Bad cut one side

[grindergary]

I'm using HT 105 at 65 amp, book specs (witch cut very close). My long axis is X and short Y

When I cut from X10 to X0 and then Y 0 to Y10, the thc bounces like crazy. The other cuts are fine. Now, this only happens on edge of plate. If I cut a square in next cut up, it will cut really nice until it reaches the end of sheet. Then it happens again. It will not do it cutting Y10X0 to x10. It will cut nice and clean.

This happens on 3/8 and 1/2 inch plate. I have tried different plate and different manufacturers. Same thing.
On 10 or 16 gauge, I have not noticed any problems.

I have clean dry air, and all mechanical seems tight.

Any suggestions appreciated.

Gary

[jimcolt]

The height control is an arc voltage feedback system. If you cut close to the edge of the plate...or right on the edge, expect the arc voltage reading (that your height control uses to adjust height) to be very erratic, which will cause the torch to oscilate up and down. You should always have at least kerf width of uncut material when cutting near an edge.

Jim Colt Hypertherm

I'm using HT 105 at 65 amp, book specs (witch cut very close). My long axis is X and short Y

When I cut from X10 to X0 and then Y 0 to Y10, the thc bounces like crazy. The other cuts are fine. Now, this only happens on edge of plate. If I cut a square in next cut up, it will cut really nice until it reaches the end of sheet. Then it happens again. It will not do it cutting Y10X0 to x10. It will cut nice and clean.

This happens on 3/8 and 1/2 inch plate. I have tried different plate and different manufacturers. Same thing.
On 10 or 16 gauge, I have not noticed any problems.

I have clean dry air, and all mechanical seems tight.

Any suggestions appreciated.

Gary

[grindergary]

[quote="jimcolt"]The height control is an arc voltage feedback system. If you cut close to the edge of the plate...or right on the edge, expect the arc voltage reading (that your height control uses to adjust height) to be very erratic, which will cause the torch to oscilate up and down. You should always have at least kerf width of uncut material when cutting near an edge.

Jim Colt Hypertherm

Jim,
I have been cutting up to 1/2 an inch from edge just to see if it would stop. The voltage remains constant at 135 plus or minus 1 volt. It does not change even with THC off.Voltage stays constant. Ground on material, ground spot just to make sure it was not mill scale. I have tried different metal suppliers. Really at a loss on this one.

Thanks for the response,

Gary

[jimcolt]

I gues I don't understand what you are saying. If the torch (z axis) is oscilating up and down.....that changes the arc voltage (voltage is proportional to arc length). If you are saying the torch is oscilating up and down...and the voltage is constant...then that is just plain not possible.

Can you describe your issue again...in clearer terms?

Jim Colt

The height control is an arc voltage feedback system. If you cut close to the edge of the plate...or right on the edge, expect the arc voltage reading (that your height control uses to adjust height) to be very erratic, which will cause the torch to oscilate up and down. You should always have at least kerf width of uncut material when cutting near an edge.

Jim Colt Hypertherm

Jim,
I have been cutting up to 1/2 an inch from edge just to see if it would stop. The voltage remains constant at 135 plus or minus 1 volt. It does not change even with THC off.Voltage stays constant. Ground on material, ground spot just to make sure it was not mill scale. I have tried different metal suppliers. Really at a loss on this one.

Thanks for the response,

Gary

[B Holmes]

I had a similar issue the first time I cut 1 inch plate. THC acted like a sewing machine, only faster. Tom from CandCNC had me turn down the THC rate. My issue was solved. Hope this is a help.

[grindergary]

[quote="jimcolt"
Jim,
I have been cutting up to 1/2 an inch from edge just to see if it would stop. The voltage remains constant at 135 plus or minus 1 volt. It does not change even with THC off.Voltage stays constant. Ground on material, ground spot just to make sure it was not mill scale. I have tried different metal suppliers. Really at a loss on this one.

Thanks for the response,

Gary

Jim, the Z axis oscillates and the DRO's show only 1 volt difference if you are fast enough to see it. If I do the same cut with THC off, The arc voltage will still only move 1 volt on that cut. If I do a cut 96 inches long on 1/2" plate, the thc will follow that cut with 1 volt difference then adjust back to preset voltage. I don't know how else to explain it.
Driving me nuts.

Gary

[grindergary]

I had a similar issue the first time I cut 1 inch plate. THC acted like a sewing machine, only faster. Tom from CandCNC had me turn down the THC rate. My issue was solved. Hope this is a help.

Where would I find this setting?

Thanks

Gary

[B Holmes]

I believe it is on the settings screen in MACH. Not near my table right now, but I will check the settings tomorrow.

[beefy]

The 1 volt difference makes sense because in the THC Cut Profiles the span setting may well be close to that (e.g. 4/4 ths is 1 volt). Anything outside that setting and the THC reacts. If your Span setting is lower (e.g. 2/4 ths is half a volt) this will also add to the tendency to oscillate. I often use 2/4 for Finecuts and 4/4 for larger nozzles. The Span setting is just a deadband area where the THC does nothing.

I too have had the "jackhammer" effect and reducing my THC Rate (Settings tab in Mach) from 40% to 15% was what fixed mine. The THC Rate does not have acceleration or deceleration when it gives steps to the Z-axis, that is why it cannot be at 100%, the steppers cannot instantly go to max rpm.

What causes the jackhammering is the THC reacts to correct the voltage but overshoots through the deadband area (span) to the other side of the volt setting, then it reacts the other way and overshoots in the opposite direction. Reducing your THC Rate helps the Z-axis stop in time (within the Span area). Also increasing the size of the deadband gives the z-axis more chance to stop before overshooting.

Keith.