THC enabling. Are we doing it wrong?

www.plasmaspider.com/viewtopic.php?f=3&t=13359#p78176
Jim Colt says:

6. Once the machine achieves the programmed x - y cut speed, the height control starts monitoring the DC arc voltage (as measured between the plasma torch negative electrode and the positive work cable) and compares the actual arc voltage to a preset arc voltage (as suggested by the plasma torch manufacturer for the material and thickness being cut), then the height control adjusts the z axis so that actual arc voltage and preset arc voltage are essentially the same. If actual voltage is too low...the z axis indexes up until the voltage matches the preset.

Its occurred to me recently that the piercing voltage spike we see is probably heavily influenced by acceleration. So a fast accelerating machine will get to cut speed faster and once it does voltage should stabilise. Here is Jim confirming that.

Where is our test pilot? this would be a perfect opportunity to learn how to use halscope and monitor torch voltage, current velocity and feed rate and see if without a THC torch voltage stabilises when you get to cut velocity. If it does, we could get rid of the THC delay and enable it at cut velocity (or some small delay after that is reached).

The more we simplify operations, the better I think!

If that is the case you could get rid of the delay now as it doesn't sample the arc voltage until it has reached 99.9% velocity.

phillc54 wrote:

If that is the case you could get rid of the delay now as it doesn't sample the arc voltage until it has reached 99.9% velocity.

Exactly my point! But when does it enable the THC?

snowgoer540 wrote:

Test pilot here... did you think I was going somewhere?

To your point, I don't know if I've ever seen a THC delay on another machine before, but that doesn't mean that it wasn't baked into the controller and wasn't user programmable.

Measuring THC delay is on my list of things to do at some point anyhow, as is learning Halscope. The two seem to be hand in hand LOL. This is following me, it will catch up and tackle me shortly, don't worry.

I think most THC's will have a delay but Linuxcnc is one of the few that has access to the motion control parameters so we should be able to get smarter...

rodw wrote:

phillc54 wrote:

If that is the case you could get rid of the delay now as it doesn't sample the arc voltage until it has reached 99.9% velocity.

Exactly my point! But when does it enable the THC?

After it has collected the sample and the delay has expired.

Edit: I should add that the sample is taken over six cycles after reaching 99.9%

Its something I never considered in the past when I was doing cut after cut in halscope. (Phill put me out of a job). I have not really done any cutting since upgrading my machine to 5 m/s^2 acceleration (0.5G roughly). I still need to stiffen up the gantry ends desperately but 50+% sales growth has severely cut into my quality plasma time! But that also means I need to cut more parts.

Thats why it needs testing on a few machines.

Following this discussion in the background......what are the units of the velocity and Accel in the ini file for each joint / axis