Emco F3 conversion - electronics and controller

[fastbike]

The next section of the schematic covers the control panel. This will follow the sketch in post #2 above with the following changes:

• an additional warning light to indicate if the stepper driver ALM output has been triggered. The drivers are CL86T from OMC StepperOnline which monitor the motors vis integrated encoders. The ALM triggers if the max current or voltage is exceeded, or the encoder determines that the motor did not move as instructed. The (previous) fault indicator is triggered by the limit switches and spindle overload.
• an extra switch added for a machine light. I'll wire this to control a 24V LED device.
• an extra switch to run the Home All macro. The software manual for Eding says it is a good idea to leave the machine in a known state before switching off so this makes it easy.
• The jog speeds are now 0.001, 0.01, and 0.1mm per press, with the Rapid override running at 1mm. So the 4 position switch has been replaced with a 3 position 60 degree switch.

Jogging
There are 4 pairs of momentary switches (S11-S17) to control jogging in the X/Y/Z/A axes. Jogging speed is controlled by S6 which provides a Rapid override, and S7-S9 giving the normal speeds.

Cooling and Aux
Flood and Mist cooling are driven by S18/S19. The feedback status of these outputs is on another page of the schematic and is driven by the output relays.
The worklight is activated by S21 - no feedback LED is required.
The Home All macro is triggered by S22.

Spindle Control
The spindle speed is selected by S24 which is connected to 3 motor contactors providing a 2 speed Dahlander arrangement (I have not changed the original spindle motor and gearbox). The nets labelled SpindleSlow and SpindleFast connect to the coils of the contactors. I'll do a separate post later explaining how that works.

The spindle is (manually) activated by S22. This energises RS1 where the first set of contacts act as a latch and the second set of contacts supplies a signal path to the controller input that controls spindle activation.
The spindle is deactivated normally by pressing S23 which de-energises the coil of RS1. The second set of contacts signals the controller that the spindle state should be stopped.
There is also a zero volt release arrangement, so if the 48V supply is lost, relay RS1 will signal to the controller that the spindle state should be stopped. This is a safety provision I added after including the limit override switch, as I did not want a stopped spindle suddenly starting if the machine was jogged back into the work envelope.

CNC Mode
The CNC mode can be set to Run or Pause via switches S25/S26. The Pause state is provided by an LED in the switch, although I have to figure out how the controller outputs that state.

[fastbike]

The Home and Probe switches are simple NC micro switches connected via a quad opto isolator and DIN style connector plugs.

There are 4 Home inputs on the controller and a Probe input.

I have yet to work through the schematic for the Probe, so am interested to hear what other people do here.

[fastbike]

Some switchgear parts have arrived and I've been testing them against the controller.

I've made a small change to the control panel setup in post #21 above. The control I'm using to select the non rapid jog speeds is a 3 position latching rotary knob from Eaton in their M22 series. Part number is M22-WRK3. This has a separate mounting plate that attaches after the actuator is fastened to the front panel and the switch mechanisms (blocks) snap into the mounting plate.

I had interpreted the datasheet to think I could snap in 3 NO blocks to drive the 0.001, 0.01, and 0.1 inputs. I turns out that the actuator only switches the 2 outside positions with the middle position being a "dead/off" position where nothing is switched.

Luckily the blocks are stackable so I've added 2x NC blocks, one on the back of each of the NO block, and wired these NC blocks in series to give a 3 position output. The relevant part of the schematic is now

[fastbike]

... so I've added 2x NC blocks, one on the back of each of the NO block, and wired these NC blocks in series to give a 3 position output.

Or I could switch the STEP001 input with a couple of diodes and a Mosfet. I will be mounting the (Limit and Home) Optoisolators on some perf board so could do the same with this. The NC switch modules are approx $22 each so a $2 Mosfet makes for a good saving.

My electronics skills are a little rusty, so am interested if anybody has got a good link for a tutorial on how to design a Mosfet switch, where the input will be driven by switches S7 and S9 in parallel. These are normally open and switch to GND when active. So I want the Mosfet to switch to GND when S7 and S9 are open and be floating (high) when they close.

[fastbike]

Back in post #15 I was pondering how to indicate when the machine is powered up - using a green LED module to illuminate the main power switch.

... Ideally I'd rather this was not being fed from the 400VAC, as it is a 230VAC part so requires extra voltage dropping which introduces thermal issues.

I've dug around on the manufacturers website, and they show an extra module M22-XLED230-T which is designed for my use case, so I will go that way.

[fastbike]

I had interpreted the datasheet to think I could snap in 3 NO blocks to drive the 0.001, 0.01, and 0.1 inputs. I turns out that the actuator only switches the 2 outside positions with the middle position being a "dead/off" position where nothing is switched.

I've found an accessory part on the manufacturer website M22-XW. From what I can see these snap into the actuator plungers and enable operation of a switch block in the centre position. I will order on to see how it works, if it does what I think the centre "0.01" step setting will only need a single NC switch.

[fastbike]

Sheet metal supplier is running a little slow, so I've down the layouts for the internals of the control enclosure. I have not modelled it in 3D, but essentially there are 4 planes with components:
- operator control panel, which has switches and status indicator lights
- power supply and main contactors, all DIN rail mounted. This is located on the front of the enclosure underneath the angled operator panel.
- floor, which has the DIN rail mounted controller and auxiliary IO board from EdingCNC
- hinged backplane (sits inside the rear access door), which has the main MCB, Dahlander spindle contactors, pump contactor, stepper drivers, and miscellaneous IO relays.
It all fits in with a little room to spare. More importantly I can now see where to run the trunking and where the 400VAC needs to be run to avoid the control cables.

And Emco came back to me with the red colour they use on their machinery: RAL3020, so time to order some paint.

[fastbike]

The control panel has evolved a little since post #2.
I've added a button to control the work light, a Home All button, and a warning light for the fault alarm from the stepper driver.
The jog speed reflects the 3 slower speeds (0.001, 0.01, 0.1 mm per press) and the rapid button advances at 1mm.
I'll print this out and mock up a panel before doing the final artwork.

[fastbike]

Another possible change to the control panel is to emulate the behaviour of the Emco FB3 power feeds in X and Y axes:

• when the spindle is off the X+/X-/Y+/Y- buttons are only active when pressed
• when the spindle is on the X+/X-/Y+/Y- buttons latch when pressed (this requires an extra feed stop button)

[fastbike]

I collected the sheet metal for the enclosure yesterday. The folding is very good apart from a folded tab at the front base which is preventing the front panel from fitting flush. I'll cut it off with a grinder rather than taking it back for alteration.

[fastbike]

And here it is welded up. Well technically not welded as I TIG brazed it with silicon bronze.
And I've done the cutout for the control panel plate, which was done after welding to ensure the box did not distort. The beauty of TIG brazing is very little heat is needed so distortion is minimal. Also a wipe down with acetone is all that is needed on the zincalume plate - no need to remove the zinc plating in the weld area. Tomorrow is hinging the rear door, and the hinged backplate. After that tack welding in the mounting plate for DIN rails etc.

[fastbike]

I've found an accessory part on the manufacturer website M22-XW. From what I can see these snap into the actuator plungers and enable operation of a switch block in the centre position. I will order on to see how it works, if it does what I think the centre "0.01" step setting will only need a single NC switch.

The parts arrived and easily snap into the 3 position rotary switch. I've fitted a NO switch block at each end, and a NC in the centre. This gives me the 3 settings for STEP0001, STEP001 and STEP01.
The Eaton M22 controls are very well designed and flexible in how they can be configured. A little spendy but that reflects the quality.

[fastbike]

Got some cabinet hinges yesterday. I've bolted one leaf to the flange at the back of the cabinet and cut the other leaf to fit the door and tacked into place (whoops - a bit much current on these thin parts).
I'll finish this off tomorrow, mount the latch and then look at fabricating a bracket for the door microswitch and the channels for the EPDM door seals.
Then I can get onto mounting the hinged backplate, and the DIN rail mounting tabs.

Photos to follow.

[fastbike]

So, I mounted the latch and the door microswitch this morning, then life got in the way. Hopefully tomorrow will see the backplate installed (and some more photos)

[fastbike]

More progress on the enclosure. Way back 3 years ago when I started this conversion a comment was made that the mechanical parts would be the hardest, with the electronics being straight forward. Hmmm, this enclosure and the electronics have taken a lot of nutting out to get the machine characteristics that I want, and to make things fit.

Anyway, here's a progress report. I mounted the enclosure door and the hinged backplane. This shows the backplane up in the normal position.


I have drilled holes for the stepper drives (room for 4 axes) and DIN rails. I've snapped on a few of the contactors, relays and the main MCB. Two views, this one with the backplane hinged down for access to the box.


And this, an internal view, from the aperture for the control panel showing the backplane in its normal position.

[fastbike]

This gives me the 3 settings for STEP0001, STEP001 and STEP01.

I hooked these switches up on the test bench to the EdingCNC UIO-10 accessory board to check how the switching works with the software.

I encountered two problems: the first is that the input for stepping 0.001mm per pulse is actually mapped to 0.01mm. Eding are looking into this and I expect a new version of the firmware will be forth coming.

The second was a bit harder to diagnose; the switch inputs were randomly being triggered. I tried a variety of wiring configurations, none of which made a difference. I have access to an oscilloscope so hooked that up to the switch wire and saw a very dirty wave form with periodic pulses. Disconnecting the power supply from the laptop running the software cleaned up the scope trace. However the laptop can only run so long on battery so I'll need to figure out a better solution, maybe some filtering or ferrites ?

[fastbike]

The second was a bit harder to diagnose; the switch inputs were randomly being triggered. I tried a variety of wiring configurations, none of which made a difference. I have access to an oscilloscope so hooked that up to the switch wire and saw a very dirty wave form with periodic pulses. Disconnecting the power supply from the laptop running the software cleaned up the scope trace. However the laptop can only run so long on battery so I'll need to figure out a better solution, maybe some filtering or ferrites ?

Eding have come back to me, so we're going to try a few configurations. In the interim I have spray primed the enclosure and door. I'm using a two pack urethane acrylic top coat in RAL3020, over an etch primer.

The backplane has been painted with a rattle can.

More photos when the painting is finished.

[fastbike]

I got a first coat of the red paint onto the enclosure and door this morning. This cannot be over-coated for 12 hours so I'll try to get the second coat done tomorrow - weather dependent as I don't have an indoor spray booth and need a reasonably windless day to paint.

Meanwhile I've started work on wiring up the backplane. I've done most of the Dahlander circuit for the main spindle, and have started the stepper drivers and auxiliary relays.
Here's a photo of the partially wired backplane.

[fastbike]

And here's the enclosure painted, actual colour has more orange in the colour than is showing on my monitor. Almost time to start installing the components inside but I'll give it a day or so to harden up a bit.

[fastbike]

I'm about half way through wiring up the control box. It's really quite time consuming, although I'm work slowly and methodically as I do not want any mistakes. I've almost finished the power supply and spindle control section (all the 400V AC stuff), so will move onto the controllers and drivers next. Here's a progress photo (taken through the aperture where the control panel will be attached.