Fusion 360 post processor handling of internal and external turret tools

What up?

Phil Trueman posted a query on the Centroid forum about rear mounted tooling. His machine is very similar to mine in having a turret that requires CCW spindle direction for external tools. That got me thinking, as I had no idea how this is handled, yet it's something I'll need to bottom out myself sooner than later.

The turret in my machine presents the external tools from the rear (ie X coordinates are negative) and expects the spindle to rotate CCW when machining. Internal tools such as boring bars move in positive X coordinates ie they operate "in front of" the spindle with CW spindle direction. Not all turrets are like this - it depends if the tool is "upside down" in the turret. In my case, external tools are mounted "the right way up" like this:

If the tools were mounted "upside down", the spindle would turn CW. The reason for NOT going this route is that the cutting forces would tend to lift the turret / cross slide / carriage. Some machines are designed for CW spindle of course - but not this one.

One downside of this scheme is that the tool tip needs to be presented 3/4" above the turret face. That's 19mm, so a standard 20mm toolholder will present the tip 1mm above centre height. Sure enough, when I measure the height of the toolholders that came with it, they have metric 20x20 shanks that have been ground down to 19mm height. Something to bear in mind when I come to fitting tools to the turret.

The other slight downside is that the "handing" of the toolholders is flipped by moving them behind the spindle. So the default tool for RH turning in this scheme is actually a conventional LH part.

I struggled to figure this business out, as it's not something I've encountered before. The Bantam doesn't have a turret, so all machining is with CW spindle direction and +ve X coordinates.

Show me!
I managed to get the tool set up correctly and the spindle direction set to CCW within the tool library definition for the threading tool (as used in my first thrading trial). But no matter what I tried, I couldn't get the tool to reside on the -X side of the spindle. Note the X axis direction shown here - flipping it within the operation setup options doesn't help, as it gets realigned when you regenerate the toolpath. Hmmm.

What's the answer, fatty?

Turns out the X coordinates is flipped within the post processor. Remember this message?

Here's where that critical content gets used:

This is also where you get the spindle direction and the "hand" of the toolholder.

But finally, you get to post the g-code with the correct (negative) X coordinates. That's the only difference between the default "turret 0" and "turret 103". 

There doesn't seem to be any change in the way the tool is called up when you compare g-code generated with -0 and -103 turret options, so hopefully the "T6600" command is all we need to operate the turret. Of course, I've not got anywhere near trying this for real yet but hopefully it will work out when I finally get there....

Butchering the frame and reassembling the headstock enclosure.

Firstly, the plate around the spindle needs to be cleaned up and fitted.

A bit of WD40 and a white pot scrubber does the trick.

And some hybrid adhesive sealer to fill the gap and hold it in place.

There are four M6 holes under the spindle nose These are good for holding it in place.

There's a slight gap, perhaps due to the draft angle of the casting or simply distortion of the plate.

Needs holding in place while it cures 

Now for some angle grinder action. First, cover everything with dust sheets.

Then mark out the frame, earmuffs on and set to it.

There was a lot of stress in the frame, so when I cut the meddle member, it sprung apart.

Finally - nobody died and no broken disks.

It cleans up with WD40 nicely.

The engine crane is handy here for holding the framework in position.

One of the crucial bolts was snapped off, so a cutting disk in the angle grinder came to the rescue.

There we go. Looks a bit odd but should do the trick.

Lets offer up the headstock cover.

Looks workable.

I need to add a couple of fixings at the foot of the housing.

Like this

And a couple of M6 holes for the fixings.

There. 

Now I need to figure out where the various compts will fit and how I'll wire them up.

How to reassemble the machine enclosure, panels, controls etc?

I need to start planning how I will reassemble the machine. I've removed the original (massive) control cabinet and RSJ framework. I'd like to be able to reassemble everything apart from the main enclosure and get all the control wiring installed and commissioned before refitting it. Given that the main enclosure provided some of the structural support for the casing around the headstock, I may need to alter the way some of these elements are secured. 

Here's the headstock cover, which also housed the display, keyboard and panel switches. There's not much holding it in place, as the fixings were all made to the main enclosure originally. Offering it up loosely in place with the engine crane, we can see what it needs to look like. That plate with the large hole fits around the spindle but has no fixings. If I can secure it to the headstock, I will have something to work with. Also, the foot of the assembly could be secured to the front lip of the bed casting - that's what's done for the main enclosure. That would provide a decent mounting for the front of the headstock cover.

But at the back, there's no longer anything to mount it to. There's a cover panel at the base that simply drops down onto 4 screws and hides the spindle motor and hydraulic pump.

Here's the massive framework that used to support the control cabinet. There are already holes to mount it to the bed casting and pick up the headstock cover. Also, the hinged door / cover thing that gives access to the spindle drive area needs something to attach to. You can just about make out the fixings for the hinge on the top left of the frame.

That's this thing:

Here's a photo of the machine before I dismantled it. 

I'm now thinking I may take an angle grinder to the framework and liberate the section that supports those 3 elements. It would account for about 1/3 of the mass of the original frame but that would still be a big saving. Besides, if I don't reuse it, I'll have to fabricate something that will look remarkably similar.

As for the controls, once the various covers etc are back in place, I'd like to be able to make use of the cavernous space within to house the control electronics.

The control panel area can be used to house a 14" touchscreen (328mm width available) and some actual switches. Then the Acorn, expansion board and PSU could fit in the cavity at the top. The servo drives and 2 VFDs should be able to find themselves some space in The Cavernous Cavity.

Firstly I'll see about butchering the framework and refitting the cabinet. I'll aim to reassemble most of the panels apart from the main enclosure itself, then look at how to situate and connect up the electronics around that.

Refitting the toolchanger turret assembly

Things are a bit tight in the workshop at the moment. It was bad enough before the Tree lathe turned up. Now it's pretty tricky moving the crane but luckily I can move it enough to be able to hoist the toolchanger into position.

Almost there. There are a couple of locating dowel pins underneath, so once the piece of wood is out of the way, it will be positively located 

4 large bolts secure it to the cross slide. And I've reconnected the switch wiring to the umbilical. The cable is routed up the drag chain along with the various hydraulic hoses etc.

Talking of which, here are the 3 hydraulic cables. The middle one is simply the coolant hose, which terminates at the manifold block on top of the turret housing.

The umbilical cable doesn't seem to have anything to mount onto. Looking back at the pictures I took when dismantling, I can't see anything missing. I'm guessing somebody has been in here before and removed a bracket of some form. Cable ties it is, then.

Here's the other end of the umbilical. The boot ferrules are not of the same vintage as the original wiring, so clearly the controls have been changed at some point.

That will do for now. I won't be able to test it any further until I've got some of the electronics installed, such as the main Acorn controller and the VFD that will be used to drive the hydraulic pump.

Turret rotary switch and prepping the tool changer assembly for refitting

This is the rotary switch. It's by Sanyo, part number RS-04068CA and although I can find refences to it on the internet, I can't find any datasheets as such, no matter where I look. No problem - it's a fairly simple device with a common connection and 8 outputs, suggesting that it's purely passive (ie no semiconductors). The shaft spins very freely and wants to end up in any of 8 positions. Clearly there's some sort of magnetic content in there. I'm guessing there are 8 reed switches but I don't need to know. Sure enough, when you spin the input, the switches close one after the other.

And the adjustment slots allow for a wide range of movement. However, there are marks where it was previously locked. With position 3 in the turret in the horizontal position (ie where it would be in use), the "#3" switch is closed. That all seems to add up.

There's also a couple of positions on the connector block for the microswitch that indicates that the turret has successfully locked.

The microswitch simply acts against the face of the driven gear. This moves axially during locking.

The coolant hose block mounts on the top of the housing.

I've removed the switch and terminal block for the moment, as they are rather delicate. Now, with a sling fitted, the engine crane can reach over the TIG welder and pick it up. Tomorrow I'll refit it to the cross slide.

Final(?) reassembly of the tool changer turret and testing with compressed air - and spindle encoder bearings:

Firstly, finish machining the LH threaded MF18 locknut for the turret spindle on the bantam CNC lathe.

Flip it round, shorten it to size and do a bit of chamfering here with the spindle in reverse:

It seems to fit nicely. It's no longer flush like the old castle nut but I can't be arsed to mill 3 slots in it at this stage. If it becomes advantageous to make the nut sit flush, I can always have another go at it later.

Here we go. The turret is back in its rightful place at last.

The bench is a mess but we are almost there now.

Testing with compressed air:

With the solenoid wired up to a 110V site transformer, I can activate the solenoid manually and supply compressed air, to (hopefully) actuate the various components.

Here it is in action. Normally the solenoid would be controlled by the Acorn board which will be watching the status of the 8 way position switch but it's not connected up yet, so I will have to simulate the operation the best I can.

That went well. There's just about enough pressure (4 bar) to drive the turret off its locating pins and start to spin the hydraulic motor. Finally, with the solenoid deactivated, the air is just about sufficient to drive the turret back onto its locking pins. I needed to position the turret at the correct angular position for that to happen, which is why the detent plate and piston is provided. 

Operation is pretty simple, only requiring a single signal (to the solenoid). Once the turret is in the correct angular position, the solenoid can simply be de-energised and the turret will be locked in place.

Obvs it's not operating at anything like the final pressure but from what I can tell, it's giving the impression of functioning as I'd expect.

Perhaps time to look at the 8 position switch now, then think about refitting the turret to the machine. Or would that be premature at this stage?

Spindle encoder:

Although it's not directly related to the turret, the spindle encoder got some attention today, mainly because it had been sitting on the dirty bench, occupying the Evaporust tray. It's now fully dismantled and cleaned up. However, I need new bearings, as the old ones make a fine old noise when spun. The encoder assembly is beautifully overengineered for what it does and was full of grease and muck. These bearings are std 6004Z shielded deep groove jobbies, so I'll replace them with 6004-2RS (double sealed) bearings. When I'm at it, I'll also get some replacements for the belt tensioner jockey pulley. These are 6207Z shielded deep groove bearings, so these will be replaced with 6207-2RD (double sealed) bearings. They are not expensive, so no point paying P&P just to have some shitty skateboard bearings and I don't plan to revisit them any time soon.

I should get these Tuesday-Wednesday (it's Easter bank holiday weekend).