Here's how it happened:
Towards the end of spring 2013 MIT Prof. Ian Hunter, my lecturer for the class 2.671 (Measurement and Instrumentation Laboratory) and director of the Bio-Instrumentation Laboratory donated to MITERS a 3-axis CNC mill.
Why, you might ask, would someone donate something as expensive as a car to a humble student shop? Well, it appears five years ago someone replaced the CMOS battery, but in doing so inadvertently reset the computer's BIOS settings. As a result, the software could not detect the machine, and it sat unusable for five years. Instead of getting it working, they bought a HAAS instead, leaving the task of getting the machine back up and running to us.
It's a Dyna-Myte 1007, manufactured by an old Californian company Dyna Mechntronics. It's named 1007 for having 10" of travel in X and 7" of travel in Y (There are also 10" of travel in Z), so it has a decent build area.
Other than the 10"x7"x10" build area, this machine sports many other great features like an external flood coolant system and filter, a pneumatic automatic toolchanger with a tool carousel that can house up to 6 tools, an extra servo drive for commanding a fourth rotary axis (Prof. Hunter wanted the original rotary axis, which came from his personal home machine shop, back) and a built-in computer running DOS with decent proprietary Dyna software.
Now if we can only get it working...
Behold, the computer system of the Ancients! Sporting a whopping 250MHz Pentium CPU, this machine is weaker and takes up more room than a $35 Raspberry Pi. The built-in software running on DOS was not bad, however. It seems capable of doing everything your standard CNC software (like Bridgeport EZTrak, which I am accustomed to) can do with some additional features.
Upon powering on and loading the Dyna4M software, we were greeted by an instant "Error 465: No Response From Driver."
Well, let's take a look at the inside of that computer and make sure everything it wired correctly. The interface card between the computer and the large driver board behind the machine is an ISA card, which we later found out takes up the RAM space of COM port 2.
After a couple days of digging through the Dyna 4M software to see if we can expose the internal software configuration, we came upon this. It would appear IRQ 3 is the hardware interrupt resource that the Dyna 4M software looks for to talk to the ISA interface card. Issue is, IRQ3 is currently occupied by Serial Port COM 2.
To alleviate this, we went into the BIOS and changed all the settings involving IRQ 3. We made sure to disable COM 2's attachment to IRQ 3, as well as any other setting that may occupy the IRQ 3 memory space.
And it worked! Freeing up IRQ 3 allowed the ISA Card to subscrible the to the proper memory space, and the Dyna 4M software is able to communicate with it!
YES! We can move axes now!However, when we tried to do anything beyond that, particularly involving the pneumatic toolchanger or spindle, the program threw errors. The former was to be expected: we didn't have air hooked up to the toolchanger to allow it to move!
We hooked up an air compressor set to 90PSI to the machine, but it turns out there was a severely leaky solenoid valve preventing the system from working properly. We ordered a new set to alleviate the issue. After replacing the set and hooking up all the hoses to their proper spaces, the fault was cleared!
But led to another one. Great. Now the spindle wasn't working.
Taking a look in the back of the machine at the spindle driver, it seems there are lots of things that can go wrong. In fact, I count 18 DIP switches, all of which could be incorrectly placed.
It also seems some of the wires may have come loose from... overzealous rewiring on our part, in an attempt to fix the machine.
Looking into the back of the machine, I see the spindle driver is a Glentek SMA8115 Brushless servo driver. A quick Google search, and I am browsing the manual for DIP switch settings. Turns out, the default hall sensor configuration for this driver is 120/240 degrees, but the spindle motor has 60/300 degree hall sensors. That got rid of the hall sensor fault, but there was still a drive enable fault. A bit more digging through the manual of the spindle driver and the axis drivers (also by Glentek, but smaller models), and I find that the spindle driver has an active disable (turn a pin to HIGH to deactivate), rather than the axis drivers and their active enable (turn a pin to HIGH to activate).
As a result, the wiring is EVER so slightly different. Clever girl... see that subtle swap between the green and orange wires? That lets the regular axis interface card output talk to the special snowflake spindle driver. After changing the cable to the proper one, IT WORKS! IT WORKS IT WORKS! NO FAULTS! TIME TO CNC ALL THE THINGS!!!!!!!!
The following will be a set of photos with no dialogue. Just imagine this song playing in the background as you browse through the following images. It's almost like art.