THC not functioning properly < resolved >

[mjf53123]

Hello all - please help

The THC on my new acorn build is not doing what it should be doing. the THC will dive or rise randomly, even if the voltage is incorrect, almost seems like it is going in the wrong direction. when it should be going up a little, it goes down, and vice versa.

I was cutting some 3/16 material, and 3/4 of the cut cuts great, and for some reason, in a few spots (material is not warped, not flat, but not warped), the torch will go +1/2" from the material and doesn't come back down. the target volts are 135 I believe and in this case, it's reading 150-160 but doesn't decrease in height.

In another instance, I was cutting some 14ga material, and all of the small stuff came out fine, keeping 130-135 volts like I wanted it to. But as soon as it begins in the border (longer sections) I can watch the voltage stay at 135 then the z starts plunging and the voltage drops to 102 before the torch breakaway/float is activated. begins the cut at .060 as it should then drops down to .007 mid-way through the cut where the material is slightly higher.

When it does fail due to the "torch break away" the program just freezes, it doesn't give me the option to raise the z and resume the cut. It did in the beginning, but not anymore. It forces me to cycle the reset button multiple times to clear the issue and let me raise the Z. Very Strange as well.

In the link below is a short clip and some pictures of the end of failed cuts showing this. (I know it's ugly, just trying things out on scrap.)

https://share.icloud.com/photos/020G4SH ... hscErZQpgA

This is telling me that the machine needs to go up, not down, and vice versa for the thicker material as it began the cut in the higher spot. Is there a setting somewhere? Everything is wired correctly as far as I can tell and it passed all of the com tests along with doing the calibration at least 30 times today.

See the attached wiring diagram that I followed and my most recent report. I have a PrimeWeld Cut60 using the Hypertherm torch and consumables. So, I used a bit of both wiring diagrams since I needed my ohmic touch-off in conjunction with my float switch.

Any help is appreciated, I am falling very far behind on some work and need to get this thing dialed in before making the long cuts that I need to do.

[Joey]

Hello MJF


Can you take some screen shots of the most recent plot captures.

Open the Plot In the F7 Utility menu.

Inside the plot select File and navigate to the cncm folder to load a different plot.

The failed cuts will create a plot file that will save to the cncm folder.

I'm not at a PC currently to open the Plot and look my self

[mjf53123]

Joey - thank you for the reply, see the images below, my apologies for not responding sooner, I was not able to get back over to my machine until tonight. Is there a way that I can open CNC 12 without being connected to the machine? My CNC PC is portable.

I have more or less confirmed that when the torch becomes too far away from the material and the voltage increases, the torch rises even higher, and vice versa when it gets too close. it appears that it goes to Z1 and Z0 in these instances respectively. Very strange that it is compensating in the wrong direction. Maybe just a setting somewhere.

Based on my previous post about the Z homing, I have since added a limit switch at the top of the axis so that Z home is about 6" above the cutting bed along with shielding the THC voltage wire, it still runs within 3" of the acorn board, but is shielding the entire length.

[Joey]

Voltage Plots look pretty good.

I see your Accel/Decell rate is at the default .5 for all axis.

Try and Tune these to the lowest amount possible. Getting to at least .2-.3 for all axis will be a lot better.

Getting the X Y and Z up to speed faster should help with THC reacting to the ARC voltage.

Most of the Calibrations look like the material was un even. I think this is throwing off the Target voltage a bit also.
For example in the plot below when the Target voltage calibrated to the Actual voltage you can see the actual voltage was never level and constantly was rising during the cut.

[mjf53123]

Joey - I can do that speed up the acceleration and see if that helps along with making sure that the material is perfectly flat. Do you think that that could be causing my issues where it goes the opposite direction and just stays there?

Do I have to calibrate every time I change consumibles, material, air pressure, etc? Or is this a one and done sort if test? Should I change the THC voltage cut off to something around 70% to accommodate for smoothing and tight corners or no?

What about the program not giving me the option to resume when the cut fails? Any thoughts on that?

[Joey]

Joey - I can do that speed up the acceleration and see if that helps along with making sure that the material is perfectly flat. Do you think that that could be causing my issues where it goes the opposite direction and just stays there?
Most of the Plots show the THC reacting correctly to the Actual voltage. I see auto sense is used. Its best to just use auto thc in the profile so that the Target voltage for the cut doesn't change.

Do I have to calibrate every time I change consumibles, material, air pressure, etc? Or is this a one and done sort if test?

Calibration is used to match the THC voltage to an existing cut chart. Prime weld voltages very and are not consistent like other cuts charts. The Best way to use THC with prime weld is to get the actual THC voltage from a specific material and set as the target voltage for the profile.

You could Just Calibrate every time the profile is changed to quickly match the actual thc voltage to the Target voltage. But keep in mind that calibration effects all other profiles target voltage. Every time a different profile is used the Calibration will need run again.

Should I change the THC voltage cut off to something around 70% to accommodate for smoothing and tight corners or no?

Anti Dive threshold is 7% by default. You can try adjusting to 10-20% to prevent the THC from getting stuck at a higher height when the voltage spikes. Anti dive threshold is the purple lines in the plot

Anti dive velocity is at 90% by default. I would increase this number to 99%. Some of your existing plots show that the Voltage is pretty high if the XY velocity is around 90% and THC reacting at less that 95% feed rate may be causing the Z to lower into the material.


What about the program not giving me the option to resume when the cut fails? Any thoughts on that?
Are you jogging the Z up after cycling the VCP rest button? Restart mode won't work until the torch is untripped. You can always jump back into restart mode after canceling out of the job also.

With your lower resolution I'm thinking you might be a good candidate to try the latest 5.09 Beta. Using smoothing in the 5.09 Beta should help keep the speeds up.

[mjf53123]

Auto THC, got it.

I’ll just plan on doing a calibration every time I use a different cut profile for now, but how would I get “actual voltage” if I wanted to update the target voltage? Turn THC off and do a cut watching the voltage and use that to set the target?

I’ll try raising those numbers but that means the THC is not going to be active below 99% of the target speed, that might create some issues when doing a large but intricate cut?

Lower resolution as in only having 2000 steps per revolution? I suppose I could turn that up if that would help, but with my axes having 4:1 and 5:1 gear reductions( stepper has the small gear and axes have the large gear) the steppers just don’t seam to have the power. It moves about 6” with one rotation. I can give the beta a shot if it’ll keep the speeds up during smoothing.

Which fields do I need to change to make the z move faster when THC is active and during touch off?

[Joey]

Auto THC, got it.

I’ll just plan on doing a calibration every time I use a different cut profile for now,

but how would I get “actual voltage” if I wanted to update the target voltage? Turn THC off and do a cut watching the voltage and use that to set the target?
Yes just make sure the cut is long enough that the X/Y velocity is up to speed and the material is flat.


I’ll try raising those numbers but that means the THC is not going to be active below 99% of the target speed, that might create some issues when doing a large but intricate cut?

Try to get the Accel/Decell for the axes as low as possible before changing the THC config settings.



I noticed in the Centroid Job the X and Y never had time to come up to speed.

THC only enabled when on a long straight, which for the Centroid job isn't until the very and at this point in the cut the Voltage was much different than the Target voltage , leading to undesirable THC reactions.
Auto sense might of had a hand in that also since auto sense doesn't select a voltage until the X and Y are at the target velocity and with a Slow Accel/Decell rate this could be far from the beginning of the cut or the job.

Lower resolution as in only having 2000 steps per revolution? I suppose I could turn that up if that would help, but with my axes having 4:1 and 5:1 gear reductions( stepper has the small gear and axes have the large gear) the steppers just don’t seam to have the power. It moves about 6” with one rotation. I can give the beta a shot if it’ll keep the speeds up during smoothing.

Can you throw pictures up of each axes. 1 rotation of the motor for 6inches of travel is a .2 overall turns ratio. Some picture of each axes would help clear this up.

Which fields do I need to change to make the z move faster when THC is active and during touch off?

The Z max rate is pretty slow at 75ipm, But getting the Accel Decell dialled in for the Z should help a lot. The Touch off rate is handled in the THC config setting as your fast and slow probe rate.

[mjf53123]

I’ll get some pictures of the axes after work and address some of the changes. All I can do for now is try to explain it.

- The Y and A axes have a have a small cog gear on them with a belt leading to a larger cog gear then it has a shaft going down to the rack and pinion. Giving me a 4:1 overall.

- The X axis has a planetary gear reduction and then the rack and pinion. Giving me an overall 5:1.

- The Z axis has the threaded rod directly attached to the stepper where 2 revolutions is 1”.

See image for the overall turns ratios that I currently have where 1” is 1”.

How can I test the acceleration rates? How do I know when I have gone too far?

Can I edit the calibration macro? I don’t think that the little 2” cut that it makes is far enough to get an accurate reading as it never got up to max speed? Maybe it will be after I can get it to accelerate faster.

I would agree that it only came on during straight aways and thinking about it, the only time it was at full speed.

The Z can pretty easily move at 300 IPM but I didn’t see a point in allowing it. Should I give it the ability?

[Joey]

I’ll get some pictures of the axes after work and address some of the changes. All I can do for now is try to explain it.

- The Y and A axes have a have a small cog gear on them with a belt leading to a larger cog gear then it has a shaft going down to the rack and pinion. Giving me a 4:1 overall.

- The X axis has a planetary gear reduction and then the rack and pinion. Giving me an overall 5:1.

- The Z axis has the threaded rod directly attached to the stepper where 2 revolutions is 1”.

See image for the overall turns ratios that I currently have where 1” is 1”.

How can I test the acceleration rates? How do I know when I have gone too far?
So the Accel/Decell is the amount of time its takes to get from 0 ipm to the max rate.

The best way to dial this in is to drop the Accel/Decell rate by %10 percent.

Every time the rate is changed run a Job that Moves the Axis back and forth at the max rate.

G0 is a rapid move, Rapids are done at the Max Rate.


Mark the beginning of the Job or Work Coordinate Zero to make sure no steps were lost and the axis returns to WC0.

Keep decreasing the Accel/decell rate by 10% until steps or position is lost.

If steps or position is lost just revert back to the previous Accel/Decell rate.

For example the X axis Job test would look like
G0 X10
G0 X0
G0 X10
G0 X0
G0 X10
G0 X0
G0 X10
G0 X0

Can I edit the calibration macro? I don’t think that the little 2” cut that it makes is far enough to get an accurate reading as it never got up to max speed? Maybe it will be after I can get it to accelerate faster.

The Calibration takes into account all the variables of the Wizard settings and Profile manager.
During the Calibration Cut a Plot is captured and I can see from the Plot that the X and Y velocity was up to speed.
The voltage was consistently climbing during the calibration which points towards an uneven material.

I made a short video of your calibration Plot you can view here

I would agree that it only came on during straight aways and thinking about it, the only time it was at full speed.

The Z can pretty easily move at 300 IPM but I didn’t see a point in allowing it. Should I give it the ability?

Faster Z movement will be needed for cutting at higher feed rates and thin material that may warp.

At least tuning in the Accel/Decell rate is a good starting point.