THC diving after pierce.......

[jlocke09]

THC diving after pierce………..
I have a DIY table that I am having issues with the torch diving after pierce. Here’s some info on my setup.
- Using Mach4 and Sheetcam
- ESS board and TMC3in1 THC
- AD1 delay checked with a 1.5 second delay
I can wait a second or two into the run, then turn THC on and it works perfectly. It’s almost like my AD1 delay is not working. Any thoughts? I’m very new to this.

thanks

[SeanP]

Depends on thickness you are cutting, might not be long enough delay, have you tried longer?

[jlocke09]

3/8" thick. I changed delay to 2 seconds and it still crashes. I've watched the voltage and it seems to be balanced out at 130 volts well within the 1.5 secs.

[robertspark]

need to give the voltage time to stabilise once the pierce is complete and the machine is near target feedrate

depends upon your acceleration.

the normal sequence is:
M3 (torch fire)
[wait arc transfer ~100mSec~ at a guess depends on your plasma cutter + thc]
G4 Pxxxx (pierce delay counter starts)
G0 Zxxx (move z to cut height [most oftern G1 is used with a feedrate])
G1 / G2 / G3 (cutting begins at a target feedrate)
once the machine is near the target feedrate (~>80%) the Arc Voltage is sampled and held or the THC just tracks the setpoint voltage.

Check your consumables are good too as they wear the sampled voltage at the cut height will change.

[jlocke09]

I'm using a Hyperthm 85 set at 60amps. Consumables were new but I changed them out just to verfy. So my sequence is just as you laid out. But after I get an ark ok the torch never stops at the .100 cut height if I have the THC on. It pierces, then plunges. It's almost like my THC delay is not waiting the 1.5 seconds like I'm telling it to.

[robertspark]

I'd suggest asking on the warp9 forum, given the tmc3in1 was developed software wise by warp9.

[Rodw]

Can your THC auto sample the torch volts? If so use that. In my tests plotting the torch voltage in an oscilloscope, ages ago, I came up with a 2 second delay was required regardless of material thickness to be safe. During this time, the pierce sees much higher voltage at the start of the cut.

But now I have other theories which I still need to test, but I think that this delay will be influenced by the acceleration your table is capable of. When you think about it, we program a pierce delay then start to move. It takes a bit of time to get to fresh material on the edge of the pierce divot and then get to cut speed. So if I can accelerate at 0.5 G and you can only manage 0.2 G I should get stable results faster than you can, so my THC on delay should be shorter.

If you enable THC too soon with a fixed voltage target, the volts will be too high so the torch will smash down to the target and possibly crash.
If you auto sample the voltage too soon, the THC will lift the torch too high as it seeks to get back to the sampled volts.

Sorry but I can't help with config of your system so get that sorted and eliminated as a cause

[jlocke09]

Thanks for all the replies.

I'll increase my delay time to 10 seconds just to see what happens then I'll reach out to warp9 forum.

[Rodw]

I think 10 seconds is overkill. Try 3
But first to a 10 second cut and observer when the voltage stabilises.
This is not the best plot but the line with the white dot on it is torch voltage.

Each vertical line is 2 seconds and the bottom squiggly blue line is the error between the desired set point and the actual so when it starts, the THC has turned on. You can see on start, the line goes off the chart briefly then humps up before setting on a natural voltage according to the cut height, material and cut speed. You can see this voltage is rock steady even before the THC is enabled (eg. before the blue squiggle).

[jlocke09]

Makes perfect sense. Thanks for the visual. It helped a lot.

[Zukzilla]

Make sure g4 is set to seconds in mach. In general settings youll find its set to ms. Hence the rapid dive