CNC Conversion
Last Updated: 10th January 2010
Shortly after acquiring my laser cutter I also bought a Sieg X1 Micro Mill with the thought that I might convert it to use CNC. While researching the options I found something on the RoutOutCNC forum about how they'd also converted a laser cutter, similar to mine, over to CNC using Mach 3 to control it.
I wasn't familiar with Mach 3 at the time but it seemed to be regarded, amongst home/hobby machinists at least, as the software to use with regards to CNC. Going off the instructional videos on their website it seemed very capable and, having almost decided that's what I wanted for the micro mill, the discovery that I could use the same controller boards and software for the laser made it too tempting to resist.
Mach 3, breakout boards, controller boards, and stepper motors were purchased from RoutOutCNC, while hardware for mounting the motors to the mill came from CNC Fusion.
The Plan
My initial plan was to convert the mill and have a good play with Mach 3 before making a final decision about how to convert the laser cutter. This was because I was thinking I might do it such that I could throw a few switches to choose between the original set up with NewlyDRAW and the new one with Mach 3. However this changed after I killed the original controller board.
I'm not 100% sure what happened to it but the fact is that I took the machine apart a couple of times to assess the possibility of increasing the bed size. I'm pretty careful with this stuff, anti-static precautions and all that. But the fact is that it worked before, but not after, one of my exploratory dismantle/reassemble operations does kinda suggest that I killed it.
It didn't make any sense for me to spend the 91 UKP it would cost for a new one so converting the laser was brought forward with with the new stuff simply replacing the old.
Software Installation
Installation of Mach 3 onto the same PC laptop I used for NewlyDRAW was straightforward; as per the videos on the Mach 3 website. However, when I did the test (see the video) my waveform was horrible and nothing like their nice clean one. I found the solution in a post on the Mach 3 forum where DaveC says:
The secret is .... disable the ACPI driver!
Open the device manager and find the 'Computer' icon.
Click on the plus next to that and you will see something like Advanced Configuration and Power Interface (ACPI) computer. Chose the 'update driver' option and select 'Standard PC' as a replacement.
Windows demands a reboot after this, and a lot of the hardware had to be 'rediscovered' and reinstalled (all used drivers it knew about already, so was painless), another reboot for good measure and now it works flawlessly.
Thanks Dave.
Motors & Switches
24v for the motors is easily obtained from the 5v/24v PSU board. On mine the connections we want are the middle two marked marked GND and 24V.
While we're here:
L and 5V are for the optocoupler that we can also see in the image. We won't be using this for reasons I'll explain later.
VGND and 5V provided a 5v supply to the old controller board. Again, we won't be using this for anything.
The original motors and home switches have 4 wires coming off one motor, and a flexible flat cable (FFC) coming off the other. FFC connectors are a bit tricky to track down so I decided it make more sense to replace most of this.
In fact the 4 wires coming off the Y-axis motor are plenty long enough to reach the terminals on the new controller. The wires coming of the X-axis motor need extending. Using a continuity tester it's easy to identify the two pairs in each group of 4. Apparently it doesn't make any different which are designated pair 1 or pair 2, or which is designated A and B within the pair.
As per the information at RoutOutCNC I set the controllers to 1 amp and 8th step. However I found that the motors tended to be noisy when stationary and the X-axis motor ran hot. I reduced the current to 0.5 amp, which seems plenty and resulted in cooler, quieter, motors.
I have purchased some micro switches to be installed as home switches but have not done this at the time of writing this. Watch this space.
Controlling the Laser Beam
The info at RoutOutCNC describes a pair of wires on the laser power board that are connected to the breakout board to give control of the beam, however my board is different. The image below shows the likely options. Note that I've placed a mirror under the board to show the tracks on the back of the PCB.
U1 is a P521 opto-coupler with the input pins connected to the red and yellow wires on J4. These are wires that I have added. More about them later.
The other side of U1 also connects to the yellow/black pair, and the black/black pair on J5. The yellow black pair goes to the power board that supplies the 24v for the motors and which also supplied the old controller with its 5v supply. This board also has an optocoupler to whose output these wire are connected (the input was connected to the old controller board). The black/black pair goes to the "Fire Laser" button on the control panel (used to test fire the laser).
The green pair on J4 go through the safety switch that prevents the laser from firing when the canopy is open.
The yellow/black pair on J5 looks temptingly like the black/white pair described on RoutOutCNC, but don't be fooled. They are on the 'wrong' side of the optocoupler and using these would connect the laser circuitry direct to the computer with no optocoupling. The voltage between the pins to which the yellow and black wires connect is a sweet 5v, but measure between them and earth, or ground on our breakout board, and we're looking at something more like 120v.
The optocoupler at the other end of the black/yellow pair (on the 5v/24v PSU board) seemed like another option but alas the input side is connected to the laser cutters 5v supply. Once again, it looks great when measured on its own but compare it with ground on our breakout board and things start to look unfriendly. Even if this were not the case we would still, to some extent, be bypassing the optoisolation were we to make use of it.
Now as I already mentioned, the red/yellow pair were added by me and as you can see, before they were added, optocoupler U1 was sat there doing nothing at all. There is much evidence that boards in this machine were designed for use in various machines, with different requirements, and this unemployed optocoupler is just one example; one that provides an ideal connection via which we can control the laser.
At this point however I ran into problems because I tried to use pin 1 of the parallel port to control it. Why I tried that, why that didn't work, and how I eventually found my solution is described, for those who are interested, in a document that I have sarcastically entitled Fun With Pin 1. All you really need to know however is that my successful solution involves connecting the red wire (that I added) to GND on the parallel port while the yellow wire connects to pin 6. Mach 3 is then set to use pin 6 as the control for the 'spindle'. I repeat: pin 6 and NOT pin 1. Read Fun With Pin 1 if you have any thoughts about using it.
In Use
The image to the right shows my first test with the new system and Mach 3 users will recognise it as being from the roadrunner G-code included with Mach 3. Things to note are:
I have pin 6 on the breakout board set to behave as if it were controlling a spindle relay. As the roadrunner G-code is written for a mill, the spindle is switched on at the start and left running until the end. Consequently we have lines cut where the head is moving from the end of one chain to another. We also have holes being cut whenever the mill's cutter would be being moved in the Z-axis.
While cutting I was twiddling the power knob on the control panel of the laser cutter (which is why we have lines of different 'thickness'.
Clearly I now need to invest some time in learning the ins and outs of LazyCAM (supplied with Mach 3) and G-code as it relates to my laser cutter.
Once again: watch this space.