Re: Newbi needs advise on torch table. - now THC

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "turbulatordude"
<davemucha@j...> wrote:

> > MACH2 is cheap, well written, constantly being improved and the

THC

> > is real and working. I cut for 4 hours today using it.
>
>
> Hi Tom, and all,
>
> I am not totally clear on the THC problem.
>
> it seems that the table requies an X/Y to move the head, but the Z
> can be independant. lower to part, start cutting, let it be
> independant of the rest of the CNC software and just keep it's
> location based on voltage.
>
> Then just wing that sucker around the table as needed.
>
> why is it so critical to have the CNC software keep track of the
> head ?
>
> a command up with an end switch prior to any non-cutting moves, and

a

> start cutting with a resultant 'go' from the independant head

control

> seems like a simple approach.
>
> After all, don't a log of guys just ignore THC all together ?
>
> My EDM is similar in that the electrode postioning in seperate from
> X/Y and even total feed is not counted as electrode wear has a huge
> effect on distance.
>
> I 'home' to home, then re-zero on the first spark, then burn to

depth

> figuring no electrode wear, then home again, then back to the
> starting zero and then measure the electrode wear. From that I can
> caluclate wear in that part and how much farther I need to go.
> Mind you none if this is critical, I'm still setting things up.
>
> Dave

Dave: You are correct. You can run the Z totally independent of the
XY. There are several commercial products ($$$$) available to do
that. In essence they use an external micro chip to measure the gap
voltage and adjust the torch height using DC motors. That has been
the conventional way it has been done for a long time.

As an Engineer I started looking for a THC solution as I was building
my table. I could not afford the commercial units and the nearest
thing was a RUTEX kit that was about 500 bucks and required a
bewildering array of sensors that had to be constructed and provided
by the "Builder" (me). Since I had already planned on making the Z
axis a full Step and Direction design so my table could also be used
for other stuff, I felt is was redundant to design and build some
sort of electronic "switchover" so that the motors could be
controlled by the THC rather than the computer.....if only I could
find a piece of software that could do the "logic" part as well as
the XY control then all I had to design was the sensing and measuring
hardware. Art of Master5/MACH1 fame agreed to modify his software to
my specs and I set out to build a real cost effective solution.

Once you get out of the manual control or the set-it-and-let-it-cut
approach (both which I tried), then you face several issues:

1. You need to be able to control the pierce cycle. You need to be
able to find the material surface, move the head down to a specified
pierce point, and start the arc.

2. You need to wait a variable length of time until the arc has been
established, lower the head into the cut then release and start the
XY moves.

3. Dynamically adjust the head during the cut to account for material
unevenness and warpage.

4. Do that cycle each time, for each cut, because the position of the
head is not known after it starts adjusting UP and Down

The reason the commercial units cost so much is that they have to
provide all of that logic in the unit. My reasoning was that if I
could use the power of the controller PC and the software logic then
I could do several things.

1. The sensing (surface zero) could be done and maintained as
a "referencing" move in a g-code.

2. The pierce cycle could be integrated with the controller to
provide XY hold and the sequence could be simple g-code and added to
my CAM post processor.

3. The controller software would adjust the Z height on the fly and
it would "know" where the absolute position of the head was so you
could skip the re-zero move on every individual cut. That would also
provide a know distance to start the pierce and lower the head for
cutting.

4. You could set limits in the software to prevent the head for over
compensating (max and Min adjustments).

The cost then becomes just the external sensing components.


We were able to do all of those things so that now cutting material
is a smooth operation and I trust my machine enough to load a piece,
hit RUN and walk off to work on cleaning up the previous piece I just
unloaded. My CAM software POST puts in all of the g-codes so now I
just draw the design, run it through CAM to define the cut vectors
and lead-ins, and load it on the machine. The whole process takes
minutes.

Can you live without a THC....yep. You can use a drag tip and
floating head and do a lot of stuff but that has issues as well.
If you cut very much material though, you will quickly decide it (THC)
is a neccessity rather than a luxury. Could you use a different
approach? Sure, but you would be giving up some features that are
real nice to have. I find my consummables (tip & electrode) costs
are almost double when I don't do things in the right sequence and
use the THC. If you cut 3 pieces a week then NBD, but if you cut 20
a day like I do sometimes then I go through a LOT of tips.

Tom C