-- StudioMill --
A guide for future maintainers of the StudioMill installed at FabLab Pembrokeshire
Contents
Introduction 3
The Machine 3
Warnings 6
Spindle Control 6
Personal Safety 6
Parallel Port 7
USB port 7
Wiring 8
Layout 10
LinuxCNC Controller 12
Sutdiomill Performance Figures 12
Spindle Control 13
Feeds and Speeds 13
Further Work 14
Introduction
At its inception, FabLab Pembrokeshire acquired the mechanical parts of a virtually unused Studiomill 4-axis educational milling machine.
The machine lacked manuals and software. The model appeared to be no longer in production. Support from the manufacturer was not forthcoming.
This manual provideds some detail about how the machine has been partially returned to service.
The current setup of the machine is described, so that continuing development and future maintenance can be carried out.
The Machine
Axes
The Studiomill hardware comprises four axes, all driven by stepper-motors and all with magnetic limit/home switches. The axes are:
- X: Up-and-down. This axis is slow. It must therefore be intended to move the spindle along the workpiece in a gradual step-by-step progress to allow a 'raster scan' style of 3D cutting.
- Y: Left-and-right. This axis is fast.
- Z: Tool in-and-out. This axis is fast. Combined with simultaneous moves of the Y-axis (while the A-axis and the X-axis are held stationary) these two axes allow a rapid etching onto the workpiece of one line of a 'raster scan' job.
- A: Workpiece rotation about an axis parallel to the X-axis. This axis is fast, but it has quite low torque. It is presumably intended to be used to rotate the workpiece into one or more orientations suitable for static machining by the X/Y/Z axes. The A-axis's low torque would suggest that it is not intended to rotate the workpiece whilst the tool is in contact. I.e. it is probably not intended for lathing type operations, where the tool is moved in-and-out while the workpiece is rotated. (A typical example of such a machine would be an airscrew copying machine.) This assumption may be wrong: careful experimentation is encouraged.
All axes except the X-axis are protected against dust ingress. The X axis is a simple length of stainless M5 studding, fully exposed to dust. This would appear to be a weak point in the design with potentially high rates of wear. However, the parts are easily and cheaply replaceable if necessary.
In addition there is the option of mounting a 5th axis, manually operated.
Spindle
A single-speed spindle is provided. It comprises a 24V DC brushed motor variously labelled as “1.2A” and “250W” (which figures clearly do not tally). The motor is geared-up via an open, narrow toothed-belt drive. The spindle itself drives a ER25 collet chuck. There is clearly a severe mismatch between the high-speed low-power motor/drive compared to the potential capabilities of the very large chuck. An extension/step-down attachment is provided which mounts a ER11 chuck into the ER25, so allowing the fitment of the kind of small tooling which better matches the motor.
The belt-drive may be another weak point in the design.
The lack of speed control limits the type of materials which can be handled. The long chuck overhang reduces the stiffness of the machine.
Containment
The entire Studiomill was originally enclosed in an acrylic[?]/aluminium case with a large door. The door has two interlock sensors, of two different types. The case clearly offers dust containment, and possibly provides some level of containment for flying broken tooling and large workpiece debris. The grade and strength of the glazing is unknown and should not be relied upon.
The case has a 100mm extract nozzle (blanked) for optional vacuum dust extraction. There are two inlet vents to allow inflow to compensate. These appear to be undersized and are further restricted by air inlet filters which are - presumably - intended to prevent dust escaping when no extraction is connected.
The door latch on the case is ineffective.
The case is fastened to the mill at multiple attachment points, with the structure of the mill providing some of the support for the sides of the case.
The general standard of assembly of the case is poor compared to the excellent quality of the mill itself. Vee-slot system and other slotted extrusions are employed the make the frame of the case but fastening methods not intended for the vee-slot system are used and many of these fasteners are misdrilled and badly aligned.
The case is difficult to remove for servicing and even with the door open the case obstructs access to the machining head and the workpiece mount. The door does not reliably latch closed.
Electronics
All the electronics for the Studiomill are enclosed in an aluminium case in the floor of the machine. The case is combined with parts of the dust containment.
The case is fitted with a small cooling fan. The outlet vents are adjacent to the fan so air-flow around the box is not optimal, but it does provide air-flow over the tight-packed array of stepper motor controllers.
Accessories
A number of accessories – packed into an aluminium case – are part of the StudioMill package.
One of the accessories is a massive 3-jaw chuck provided with both external and internal jaws. This could be mounted on the A-axis turntable using bolts* and the included register-plate – but to what end is unclear – possibly for lathing type work (see A-Axis above). Users should take care when using the heavy chuck (or indeed when working on any massively heavy workpiece) to reduce the maximum acceleration of the A-axis, so that steps are not missed or overrun due to the increased inertia.
Other accessories include multiple collets for both ER25 and ER11 chucks, spanners for the chucks and several cutters.
*The bolts – M8x55[?]mm socket-head – are missing from the accessories.
Workpiece Materials
The mill is known to have machined balsa in the past. The manufacturer claims that suitable materials are: ABS, TryCut, modelling wax, woods, resins, foams. No other information is available and it seems unlikely that ABS will cut successfully with the high-speed spindle unless great care is taken to minimise cutter CED and maintain high feedrates to prevent melting and binding.
The mill has a number of features which render it incapable of machining hard materials:
- Long spindle overhang.
- Long vertical workpiece overhang with no restraint option at the opposite end to the A-axis turntable.
- High spindle RPM.
- Limited overall rigidity.
- Magnetic limit switches (susceptible to ferrous dust).
The machine appears realistically to be capable of machining foam plastics including foam uPVC, balsa wood, and perhaps light softwoods. Experimentation is required.
Warnings
Spindle Control
The spindle is switched via one of the parallel port output pins. Some PCs leave the port outputs floating when not in use, resulting in the spindle state being indeterminate when the PC is turned on with LinuxCNC (AXIS) not running.
So if the Studiomill is switched on but LinuxCNC is not running, then the spindle may immediately start turning the moment power is applied to the mill. Or the spindle may subsequently start turning unexpectedly due to a random state-change on the parallel port.
Bear in mind, too, that LinuxCNC comes with absolutely no warranty. Software errors, power surges, hardware failure or operator mistakes may result in the spindle starting up unexpectedly.
Overall, then, there is a significant risk that the spindle will spin-up unexpectedly whenever power is applied to the Studiomill. Therefore do not work on the spindle unless the machine is powered off, especially if the PC is not both powered on and running LinuxCNC.
Personal Safety
The modified StudioMill at FabLab Pembrokeshire lacks three safety features normally found on educational CNC mills:
- Safety containment: the original glazed case has been removed, perhaps permanently. Machine operators are therefore exposed to increased risks of noise, injury from the mechanism and injury from broken cutters and debris.
- Dust extraction: without containment (and indeed even if containment were re-fitted) operators are exposed to airborne dust.
- Emergency stop: the original emergency stop which was designed to be triggered by the containment door has been removed.
In addition, the CE markings on the machine are invalidated by the new electrical work that has recently been carried out by an unregistered/unqualified person.
Therefore, this machine should be considered to be an ongoing experimental project - not a ready-to-use workshop machine - until the following are in place:
- A risk assessment.
- An emergency-stop system - for example an accessible slam-button wired into the power cable to cut mains power to the machine.
- PPE for the operator and associated notices.
- A location for the machine where it can be operated without risk to non-participants.
- A RCD-protected electrical supply and any electrical inspection required by law.
While the machine could be regarded as a useful workshop machine in the hands of an operator who understands the safety limitations, it is not presently suitable as an educational demonstrator.
Parallel Port
Users who plug a PC or other equipment into the parallel port on the Studiomill do so at their own risk. While the board is claimed to be fully opto-isolated, there remains no guarantee that the parallel port will be free of noise and potentially damaging voltage spikes.
USB port
The USB port on the Studiomill should be blanked. No equipment of any kind should be plugged into it.
Wiring
At the heart of the changes to the StudioMill is a new break-out board. A popular parallel opto-isolator board was chosen for simplicity, for compatibility with LinuxCNC and for ease of obtaining replacement if necessary.
The board is wired into the mill in the conventional manner, as shown by this diagram. Note that the various grounds are not common on the board, but are wired common externally.
Layout
The main power and control components of Studiomill are laid out as in the image, below:
Notes:
- The breakout board requires both 24V and 5V inputs. The sensors and probe will work only if the 24V supply is present.
- The breakout board's 5V requirements would normally be supplied by a USB cable from the PC (that cable would not be used for data transfer however - only for power.) In the Studiomill a 5V PSU was already fitted, so this was wired direct to the breakout board instead. (In theory, the USB port on the breakout board could now be used to draw power from the PSU for an external device. This is inadvisable and the port should instead be blocked to prevent temptation. A grommet will be provided for the purpose.)
- The 24V supply is from a L7824CP regulator. Specification is max 40V input. It is run here from the nominally 48V line which supplies the stepper motor drivers. To protect the 7842, the 48V PSU has been wound down to minimum output (40.5V). For ease of installation the 7842 has simply been packaged in heat-shrink along with its required capacitors then clamped directly into the terminals of one of the stepper drivers. Given that the 7842 is virtually idling in this application, it seems very unlikely that it will come to harm even though slightly overvolt and without heat-sink.
- As a result of reducing the 48V PSU to 40.5V, the motors offer slightly reduced torque. Initial testing suggests this is not an issue.
- There are five new debouncing circuits (see the circuit diagram) wired between the sensor inputs (P10/P11/P12/P13/P15) and GND and PC5V+. These serve to prevent detection of noisy make/breaks in the physical switches. The circuits include also the pull-up resistors for the switches. Four of the debouncing circuits are packaged on one small circuit board which is then double-bound in heat-shrink. The fifth is simply packaged in heat-shrink. The flying leads from all five are wired into the breakout board. Both packages are tucked loose between the breakout board and the case wall.
- A new relay to switch the spindle has been fitted. It is a 24V vehicle-spec item driven from the 7842's 24V output, switched via the smaller onboard relay on the breakout board. The original Studiomill controller switched the mains power to the spindle PSU. The mains power is now wired permanently-on and - for safety reasons - the replacement controller switches the spindle PSU's low-voltage output instead.
- The signals to the four stepper drivers are wired conventionally (common anode).
- The A-axis motor (turntable) has had the direction reversed, by swapping its red/black wires at the driver.
- All stepper motor drivers are set to the same current and microstepping values which are unchanged from the original manufacturer[?] settings. All employ half-current hold. The configuration switch settings are (from 1 to 8): On/On/Off/Off/Off/On/On/On. Assuming these settings correspond to typical 542-type controller settings: 5 microsteps (1000 steps per rotation); 3.0A peak current; half-current hold.
LinuxCNC Controller
A redundant PC has been configured as a LinuxCNC controller, dedicated to the Studiomill.
Should the PC fail in future, the process of re-installing LinuxCNC is quite simple:
- Download LinuxCNC from linuxcnc.org
- Install it.
- Connect to internet (via ethernet) and update to latest version.
- Either:
- Use stepconf to create a configuration, then add software for the tool touch off.
- Or - more simply - copy the configuration files available at [FabLab long-term storage share/directory/file required here].
- Connect the PC to the Studiomill with a parallel cable only. Do not connect a USB cable from the PC to the breakout board USB socket.
Sutdiomill Performance Figures
In the absence of existing configuration files to copy the following very basic essentials would be required to set up LinuxCNC. The setup of the tool touch-off is non-trivial. Help is available online but must be aggregated from the manual pages, several howto's and other forum posts.
The figures below have been obtained by experimentation and are a good starting point:
- Axis 0 (X):
- Max velocity 10mm/s
- Max acceleration 1000mm/s2
- Limits 0.0mm to 235.0mm. Home at 0mm
- Tip: Don't use the stepconf 'Run' test facility to constantly run the x-axis up-and-down because the leadscrew is not specified for constant use at high speed. The M5 leadscrew studding will rapidly overheat, bind in the nut, and cause damage.
- Axis 1 (Y):
- Max velocity 80mm/s
- Max acceleration 3600mm/s2
- Limits -80mm to +80mm. Home at -81.85mm
- Note: This axis may need directions inverting for correct axis orientation.
- Axis 2 (Z)
- Max velocity 80mm/s
- Max acceleration 3600mm/s2
- Limits 0.5mm to -148mm. Home at 0mm.
- Tip: If the min limit were 0.0mm then there would be risk of tripping the limit switch during a G53 G0 Z0 withdrawal of the tool. Hence use 0.5mm.
- Axis 3 (A):
- Max velocity 90degrees/s
- Max acceleration 3000degrees/s2
- Limits -9999 to +9999. Home at -5.6degrees
Spindle Control
Spindle control is presently simple on/off with no speed control.