Need Help!- Z height micro switch control

[MonoNeuron]

Hi Guys,
Could someone enlighten me on how to set up the Z axis for my plasma?
I have bought the MP1000C from CandCNC.com and Tom says I need to install a micro switch for the height adjustment and torch set off procedure. He spends quite a lot of time answering other peoples questions so I thought I would come here and leave him alone for a while and see what you guys can come up with.

I have no idea how to set this up.

Do I install the micro switch at the bottom of travel and zero from the top of the plate or do I set the switch at a known height from the table and input the thickness into Mach3 and work from there?
Where do I set the switch? 1" above the table. 3" maybe?
Do I make it adjustable so I can slide it along and set it anywhere for any height?
Will it get in the way of the torch when it cuts corrugated iron, for example, if it's set too low?

I am in a quandary so any help will be gratefully accepted.
Richard.

[Torchhead]

You missed an important part. The switch is part of the Floating Torch Holder. It allows the torch to ride up and down a few inches independent of the Z movement

It's like the coil springs on a 58 Buick Roadmaster. When the torch moves down for a Reference (Z Home) the torch tip strikes the metal and starts to shove the floating part up. A switch is mounted so that upward movement of the torch on it's floating holder trips the swtich after "?" amount of movement. That stops the REF move. The overtravel between the time the tip just touches the metal and when it trips the swtich is called the "SwitchOffset". That will vary from machine to machine but will be a constant on any one machine.

The switch is mounted so that the action of the floating holder moving up trips it. It's far above any cutting or material.

By doing this it will automatically find the top of any thickness (or warped) piece of material and reset Z so when the code says "lift .125 and fire" for a pierce it's .125 above the material at the pierce point.

SheetCAM has variables in the POST to enter in your SwitchOffset value and a RefDistance value (in MM) that determines the Pierce Interval (how often the torch touches off). It then does auto-touchoff at preset intervals to compensate for warped or unlevel material and get a perfect pierce each time.

So the switch does not hang down or touch anything but the floating torch holder
slide. There are several photos on the CandCNCSupport Yahoo Forum in the photos section and there is a file with pictures in the Files section.

You gonna cut corrugated iron? You will need to make sure the Z is fast enough to keep up with X & Y. Remember that the Z under THC control runs at a percentage (THC Rate) of full Z velocity to prevent lost steps. Too many rapid short moves in opposite directions to use any Accel or Decel numbers during cutting. The lower velocity compensates for infinite Accel. If you make the Z too fast it will decrease the resolution and that causes overshoot (head oscillation) problems.

TOM CAUDLE
www.CandCNC.com

[MonoNeuron]

Thanks a bunch yet again Tom

Ahhh, The old 58 Buick Roadmaster trick. I should have thought of that earlier.
I have seen a lot of photos but could never see what or why the extra slide was there but the pics you directed me to are just great and now I am armed with a lot more information and everything is getting a lot more clearer.
I couldn't find any posts on the forum here on the Zone outlining this.
I am still having problems with the controller and it may be a dud parallel port card. I tried to get one in town over the past few days but either they had thrown them out because no one wanted one for the past 4 years or they just don't stock them anymore. I ordered 2 from the Philippines for $24 which should arrive in a few days.
Until then I will get into building the floating Z axis and rip out the parallel port card from my laser and try that.
As it stands now with no plasma cutter connected I have a continuous light on on the arc good on the front panel and no matter what I do I can't get it to turn off. It also pulses on and off at about 1 sec intervals when I jog the Z axis up and down then stays on when I stop jogging.
None of this is still making any sense to me so I will get back to you when I install the new parallel card.
I think building the laser table was way more easier than the plasma cutter, even with all the problems I had with that
Thanks for your continued help, Tom
Richard.

[Robin Hewitt]

I drop the torch under it's own weight (approx 2 lbs) until it rests on the metal then back off.

Of course if the weight of the torch flexes the sheet then it follows the torch up when I think I'm lifting clear.

If this is a work around I would love to understand it. Unfortunately I'm not au fait with the 58 Buick Roadmaster

[Big John T]

Originally Posted by Robin Hewitt I drop the torch under it's own weight (approx 2 lbs) until it rests on the metal then back off. Of course if the weight of the torch flexes the sheet then it follows the torch up when I think I'm lifting clear. If this is a work around I would love to understand it. Unfortunately I'm not au fait with the 58 Buick Roadmaster

If you have problems with the sheet flexing you can counter balance your torch or use a micro switch that is extended with a small air cylinder to feel for the top of the material then retracted out of the way before cutting.

John

[Torchhead]

A simple bolt and spring at the top of the Floating holder to provide up bias against the weight works fine. I mounted a small block with a through (clearance hole) on the Z plate and tapped the top of my floating torch holder. With a long bolt and a spring between the head of the bolt and the small block it wants to lift the bolt Up. When screwed into the topof the floating holder it pulls the holder up. The tightness of the bolt determines the uplift force. I can dial in whatever amount I need. For really thin material the table slat setup needs to be closer spaced to provide more support under the material.

[Big John T]

Torchhead,

Elegant and simple solution. Very nice. I tend to over design things sometimes

John

[Robin Hewitt]

Had an idea but I don't quite understand how a plasma cutter works...

There are two connections to the torch, the loose wire goes to the cutter nose. If I could fire a a low voltage spark down that I could tell if it was touching metal same way a spark eroder knows when it's touching metal.

That wire can't be a solid ground, so what does it connect to inside the plasma box? Can I switch it with a fat relay?

[Torchhead]

It's called ohmic sensing. It not a spark but a low current where the torch tip shield is isolated from the body and has a connection back to a sensing circuit. When the tip touches the metal (other side of the circuit you get current flow) It sounds simple and elegant....only problem is it sometimes fails to work if the tip gets dirty or the surface you are cutting is oily or rusty, so they normally back it up with a torque sensor on the Motor to keep from driving the torch too far and busting the tip.

A good engineer knows when to stop designing. The simplest circuit that does the job and is the most reliable wins out over elegance.

BTW low voltage won't spark across a gap. The insulation factor of dry air at sea level pressure is 10,000 volts per inch. To pass current you have to exceed the breakdown insulation of the media. That means .01 inch needs over 100 volts. To get a good spark you need many times that.

You might be able to measure a current between the electrode lead (which in some machines is in contact with the tip) and the work clamp when the tip comes in contact with the metal but it would need to be integrated in with the plasma electronics. I know that a TD machine will not fire if the tip is setting on metal (at least on some models) so they have to be able to sense the short. An external circuit would have to measure millamps of current at low voltage then be able to survive hundreds of amps at hundreds of volts. It could also suffer from the same problems as ohmic sensing so a backup method to stop the probe move would be needed. If you abandoned the floating head the tip would have to stop pretty quick or you would end up with a box of busted ($$$) barrel rings and tips.

TOM CAUDLE
www.CandCNC.com

[Robin Hewitt]

Originally Posted by Torchhead A good engineer knows when to stop designing. The simplest circuit that does the job and is the most reliable wins out over elegance.

Engineers do not necessarily make good designers, too tied up with worrying about the ins and outs of HOW to make it. A good designer simply puts in everything he wants, the engineer then has to try and get him to compromise for what can reasonably be made

If I took over the connection to the torch tip, applied some capacitance and charged it 50-100 volts before starting the descent to look for metal, a sudden drop in volts would indicate that I had found it. The dialectric would be either rust or guck on the sheet to be cut. I just need a few volts to cut through that guck.

The problem is where does the wire go inside the cutter? Presumably the cutting current goes to the inside electrode. The other cable must be for the initiator spark. I think the initiator sparks to the central electrode to ionise the air coming out the tip. Everything I read is a bit vague on initiator sparks, probably because thee are different ways of doing it

I need to know what goes down that wire that connects to the outer electrode.

[MonoNeuron]

I finished the floating head a few days ago and my new parallel port cards arrived today so I will be busy out back finishing the plasma cutter off over the next few days...in the cold...Brrrrrr. damned Antarctic wind!
I have 1 question though and that is where do I connect the switch for the height sensing?
Does it go on the Z axis limit or the Z axis home or is it a different switch I know nothing about. Can't seem to find any info on this :-)
here are some shots of the floating head for all and sundry.
It has about 8mm of travel which I think, if I understand the concept, will be enough travel for it. After all, it only has to trip the switch and then move a known height which I think I can find out by jogging the Z axis in Mach.
I just hope I can get it all together soon.
Rich.

[Torchhead]

Limits on a machine just stop the motion signals....You use the Z home signal since the touch-off sequence looks for a specific signal and stops then backs off slightly. The code to do all of that and reset the Z to the proper 0 height AND then raise to the correct pierce height come sout of the POST in SheetCAM.