Thc rate

[4130metalworks]

Could i get some clarification on what the touch off feed rate and thc feed rate is. I was getting a fairly pronounced jagged edge to my cuts and reduced the speed for both of these which cured that problem, but now my torch is crashing into the material constantly. To me it looks like the thc isn’t reacting to variance in heights and not raising when it should

I would like to speed it back up but don’t want to do that without knowing which each rate is for
Thanks for all input

[djreiswig]

Guessing the touch off federate is for the IHS referencing moves and the THC federate is for when the THC is moving the Z when actually cutting. If the touch off federate is too fast, it might over run your switch and give inaccurate pierce heights. Incorrect THC federate would make your THC react slowly. I'd adjust the THC federate back up

[weldguy]

Sounds like you need to find that sweet spot between the THC feed rate that was fast enough before and the lower speed your currently at. I would try to keep the THC feed rate as low as possible if it is providing a smoother edge but fast enough to react in time as to not allow the torch to contact the material when it warps. If the rate you find that works is providing a rough cut edge there may be some additional THC tuning that you can do depending on what you are using for a THC.

[sheetcamCS2]

I agree with prior replys. Now consider that since your linear (cutting) feedrate is 1350mm/m, if THC Feed Rate was the same, that would allow your THC to react quick enough to navigate a 45deg slope, you should consider equal feeds then to be fastest case for THC feedrate. If that is too fast and affecting your edge quality, then make THC feedrate 1/2 of linear feedrate. I believe you are using a Masso or similar post processor, which is adding 'Stop after pierce' and later fields to the Jet Tool params. So it would be entirely possible to just make THC Feed Rate a function of linear feedRate in the pp, using whatever percentage of feedRate that you like on a consistent basis.

Let us know if you need help with that after you test it out as is with THC Feed Rate adjustments and test cuts.

[4130metalworks]

Thank you for the replies, i will try increasing the thc feedrate back up to half of the plasma cutting feedrate and see how that goes before going any higher.

[4130metalworks]

So i tried increased thc feedrate and left the touch off as is, 350mm/ sec was the max best setting. At 500 the torch was bouncing up and down like a pogo stick, strange it was that low a number

[sheetcamCS2]

...strange it was that low a number

Just a hunch, but it could be due to a couple of factors such as acceleration of Z and sample freq of arc voltage. Regardless, you seem to have gone through the exercise of tuning it. Well done. Also, did you mean to report it is 350mm/sec ? or /min ? big difference.

[4130metalworks]

sorry that was a mistake, 350mm/min
I do have the acceleration and motor speeds quite high, the z axis has a maximum feedrate of 8500 mm/min and acceleration 6000mm/sec^2 on a 5mm pitch ball screw
might look at turning down all max feedrates because the x and y can get quite violent

[sheetcamCS2]

might look at turning down all max feedrates because the x and y can get quite violent

I believe you're going to find that reducing acceleration will have much more of an effect on reducing the "violence" [mechanical resonance]. 6000mm/sec2 is pretty stout acceleration.

For comparison, I run nema 23 steppers setup to deliver 270 oz-in torque at 1400mm/sec2 accel, 16000mm/m max speed, but I only do that when raster etching with laser and router process with X (least mass axis). For plasma and laser cutting (XY vector motions), I dial down to 400mm/sec2 accel, 8000mm/min max speed. I should run 800mm/sec2 accel for plasma though, to keep the dross down when cornering and small arcs. Z is configured for 400mm/sec2 accel, 4000mm/min max speed all the time.

[tcaudle]

So i tried increased thc feedrate and left the touch off as is, 350mm/ sec was the max best setting. At 500 the torch was bouncing up and down like a pogo stick, strange it was that low a number

There is a lot of electronics involved but Ill try to simplify:

The THC reads the voltage at the cut which is proportional to the height (Arc Gap) given other variables (feed rate and air quality ) are constant.

The voltage feedback to the THC creates a classic "Servo" loop. Much like verbal communications the more they are delayed the less they make sense. I boils do to reading the voltage and moving the Z mechanically . Any delay in that process causes a delay in the actions both starting and stopping the motion /direction. That in turn creates overshoot. Meaning the Z moves too far and then once that value hits it corrects in the opposite direction but once again, overshoots and you get oscillation up and down . The two factors are how fast the UP and DOWN signals can occur versus how fast the Z can accelerate and move. When the response (feedback) in the THC is slow you simply cannot force the Z to move faster than it can do to form a stable "loop". So its NOT just motor tuning. Its timing and having the acceleration and speeds too high for the round trip loop response time will result in the sewing machine effect. Couple slow loop response (loop delay) with any backlash and it gets worse. The coarser the Z motion the worse the jitter.
So......you have to find the rate that does not cause the overshoot and the point of oscillation regardless of how slow or how fast that is. Much like an audio system too much GAIN causes bad oscillations too much GAIN (gain being the Z motion speeds)

To get faster Z response you have to tie the actual voltage changes to the actual federates as tightly as possible. Most Windows applications prevent that because Windows controls all system timing and there are lots of things that take priority over Input / Output (I/O) signals to hardware.

I wont go into things like "S-Curve Acceleration" Other than to stop overshoot, the loop has to vary the response speeds depending on how far away from the "target " voltage it is. its non-linear and helps dampen oscillation.

Touch off speeds need to be slow enough to not get overshoot . Once again its timing. The time it takes for the signal from he touch off sensor to reach the control software and stop the motion. The slower you make the touch off the better the accuracy (less overshoot) . Once again its about timing and how fast the control software can read the Inputs

It boils down to the physical and electronic response and matching them up /

[4130metalworks]

Thank you for the in-depth explanation tcaudle, sounds like I need to do a bit more tuning on my motors

Thanks again Lou, i will definitely turn it down. My x and y are 35000mm sec with 6500 mm sec acceleration. A good starting point i think will be to set the max feed rate to be slightly higher than listed in the Hypertherm cut chart

[tcaudle]

In motor tuning you set the motors to the physics of the table . The ACTUAL feedrate is a call from the g-code for a specific material and amps setting. Its the ACCELERATION (and Deceleration) that is not changed from code. Then you have the same parameters for the Z. Depending on the THC you either have to live with the global settings in motor tuning or you have a separate set of parameters used when the THC is in charge of moving Z. A good number for XY on most tables is in the 500 IPM range (12700 mm/min or 211 mm/sec) then acceleration needs to be no more than about 35 IPS/sec or about 890 m/sec/sec. There are reasons to have that higher on some machines where the THC is not connected to the XY motion and tends to dive on any motion slow down.

For ZS the numbers are more consecrative. Most leadscrew Z's can't move much faster than 100 IPM with stepper motor and usually acceleration is lower than XY. The THC if its controlling the Z may have a separate speed and acceleration and THOSE need to match the material being cut . Often acceleration is handled by the THC (old MACH3 built-in THC had infinite accel because the moves were supposed to be very short. Of course lack of acceleration curve begets overshoot when the Rate gets too high. The MACH3 internal rate for Z maxed out at about 20 IPM (what ever that percentage of the Z max velocity was set to) Some would go as high as 30%. over that they oscillated .
Once again its about the Loop delay and the longer that is the lower the Rate has to be.

You are at the mercy of the electronics and Physics of the motion. The dumber the THC the more conservative you have to get with the Z settings .

You tune the Xy for max velocity for rapids not to exceed the motor max RPM . You yune the acceleration as high as you can to not get slow down on corners but not so high it runs really ragged.

Then the Z has to consider its gearing and the loop response of the THC .

Just because a table can do 1500 IPM or more does not mean is needs to be set to that. High speeds and high acceleration puts dynamic stresses on a table that are not necessary. All you gain is faster rapids The overall time to cut drops very little even in a whole sheet of parts and usually setup (loading material and zeroing the machine) ad unloading and clean up far exceed any accumulated rapid time. Maybe in production where are are cutting a thousand parts over and over a minute or two saved on a sheet will add up as long as the operator never takes a break.