Re: CNC PLASMA - THC schematic ?

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

>
> > Yes. I guess I need to move it over. While the basic voltage
> > sensing circuit of the posted schematic remains the same we have
> made
> > several additions and modifications to the version that will be
> > available from Bob Campbell. We have added a universal current
> > sensor for the plasma side that works with any machine for

the "Arc

> > Good" signal. Jim at Sound Logic is doing the design and has

added

> > the buffers and isolation for the second parallel port that is
> > usually need for the added inputs.
>
> Wow,
>
> The schematic looks good, I have an earlier verision that is

similar

> but with only 2 channels of comparitor. I think that was from

when

> the discussion first came up there.
>
> I was looking at a simple DC motor instead of a stepper as the
> control will be open loop for posistion, but closed loop on arc
> voltage.
>
> will any of the mods be available for the group ?
>
>
> Dave

The circuit was designed early on to integrate with MACH2. In looking
at designs of stand alone units I realized they had to have some
rather complex logic circuits to be able to do all the things a good
THC should. It does more than just adjust the head as it goes
along. It needs to sense the arc current and hold program execution
until you get a good arc on pierce. If not you will end up with gaps
in your cuts. On start,you can program in a delay, but often my
torch (Gertie has a mind of her own) will turn off its purge air
during a long rapid and the start process takes several seconds under
those conditions. If you start and pause long enough to cover all
conditions, on normal starts the torch cuts a larger and larger divit
until it starts to move.

The biggest issue was to keep the start height and pierce height
constant. There are several ways to do this. One used on stand-
alone THC's drops a sensor down to "feel" for the surface of the
material and then zero's itself so it knows to start say at .2 inches
and then plunge to .135 BEFORE the actual THC Up and Down control
takes over. At the end of each cut you never know where you are so
you have to go through the touch and re-zero on each cut. No big
deal when there are a dozen cuts but a PITA when you have close to a
hundred! Then there is the issue of the sensor landing in a void
that has already been cut (common on close cuts) or on a piece of
surface slag.

I reasoned that all of that logic could be handled by the PC and the
controller program instead of a separate microprocessor in the THC.
If the controller program controlled the up and down it would always
know where the tip was in relation to the material. It could keep
track of the up and down movements. It could also sense the Arc Good
condition and hold the movement until it was valid. It also has
software excursion limits to prevent over travel.

Art was able to add the features to MACH2 and using a much simplier
external circuit it handles the logic of the THC. It has to have
control over the Z axis which means it needs to be a Step and
Direction control. I used a simple stepper and a Gecko but a low
cost unipolar or bipolar stepper driver with a small 23 frame stepper
will work great. The speed required for THC is only about 20% of the
normal rapid on that axis.

You could in theory take the output of my circuit and control an H
bridge to move the DC motor up and down. You would need to provide
safety limit switches in case the voltage started to get out of range
and the motor moved too far. You would still have to design the
surface sensing and move to pierce height circuits. If your torch
has an arc good signal then you would be all set; if not you REALLY
need to design in that feature or spend a lot of frustrating time
finishing cuts by hand.