Voltage Scale and Auto Volts

Hi all,

We have a Hypertherm 125 for our plasma table, so we can cut up to 32mm plate. We are using a THCCAD-10 on a 50:1 voltage divider. and have set the voltage scale according to the sticker on the THCCAD-10 and the online calculator. We use MODE 1 with the ARC OK signal from the Hypertherm.

We are having trouble cutting thicker plate, 20mm and above. The machine cuts alright if we use IGNORE ARC OK, because it locks the THC. As soon as we unlock that the torch or goes up, or dives in the material. We use AUTO VOLTS. The Hypertherm manual gives a cut voltage of 169V for 25mm plate, but we see only approx 120V on the screen when it is cutting 25mm plate.

What does the AUTO VOLTS feature do ? Is it trying to get the plasma controller to cut at the voltage in the material table ? In which case we are way off the mark with our voltage. Or is the plasma controller sampling a voltage after ARC OK (when it presumably is stabilized), and tries to work around that voltage for the torch height ? In that case the voltage scale could be the culprit ?

We do not have full knowledge on how QtPlasmaC handles input voltage and works out the torch height, which makes it a bit difficult to troubleshoot for us. Any idea are welcome.

Cheers, Peter

If AutoVolts is enabled then the THC target voltage is sampled from the actual cut voltage after the velocity has reached 99.9% of the g-code feed rate and then after the THC Delay time has expired.

If AutoVolts is disabled then the THC target voltage is read directly from the Cut Voltage in the selected material.
 

Hi Tommy and Phil,

Thank you for the info. @Tommy : out of curiosity, what is the reasoning behind lowering the acceleration ? I will give that a crack, just wondering why it affects the THC ? Do you also change the acceleration on the Z-axis ?

Cheers, Peter

I like auto volts. There is a natural voltage at a specific height above the material that stays constant.

But I would fire up linuxcnc's halscope and monitor torch voltage when cutting including a pierce and look at how long it takes to settle to a constant voltage with THC disabled and what the voltage is. That will be quicker than trial and error as you can do it with one cut.

I think you will find that you need to extend your auto sensing delay on thicker plate. The voltage is quite erratic on piercing so if its sampled a low level (possibly too early), as the THC is disabled , the torch will dive to bring the voltage down from the correct height to the incorrect sampled height.

Move it out to  2 seconds, 1 second, then 0.5 second (the minimum I would use)
2 seconds will be fine on that kind of material at slow speeds

 

Hi all,

Thanks again for all the info. I have tried different things as per all suggestions from you all, but the main problem is still there and seems to be this : the machine starts cutting and after a undetermined time it just rapidly lifts up the torch and of course then it loses the arc. Sometimes it goes the other way around, the torch just dives into the material. On 25mm plate the feedrate is not very high, so the THC should be able to maintain height at its leisure. But it just goes awry in the blink of an eye.

The problem is that this happens very erratic, sometimes we cut half a meter without problems and then all of a sudden the torch decides to go whichever way it wants to go, buggering up a part. And sometimes this happens after say 50mm of cutting, and sometimes it doesn't happen at all. Something seems to trigger the torch to go up or down without a feedback loop to correct it.

Most of the time the THC works splendidly, it is very fast and is highly responsive. It's just this gremlin we are chasing and we can't put our finger on what it is.

Cheers, Peter

Yes, plotting the voltage with halscope when idle and when cutting
should help pinoint what's going on. Normal operation followed by
crazy behaviour at random times sounds like a hardware issue of
some kind.