Plasma cutter out of sync with software

[BrianLee]

I'm trying to cut small letter out of 20 gauge metal and have found that the speed of the machine is not able to keep up with the software. Which is understandable, but it is creating a huge problem with my THC. My THC (it's a Neuron Simplicity) controls the Z-axis directly, so when the UCCNC software reads the line of code to switch on the THC, the machine is not always in the proper (X,Y) position and if the machine is actually at a corner then turning on the THC results in the torch plunging into the metal.
The pic below shows where my machine was on when THC was turned on. (The letter is about 2" high)

This pic shows where the software was when THC was turned back on. It was actually starting on the next letter.

My question is can the software be adjusted so that is more in sync with the machine position? Or is this a situation where because steppers provide no feedback position to the software servos are needed? Which then brings up my next question, can the UCCNC software communicate effectively with servos in order to keep things in sync?

[cncdrive]

The M205/M206 are motion syncron commands and are in syncron within 1milliseconds (1/1000 seconds) with the motion, so they are very well in syncron with the motion. That time is so short that it can't cause an issue like you showed.
Servos will not make things better, they can make it worse if they have a following error while steppers never have more following error than one step. (Except if they loose steps.)
I think the issue will be something else than the THC turning on/off, but yet I don't have an idea what it could be.
One thing: Are you sure you switching the THC on the proper places in your code? I would check that first.

[BrianLee]

I'm not questioning if the M205/M206 codes are in sync with the software, I'm sure they are. The issue is my cutter is not in the same (x,y) position as the software. In other words, since I am using steppers the software has no idea where the machine is at and in my case the machine is lagging behind the software by a second or two. When the software then sees the M205 command this command is sent to the Neuron controller which directly controls the z-axis. The Neuron controller doesn't know either that the machine is lagging behind and automatically initiates THC when in fact the machine isn't anywhere near the correct coordinates for the THC to be turned on.
This was the first time I noticed the issue. The machine was at a corner when the THC was turned on.

This was where the software was when the THC was turned on.

Here is a short video I made of what is happening.
https://youtu.be/5f5c7NMIOUY
You can clearly see the THC was turned on because i have a floating head on my machine and you can see it moving as the cutter is plunged into the material.

I rewrote the rules for Sheetcam so that the THC would take an initial reading at the start of the letter and then turn off. I thought this would at least cut the whole letter out and since it was a small letter the THC wasn't really needed. But the machine plunged at the new position as shown in the first picture of my first post and when I looked at the software and saw that it had at that point issued a THC ON command for the next letter.

[cncdrive]

What do you mean by "The machine is lagging behind the software"?
What is lagging compared to what?
If you mean that the machine's position is lagging to the M205 command position in code then I have to say that it is not possible that it is lagging. It can lag with a maximum of 1milliseconds which is a so short time that it can't cause an issue like this.
The M205/M206 works with the software putting those commands into the motion buffer between the motion commands and the motion controller is directly executing the motion commands and also the M205/M206 commands between the motion commands. And so is why that high lagging is not possible.

[shad]

Hi Brian!
I already told you in a email that this is looks like target voltage is lower then optimal for 1.5 mm cut height. When Automatic Voltage Control (AVC) enabled (after then AVC delay is ended) controller compare target voltage and real. If there is a difference, controller try to compensate error to zero. On the video controller move torch down, because Set Voltage < Real voltage.
First - check position of the divider selection jumper on the controller PCB. If you use 1:20 input the jumper must be set on 2-3 pins.
Then check how the UCCNC THCOFF signal (THC en. pin output must be connected to the THCOFF input of the interface board) work.
If THC enabled in the UCCNC - led on the controller face (which is marked as an up arrow) must be OFF. if THC disabled - led must be ON. In this case THC ON/OFF command from gcode will be transferred to the controller correctly.
Then try to cut straight line without AVC (turn off AVC using center button on the controller) and controller will show the actual value of the arc voltage during the cut. This is a real arc voltage value for cut height. Set this value as target, turn ON AVC and cut.

[BrianLee]

From my observations it appeared that the THC was being turned on too soon--before my machine was in the correct (x,y) position. I assume the software can only calculate the (x,y) position on my machine but would never know the absolute position. That's what I meant by lagging. I thought the THCON signal to the Neuron controller was being sent too soon.
However, after running more tests today I now know that what I had previous thought is not actually the problem. It appears the problem may be with the THC controller. I'll have to do more checking tomorrow and will just email Andrew instead of continuing this thread since it now appears not to be an UCCNC issue.
Thank you for your time and help.

[Robertspark]

Are you using servo or stepper motors?

Can you post a code snippet?

Can we have a look at your profile?

[BrianLee]

It turned out to be a user id10t error. I didn't populate the THC en. out pin and port settings on the I/O page. In my defense I thought those setting were used when THC was being controlled by the UCCNC software. Added the needed pin and port setting and everything is working pretty good. I still need to make a few adjustment here and there to zero everything in but it's performing pretty good now.

[cncdrive]

Good to hear it works good now.