Back at the end of 2014, I bought a cheap
Chinese Plasma Cutter, called a Giant Cut40 from a local
trader, at £185, it seemed to be a reasonable buy, the
trader was local in case of problems, so it was a
reasonable safe purchase.
By January 2015 I was decided it was project time and I
started to acquire parts to make a CNC frame for the
plasma, Nema 23 stepper motors, timing belts and pulleys
were purchased via Ebay along with a ball screw and a 4
channel controller. The software in use at the
time on a CNC milling machine was CNCPro an old DOS
The frame slowly started to take shape over the next few
months, I used 50 x 25 mm box section.
By this time it was June 2015, so I was six months into
the project and to be honest wondering why I was
At the end of June 2015, I thought the stage had been
reached to move the frame from the floor to a bench
position and apart from blowing controllers to useless
pieces dead electronics every now and then I managed to
achieve some good cuts. I use Qcad under Linux for
my drawing package and purchased a copy of Sheetcam to
translate across to Gcode.
The following video was June 2016 so still using the dos
program CNCPro But it toom me to this stage to stop the
controllers from failing on arc start. A home made
filter system of caps Zener diodes and MOV's between the
motors and the controller did the job for the most part.
One of the issues was cutting arc start up. To
this point I was dragging the plasma tip across the
metal, that was the only way to ensure the arc
started. But I knew I had to improve things, so
that meant investing time in better software such as
Mach3 even though Mach4 was coming out soon.
I looked at how I could get some form of pilot arc from
the cut 40 so I started to experiment, I did try it in
the early days with some success but for some reason
stopped, now was time to try again.
So I proved a system that would work and improved on it
in stages. It is worth note that in 2016 and 2017
I started to see Cut 50 Plasma cutters with a pilot arc
built in, Oh well!
When I changed over to a SG55A straight torch (see later
videos) I also changed the resistors slightly to three
100 Ohm, 100 watt resistors wired in parallel. I
had a few requests for a circuit, so here is something
The next stage was to probe the top of the work piece
for Zero height. If I had pilot arc I did not need
to drag the tip any more? So a micro-switch was
fitted to the 'floating head' design I was using, wired
into the probe input of mach3 and some macro code worked
out with help from the Mach3 forum and away I went.
It was now December 2016, 2 years from buying the Plasma
cutter and nearly two years into the project of CNC'ing
the unit. It has been suggested I should have
bought a Hypertherm, in fact I still get those comments,
perhaps when I can have a table bigger then the size of
a desk I may, but for now :)
The next stage of improvement was to add Torch height
Control, THC. Mach3 has a means of taking signals
from some basic units and moving the Z axis up or down
to suit, but prices again can be expensive and would
cost me more that the Plasma cutter cost in the first
place, so a sub project was started, please see a little
Filtering on the CNC Plasma
I have had a few comments about filtering, I guess that
as people start to build their own units, like me they
have issues with the HF start on the Chinese cut
machines getting back into the control electronics and
So here is the circuit of the filter I added to the
output from the motor controller to the stepper
motors. The diagram only shows the output for one
motor/axis, so if you use 3 axis, you need 3
filters. This worked on mine, or shall I say, it
helped a lot, each machine set-up will be different, so
best experiment yourself and see what is best for your
Here is the Youtube video that accompanies the circuit.
A few repairs and a better camera!
June 2017 and I have just updated the table, for
the table it is a major update and potentially throws
all my previous cut setting out of the window! The
ball screw and number of steps per mm, meant the maximum
the table would run the Y axis at reliably was a maximum
speed of 1200 mm per minute. I would cut at 900 mm per
minute to make sure it did not miss steps.
So the Y axis was changed over to the same belt system
as X axis and I could go to a maximum speed of 7,000 mm
per minute. (275 ipm).
So new learning curve awaits as I figure out Amps
verses speed on my machine.
I have been playing again with settings and on
1.5mm steel cutting at 27 Amps at 2,200 mm/min.
Also a repair, or more like permanent fix of the Z
height probe switch as it kept moving!
So once I sorted that, added an aluminium shield to
protect the Y axis ways from the splash-back I had
another go with what I call a test piece. Apart
from desperately trying to pick out the tip-ups before
they caught on the head, this cut without any errors.
Total running time was 18 minutes 15 seconds.
Anyway this was my test piece for tonight, 450mm
I am saving my newer cutting
parameters and will update the table when I have
played a bit more.
The above mentioned Z Probe switch
meant I needed to remove the axis to sort it, so I did
a couple of videos on my way of doing the Z zero
setting. details of mechanical switch, macros and
Here is a link to the M800 file as mentioned in the
Not updated for a while, here is a simple video, just to
show it is still here and working.
Just playing, it may give ideas for the festive season,
you can draw and cut similar items.
I had a few questions about the
Pilot arc system I used, so the best way I could
think of answering is with a video or two!
I find when the time comes to
change or get a new batch of consumables I have
issues, I have been finding the quality of fit,
manufacturing etc. to be different from one batch
to another. So I am trying a different style
torch, the SG-55. It is physically bigger
and meant for more amps typically the 50 to 60
plus range, but I find you can get 40 Amp
consumables for it.
The above video is the first explanation and the
next one or possibly more will be about
use. First I have to make a mount to fit
the SG-55 to the Z-axis.
I eventually realised that
the handle part of the Sg-55A torch is 28mm
diameter and that is the same as some standard
plumbing fittings, so from the local hardware
store I picked up some plastic pipe clips the
sort that wrap all around the pipe. Took
the securing hole out to match a M6 screw
thread and I could mount directly to the from
of the Z axis. This made up for the
extra weight and then some as I could do
without the steel mount for the PT-31
torch. Wired in the Pilot wire back to
the resistors and ready for it's first go:
This was very encouraging, so time to make
something else a simple 'key' holder and
time for a second go just to make sure it
was not a fluke:
So happy with the
new torch, seems to perform better, I
believe there should be less variance
in the parts, but time will tell, it
does mean the custom end to provide
Pilot arc is not used so the air
passing the nozzle should be more
uniform and keep things a bit cooler,
perhaps helping shift the dross?
Next up, was to
help someone having difficulty with 1
mm aluminium, so a response via email,
but testing it myself should help
more. I went to 25 Amps at 5
metres per minute, just under 200 inch
per minute, think I could go a bit
faster, but ran out of aluminium for
the time being.
So that brings the page up to date
for a little while.
Oh, I am
trying to cut sheet brass 1.6mm
thick, a simple clock. I
basically used same speeds as I
would for steel, but the cut edge
not as sharp, more of a melted
edge and it seems the melted metal
can solidify causing a kerf/dross
line on the bottom edge of the cut
that is not as easy to remove.
Next is a
general video, in this I used a
free to download 3D Puzzle file
from the internet.
Dragonfly, I guess it
was originally done for 2m ply
wood or similar, but here we go in
1.5mm steel. I had a few
issues where I placed the starts
in Sheetcam in the wrong place, it
allowed the cut piece to drop out
before the cut had actually
finished, so the plasma was trying
to fire into nothing, so the HF
would stutter and try to keep the
I also forgot to make sure the
earth clip was on the table from
the 20 Ohm pilot arc
resistors. Think memory must
have been suffering a bit!
I am always
looking to see if things can be
improved, I will try things out
and see how my ideas work.
In another video I was cutting 8mm
steel, which to be honest was a
bit adventurous for this machine,
but one thing I noticed was the
dross/blowback when piercing could
get stuck on the tip or nozzle, so
a change to the nozzle by removing
the castellations was tried.
By removing the end, and not going
into the jet/arc hole I was able
to shave 0.5 mm off the tip.
Now while this may not sound like
much, it will, I hope allow me to
have a bit more clearance at
pierce height, as the new height
will be, say, 2 mm rather than 1.5
but the hole is still at the same
Also if I continue to use 1mm
cutting height then the hole has
come closer to the work from 1.5
to 1.0 mm. All by removing
the end bit of metal from the tip.
Been a while
but here is a bit on Home
switches, soft limits and G28.
I tried to answer a few questions
on the Internet, this made me
think and look into using my last
input pin for home switches, plus
I had to change some button
settings on the Mach3 screen so
here is a video about that.
I mentioned in the video I would
have the Visual basic script on
here so here it is:
' Optional to turn on my
table if not on!
' Wait for movement to
so other threads can run
while we wait
' Reference Y
' Reference X
' Reference Z
' Wait for movement to
so other threads can run
while we wait
' Go to