Butchering the chip tray - mounting the drag chain tower

The drag chain tower (aka the base for the fixed end of the drag chain) is a fairly simple concept - a lid holds the chain in place, as well as sealing the tower. The channel down the middle of the tower is big enough for me to drop the encoder and limit switch connectors through without needing to disassemble them. I should be able to run those cables without majorly dismantling all the neighbouring components.

Some dimensions, for when I come to butcher the machine:

Let's try a mockup:

And check the movement (450mm) will be accommodated without clashing:

Mark out the positions of the holes:

After some stich drilling and die grinder action, it's done. So I then got a bit carried away and cleaned away the swarf and gunk from the chip tray. I'm not about to paint the thing but this seems like a good time to get rid of that nasty muck.

From the front as well:

The machine is old and tired but having stripped it back, adjusted the myriad gib strips and fitted the Linuxcnc system with closed loop servos, it should be good enough for the likes of me. And it can do stuff the Tree CNC lathe can't, like faceplate work and larger, odder shaped workpieces. The Tree was really intended for repetitive machining of bar stock, rather than jobbing work.

When the drag chain tower is done (tomorrow), I will assemble the cables, drag chain etc and hopefully move on to mounting the cabinet and routing the cables in a semi professional fashion.

Designing and printing the drag chain tower - and testing various filaments

I have some 3.4mm diameter o-ring cord(?) - the groove dims shd be around 4.5mm x 2.5mm for about 25% compression.

And I shouldn't forget to put some tiny o-rings around the 4 fixing screws. You might argue the screws would be better placed within the large o-ring but that's where I ended up.

Printing this in PA-CF (carbon fibre loaded nylon) didn't look like an option, having tried to print out some of the drag chain covers using the stuff. Apart from the surface texture (rough as a bear's arse), the things snapped along the layer lines without much provocation. They were also pretty feeble in terms of stiffness - all in all a complete disappointment. No amount of fiddling with the slicer settings is going to help. Perhaps it's best suited to large, structural parts.

It fits - but distorts if you so much look at it. And it looks (and feels) shit:

Let's try some PETG and see if that's any better. I'm rather hoping for something that is reasonably rigid, with a clean finish and possibly even a degree of solvent (coolant and way oil) resistance.

Well that is shit - but in a different way to the PA-CF.

Let's try the Creality ABS and see if that's any better:

Well, it looks better - but isn't a whole lot stronger (snaps in half easily). I may simply revert to the PLA, which actually worked well, if you can forgive the fact it's grey, not black. I could always get a reel of black of course.....

However, for the drag chain tower, it may be well suited, given that strength isn't a big issue there.

So, back to the drag chain tower. Let's print that out in ABS and be done with this part of the task. ABS is more fussy when it comes to printer settings, requiring a heated bed, hot chamber and various additions such as extended bed, multiple base layers etc to prevent it lifting around the base.

Well, this looked OK to start with:

The base is lifted at the edges ie it won't sit flat on the table. NVM, I will bugger about with the settings - but in the meantime, let's check out the other features. Those M5 holes for the lid seem to tap out ok:

The lid(s) are both badly warped.

This is with a raft, 5 layers at the base, hot bed, fan almost off. I stopped the print to check how it looked. 

This is a lot better. Pretty decent finish and no warping at the edges. I will now print the tower, noting that it estimates about 10 hours. Ideally it will be ready in the morning then.

Completing the drag chain installation on the Bantam

I need some reminders how I'd intended to install the drag chain for the carriage electrical (and coolant?) connections on the Bantam.

Luckily I recall making a CAD model of the machine. This wasn't simply a vanity or boredom activity - I actually used it to design the various components used in converting it to CNC. One advantage of not having changed CAD systems for many years is that I can still access and mess with the work I did before my memory's horizon!

Here's the latest version of the complete machine:

Ooof, that's not right. I'd forgotten I'd actually continued to evolve the design so that the Z axis ballscrew and linear encoder were repositioned at the rear of the machine.

This is actually what I've got standing in the workshop right now.

And yes, the cross slide, carriage ad drag chain all articulate. Arguably I could do with finishing off the appearances (ie get the green paint brush out) but that is just superficial.

The original aim here was to find out what I had planned in terms of the drag chain. That is becoming a bit more obvious and the clouds of memory are starting to clear gradually.

I remounted this stainless steel cover plate that I seem to recall was intended to provide a mount for the travelling end of the drag chain. Sure enough, the drag chain fixings line up nicely.

With a few links fitted, it emerges at the rear of the X axis assembly.

And in the CAD assembly, it seems I had intended the fixed end to be mounted somewhere around 19mm from the end of the bed. This is with the carriage at the furthest extent from the headstock (at the home position, FWIW):

The height of the fixed end looks about right, as the separation of the top and bottom mounts measures about 85-90mm. 

That places the bottom side of the mounting bracket around 88mm from the bottom of the chip tray.

I'm measuring about 200mm from the bottom of the top bracket to the chip tray.

Looks as if the height of the fixed end is fairly clear. And the horizontal position (along the Z axis). So let's design some sort of base to locate the base on. Perhaps also forming a degree of shielding of the cable from swarf. Next time....

Got a few minutes on my hands later on and did the paint job:

Looks better.

IGUS energy chain modelling and printing

I'm planning to finalise the control cabinet for the Bantam lathe. This will involve hanging it on the tailstock end of the machine, using a couple of holes I've already tapped on the machine stand that line up with mounting holes on the cabinet.

To do this, I need to route the various control and motor cables in a vaguely professional manner. Previously I made a start on this but it was such a long time ago I've almost completely forgotten what I had in mind. What I do recall is messing about with some drag chain and at one point I even modelled it in Fusion as a live (movable) assembly. Nobody knows why I did that but perhaps I was just bored at the time.

Here's a previous post that describes that modelling work. And it seems I ordered some R48 series "energy cable" https://www.igus.co.uk/product/series-R48 For some reason I also got some smaller drag chain - what was all that about?

Here's the smaller stuff:

Looks the business - it's from IGUS (German), which is rather expensive but s "proper" industrial stuff.

On the face of it, this would be a sufficient length to accommodate the carriage movement. That's the only need for drag chain on this machine.

But there's one snag with this smaller stuff - it's open at the outside, which wouldn't be ideal when we are generating swarf and moving the drag chain back and forth endlessly.

Which explains why I (subsequently) bought the larger drag chain. This is the R48 model and the distinguishing feature of this version is that it is fully enclosed and is designed to withstand swarf - or at least allow reduced ingress. It has a removeable cap on the "outside" of the chain and the other 3 sides are closed, notwithstanding the need for articulation.

This shows how the covers are removed:

Like many industrial products these days, 3D CAD models are provided. But on closer inspection, it becomes clear that the components are not fully representative of the shipped products. Or to put it another way, if you were to print out those models, you'd have a hard time trying to remove the covers. 

I guess this is to stop (or slow down) the Chinese cloners. Why would you want to make your own parts when you already have the stuff in your hands? More on that later....

I've moved some of the furniture around so that I can get to the back of the machine - "let the dog see the rabbit" etc.

I need to remove the splash guard thing:

You can see the X axis servo motor here - the cables are just sitting in the swarf tray. The job got abandoned before it was finished.

With the rear guard removed, access is a bit more reasonable.

Here's the R48 drag chain:

This is one of the brackets that are required to secure each end of the chain:

That cover snaps on and off so that you can run the cables.

There's a shorter one that fits on the bracket itself

The 2 different versions are almost identical, apart from the curved extension which accommodates the articulation without allowing swarf ingress.

Those tabs on the bottom engage with these locking ramps on the main body.

Here's the problem - it's clear that The Stupid Fat Bloke removed ALL of the covers when we were originally messing with the drag chain. In the meantime, he obviously put them somewhere "safe". And now I have a length of drag chain without any of its covers. That is chocolate teapot territory. 

After spending the best part of an hour going through the storage in the main garage area (and finding nothing), I'm left with just this rack to save my bacon. Can it be in there please?

No chance. Another 20 minutes of my life has gone - and still no covers. So onto Plan B - 3D print some, so that the originals are then provoked into reappearing once they are complete.

So that requires me to model them up in Fusion. Luckily I have one solitary example tat The Stupid Fat Bloke overlooked. Let's get it done:

Print it out with 100% infill. Well that looks really shit:

But the fit is almost there.

It has a support raft but the issue is really that the layers show up when you have such a small angle to the table. 

The solution is fairly obvious - flip it on its end and use tree supports to avoid sagging of critical faces:

Like this:

That's better:

A couple of minor adjustments to "optimise" the fit and we have a solution. This is absolutely spot on - no slop or misalignment - clicks into place just like the genuine part:

I've ordered some black PA-CF filament (PA6 Nylon with carbon fibre) which should be here tomorrow. I may need to fiddle with the the settings, having changed from this grey (unfilled) PLA but otherwise, I should be able to print out a load of these.