most people have not been voicing their concerns, or they simply do not have enough experience or knowledge to know better. But I've been through many forums through out 2018 discussing a new trajectory planner for dealing with arc segment slow downs. It's now 2023, which is 5 years later, and I'm hoping there has been some improvement or the new planner will exsist sooner than later. For cnc mills, and routers, not a huge issue, because these items can be have nicely mapped finishing passes. But for plasma these arc slowdowns create more dross, which also creates more bevel for these quick slow down points, on top of leaving torch indents at the slow down points. It is not practical to have this feature be left unresolved for uccnc plasma, as you've made a dedicated launcher for your plasma interface. You have to spend a countless amount of time tweeking corner error, arc radius tolerance, and linear error tolerance as well as adjusting acceleration in combination with these other features to try to create this balancing act to mitigate arc slowdown. And in the end as much as uccnc might be much more accurate than in path trajectory than other software, it comes at the price of not being able to perform properly enough to run a highly accurate plasma machine without comprise.

[cncdrive]

Well, I disagree with what you wrote, if you setting the acceleration of the axes high enough and the tolerance low enough then the UCCNC outperforms Mach3 in execution speed doing less slowdowns.
Please remember that in UCCNC you can set the axes accelerations higher than in Mach3 (if you came from the Mach3 world) because Mach3 has a bug in the trajectory planer.
Also remember that you have to set the tolerances properly for your work, I mean if you setting the tolerances to micron precision with low accelerations then ofcourse slowdowns will be visible, because then it is mathematically impossible to do it elseway.

[Travis]

So that's a pretty poor opening statement, just because you disagree doesn't mean it's not true. What I said above is a statement of fact. I have two machines running side by side, with the same components one running on mach3 and one running on uccnc. Regardless of what liberties mach3 is taking for its path control, it does not die out through corners and the corners are still plenty accurate. You can not say the same for uccnc, as soon as you set the acceleration high enough to mitigate most of the slow down, it's too much acceleration for the machine to handle. Then you go back and adjust your tolerance and lower your acceleration down to the capability of the machine, and sure you get a somewhat smooth motion but not your out of accuracy by well over .05". Okay so now back to bringing the tolerance into an acceptable range, oh wait now there is more arc slow again, which again when the torch slows down it melts metal more in that spot. So in turn it leaves a whoop in the part, from being in one place longer than it should have been. It's obvious that you don't actually run a plasma table because if you did you would understand why this is so detrimental to quality functionality and part production. The truth is I was behind uccnc for a long time now and all my machine run your software. This is the first plasma table I've put together and used your motion controller and software with. It's performance plain and simple in corners in not acceptable.

You can say all you want that there is plenty of adjustment to take away the arc slowdown, but it's not true. I don't care what type of machine you have every machine has an acceleration limit. You can't expect to solve a someone's problem by telling them to set thier acceleration higher. My machine has 1000ipm Rapids, my acceleration is set to 20. If I go higher than 20 the acceleration and deceleration is just way to abrupt. I can easily find a nice take off and stopping speed below 20. The next thing, I have my corner error all the way up to .1 for error allowance.......IT STILL, slows down through corners. Okay well then you'll mention my capability as a designer because my vectors must be junk. Well they are not I have 13 years in vcarve, 10 years in rhino3d, 5 years in fusion 360. I'm very well versed in creating perfect vectors for machining and manufacturing.

I'm also annoyed that the answer you provided is the same answer I see everywhere and saw back when this was a popular topic. Believe it or not some of us are very well versed in what to expect from normal tolerances. Obviously I'm not trying to run my tolerance factor at .001" because it's pointless to do so for plasma, but should I expect a .03" tolerance, absolutely, that is 100% practical. And every single part that comes off my machine with the mach3 controller is a consistent repeatable .03" tolerance.

Now with the tolerance set to .03" in uccnc the arc slowdown is out of control slow, I can't cut any true circles, they all end up with a whoop in them, and it isn't because I'm not leading in and out correctly because I have very fluid lead in and lead out parameters. Yet on my machine running uccnc I end up with a junky hole because it slows down on the lead in and lead out leaving a spot where the torch is allowed to burn that area longer, which leads to crap looking hole. This happends on every single arc, and is unacceptable.

I suggest you work on a solution for this because I know you won't refund me for my motion controller and the software for this machine. And again as much as I really like using uccnc for my mill and routers, the planner is not working for plasma at all period. Again this is not an opinion, opinions are things that you have when you have a preference towards something. This is a proven fact that it does not work for plasma.

[Travis]

Well I've waited a while for a proper response. So far it seems no one wants to discuss this further. What is sad about this, is I've been reading through many areas of the forum, and this seems to be a trending deal. One guy basically pleading with you about cv settings because it's providing poor surface finish results. As well as trying to get help going back to mach3. Well I am having literally the same exact problems, identical to what he described, using uc300eth, ucbb1, clean installs. Can't get homing to work on the y axis with the a axis slaved. Works fine in uccnc actually everything on the plasma works perfect, all thenrelays, touch sensors, homing, except this cv tolerance. I've tried 1000's of combinations at this point. Put in what should be an unnecessary amount of time in file editing to try to node edit to "trick" uccnc into moving through corners faster.

I want to say even with all this happening, I still just purchased a new uc300eth, and ucbb1 set from cnc drive. The reason, is because that setup is for a router, and in terms of that it has always been 1000% reliable.

The reason I said the above is to specify I've made the decision to go with uccnc 3 times now, and I've spent 6 months creating my own custom screenset from scratch a few years ago, and I enjoy the interface so much I don't want to be forced to change.

I'm not some random guy being a bigot here, I'm a guy who runs a small business, much smaller than uccnc, trying to survive. The problem is in my eyes, is you need to start coming up with better excuses than just saying something bad about mach3, or referring everyone to the manual. We have all read through the constant velocity setting a million times by now, and reading through it isn't going to make it any better. It just flat out needs to be fixed, and I'm sure that sounds like a huge undertaking. I'm sure it is because I don't code, although I recently just signed up for c++ courses and I'm about 3 months in. For guys like the team at uccnc, if you wrote the program from.scratch, you ought to be plenty smart enough to think through a way to fix it and or adjust it for plasma use. Even mach3 has a small coding deviation for increasing acceleration during plasma use specifically. I've seen it done with cheap hardware and cheap motors, on a lack luster machine platform. Uccnc ought to be running circles around mach3 in this aspect.

Please create a thread and updates regarding the cv setting within the trajectory planner. So you can become the best control software period. And for the sake of all your loyal customers who continue to buy even knowing this large pitfall.

[cncdrive]

Hi,

Which version of UCCNC are you running.
The arcs were optimised several times in several versions of the UCCNC.
Did you try 1.2115? Or ?