QTPlasmaC issues / questions

Hello all! 

First of all thank you for the great effort and support for DIY people like me. I managed to put together a neat little plasma cutting machine using the forum and youtube-videos as guide. I got it working yesterday with all systems running as they should.

I came upon few issues while building the machine and these two things still puzzle me. I am a new user as you probably figured out so these might be very easy to answer  but I would be very happy if you could help me on these.

1. During PnConf-wizard there is a tab where Mesa THCAD settings are selected for THC. Where do these values go / where they are stored in the machine configuration files? I can't find the set values either from xxx.hal or xxx.ini files where other stuff seems to be written. The THC works nicely though.
   
   
2. I have Hypertherm Powermax 45xp as plasma unit and it has a feature where changing  consumables or disabling - enabling torch from the safety switch leads to a situation where plasma does not fire from next start signal. It instead sends out series of short bursts of air indicating that the next start command will fire the torch. This all is fine but the LinuxCNC seems to get totally stuck if this happens at the start of the cutting. The machine will start running the program and goes on to probe the Z height and then just waits there at pierce height indefinitely. I have not found out how to recover from this. Starting program again does not help, nor does switching motor power on and off. Only way to recover from this is to restart the linuxCNC. Have I missed something?
   
   Here is the paragraph from hypertherm manual:  
   The first time you pull the torch trigger after you set the torch-disable switch to the “ready to fire” ( ) position, the torch may emit multiple puffs of air in quick succession. With each puff of air, the torch vibrates slightly, and plasma power supply emits a pressure-release sound. This is a warning that occurs when you lock and then unlock the torch without turning OFF the power supply. (The Torch Cap LED also illuminates – see page 148.) This does not indicate a fault condition. The purpose of the warning is to alert you that the torch is activated and will fire a plasma arc the next time you pull the trigger. See page 45
   
   
3. Is there a way to stop all movement on the CNC except pushing the estop-button? I mean if I mistakenly click homing button and I know something is wrong I would like to stop the movement there and fix it before it crashes. Another example would be a incorrect MDI command which I would like to stop. Is the estop the intended way of doing this or is there some machine-stop signal I could tie to a physical button?

Anyway thank you for all of you guys involved in developing this awesome environment!​​​​​​​

1. All plasma parameters are stored in xxx.prefs in the config directory.

2. That seems odd, if the torch doesn't start you should get an error message and the torch should move to the top and wait for either a Resume or Stop.

3. If you have Keyboard Shortcuts enabled then Esc will stop most motion, you do need to understand that it is userspace so there may be a short delay.

1. Thanks! I will check that next time I am at my workshop
   
   
2. I could maybe record a video and gather more information about this. I have had this happen few times while testing out different consumables etc. At first I didnt realize it was a Hypertherm thing and I was just confused why torch wouldnt fire. If I remember to do the torch pulse before starting the process everything works fine as the torch firing warning event is triggered beforehand.
   
   I have lost arc couple of times in the early stages of testing and that indeed stopped the program and LinuxCNC was waiting for either resume or stop command. This somehow totally bugs out the system. Ill list the things I tried to recover
   - Stop current program and load a new one from file
   - enable / disable torch
   - Pulse torch manually
   - E-stop the machine and repower the drives
   
   
3. Thanks! I dont know why I didnt think of testing the Esc-key. The slight delay is fine I think, just good to know that there is a way to stop it. Is it possible to tie this stop command to a physical button on my control panel instead of using the keyboard? 
   
   
   ​​​

2. The more information that you can give us the better...

3. Yes: linuxcnc.org/docs/devel/html/plasma/qtpl....html#qt_ext-hal-pin
 

 
Hello again!

I figured the issue out today.

It seems the retry delay for failed torch start is 60s by default. The second try indeed happens after this timer runs out, one just needs to have patience enough.

The odd thing as new user is that the timer is not cleared when job is stopped and even when new job is started with new cut pattern. If the torch misfired the user needs to wait that retry delay before torch can fire again. Anyway this mystery was now solved!

I started cutting some test stuff and immediately ran into problems. Can I ask these things in this same thread or should I make a new one? I mean these are clearly not related directly to PlasmaC but rather some general CNC plasma operator questions.

My cuts on 2mm mild steel using normal consumables are in quite a noticeable angle. I thought the plasma could do at least a bit better in regards of the cut edge quality. Start and stop points also seem to get splatter and smaller circles are totally out of shape. 

The testing results are really confusing to me. I tried decreasing the cutting height from 1.5mm (Hypertherm manual) to 1.2mm but it did not have any effect. I have tried enabling / disabling THC, changed consumables, checked air pressure, calibrated Z axis, by eyeball I would say that the ohmic surface detection correctly zeroes the Z height and then cuts from about 1.5mm height. I also tried cutting with higher and lower speed than what was suggested in the hypertherm operator manual.

Oddly nothing seems to have any impact on the quality. There is almost no dross but the bevel on the cut edge is very high. 

Can anyone of you veterans give some hints where to look or what I might be doing wrong.

Just a thought is the torch at right angle to the table

Clive wrote:

Just a thought is the torch at right angle to the table

 

I will have to check that. I assume it is but I have been wrong before.

What I forgot to mention that clearly all inside-cuts are beveled outside and outside-cuts are beveled inside. I would assume that this is a telltale sign of torch being too high during cutting but I measured it and it really seems to be about 1.5mm from the surface. And as I mentioned that I even tested on lower torch height setting.

If you use Auto Volts do you have a reasonable THC Delay?
If it is too short then usually the torch will rise after THC becomes active.
For my PM30 it needs a minimum of 0.3 seconds, I mostly use 0.4 to be safe.

phillc54 wrote:

If you use Auto Volts do you have a reasonable THC Delay?
If it is too short then usually the torch will rise after THC becomes active.
For my PM30 it needs a minimum of 0.3 seconds, I mostly use 0.4 to be safe.

I am using 0.5s as delay it seems. All THC settings are still at default.

I tried cutting a bent piece of metal and the THC worked great at least by eye. It successfully managed to ride the plate down the hill and then back up again on the return trip. Neat system you guys have developed here!

I used HalScope to record few cuts. Is there way to increase the recording length somehow? I needed 4 signals so the recording is only 4seconds long. I guess the storage space (RAM) is not an issue with current PCs so more samples would be ok.

From the HalScope it seems that the Arc voltage value steadies within 0.2s or so after torch has reached desired cutting height. 

I adjusted the X-axis torch alignment. It might have been a bit off. Now I got some of the edges almost straight but the diagonal is even more beveled. The Z axis carriage I built seems to be pretty flimsy. I f I move the torch hose from where it exits the drag chain the torch head seems to be moving slightly. I wonder if this, quite small, movement can have such a drastic effect on the quality? The misalignment in X axis direction was also quite minor and it indeed seemed to change things.

I would imagine that 0.5s would be plenty, but to scope it, try using plasmac.thc-active as the trigger instead of plasmac.torch-on, and set the horizontal to 20mS 100mS or so. You want to ensure that the arc voltage is stable before THC becomes active.

Edit: oops,...