THE CNC PROJECT

Part 2

DUST AND MORE DUST

All milling machines make dust – or swarf, or shavings or whatever you want to call it, but the faster they work and the better they cut, the more dust they make and you better have a way to clear that away for your milling head or it will clog up, overheat and maybe even break a tip.

So you need a dust collection system.

Yes, you could just get a shopvac but that won’t cut it: once the vac bag gets about half full its suction drops off and you are forever going to be buying new dust bags, and this thing is going to be sucking a lot of dust over long periods.

What works so much better is a cyclone filter. This is a specially shaped chamber that causes a “cyclone” in the air that gets the particles in the air to drop out. This goes in front of the vac so that the air being sucked into the vac is almost clean while all of the dust is easily collected out the bottom of the cyclone.

At first I thought of putting the whole thing under the CNC but this is a very useful thing to have: almost every workshop power tool makes dust and they have a collection port that you can hook up the vac to, so why not have it on a mobile trolley for general use?

The cyclone cost $32. The Shop Vac was $100. Other than that there were some bits of 50mm PVC plumbing and a metal 20L drum I found for free at my work. I had a trolley that I got for some other use that just happened to be exactly big enough to fit the shop vac on the bottom shelf. I estimate the total cost at $180.

Just have a look at how much the fancy shop vacs go for and note that these do not usually have cyclones on them: you will be buying a lot of dust bags if you want to use them for a CNC mill.

The big pipe has a threaded joiner in the middle that acts as a swivel so that the top can be removed from the cyclone (the black bit) and the drum has a clamp on it that can be undone to get the dust emptied. The fitting on the bottom that locks onto the vac has a swivel that is airtight and has slots that lock on the pins on the vac body so you can remove the bits above it easily to empty the vac itself (although this won’t be often).

There are other neat details but the main thing is that it all works well. See the picture below.

As a project this worked out really well. It made me think and it was good practice putting various parts together to make something that works properly.

It also kept me going while I waited for parts of the CNC to arrive.  There was a lot of waiting involved.

 

THE CNC PROJECT

Part 1

Overview

What I really want is to make nice clean things that work because the parts all fit together neatly. I want to make some spinning motors and such in future and that needs high quality accurate machining. It will take time to develop, yes, but the results will be usable as opposed to rough-as-guts hand manufacture. I know that only too well.

The best way to do this is with a CNC Mill. That’s Computer Numerical Control. When I worked in a machine shop as a teenager the CNCs were big and expensive. We cut stainless steel like it was cheese.

Nowadays there are small, cheap and low power models that you can run from your garage. This is defined as subtractive machining – you take a lump of material and subtract the bits you don’t want from it. The opposite is additive machining, or 3D printing. I have one of those too – they are good but the parts it makes cannot equal the strength of a machined aluminium part. Having both in your workshop is the best option.

I looked around the net and after careful thought I decided on the OneFinity Woodworker X-50 32 x 32 inch. It cost about $5K AUD including delivery from Canada.

Why did I choose the OneFinity?

1. Ball screws. It has Ball screws which are solid unlike a lot of other machines which have belts. Belts are fine for machinery that does not take a lot of force such as 3D printers, but for milling they are just too weak imo.

2. Solid construction with linear bearings. This is the only home CNC that has decent rails and lead screws.

3. Flexibility with the choice of machine head and software.

The maker does not recommend using it for cutting steel and that is fine. They also don’t supply a milling head or drag chains and that’s fine, I will get my own. They supply the three motion axes, the controller box for them and the control panel. There are a few extras that are worth getting from OneFinity such as the touch probe but everything else you get for yourself and assemble yourself.

So now there is a whole lot of work to get all of the parts together and assemble them into a working machine - metalwork, woodwork, electronics, plastic work, painting, design . . . . there’s all of them needed and a few other things too. Then once it is all working I need to work out the software to make it all go.

I actually started the whole thing in august last year. There has been a lot of time spent waiting for parts to arrive and a lot of frustrations when things did not go according to plan, but now it's looking pretty good.

The Bench

The machine needs a solid bench and I bought one, twice.

The frame of the bench needs to be solid and it looks like the best one is made by a company called Kreg. It comes in parts that bolt together and it is thick steel – but for some reason they are very hard to get here in Oz. The first time I ordered it they refunded me a week or two later because they could not get one.

I ended up having to buy a smaller one just for the vertical legs and a set of long horizontal bars direct from the US. It was worth it though, I did not see any other item that was as solid.

I tried out other options -  (getting cheap chinese metal benches and widening them) - but they were so flimsy and wonky that the bench would never have been stiff enough to use. They are okay for what they are – I have three in the garage, but no good for mounting the CNC on.

Then there was the bench top.

I could not get plywood thick or wide enough for my initial plans so I ended up getting thinner 12mm sheets, redesigning things so the bench was only 1220 deep (max plywood width) and then gluing two sheets of the ply together to make one solid plate 24mm thick to mount the CNC on.

I don’t recommend this to anyone: with no proper tools or even a bench 1220 x 1300 to sit them on it while gluing it’s surprising that they turned out okay. Even then the whole thing is not quite flat: it has a slight bulge in the centre.

I had hoped that gluing them together would flatten it but nope.

Now that the machine is bolted to it I will use the machine itself to mill the middle of the bench flat before I put the spoil board on.

Other bits

I added a lower level shelf from some bits of a cheap metal bench and another piece of the same plywood to stabilise the bench a bit and give me somewhere to put other parts of the machinery out of the way – the electronics, the water cooling system for the milling head and the suction hoses and so on.

------------------------------------------------------

The plan

I didn’t just throw all of this together, I made plans from the start – and I designed, re-designed and re-re-designed things in 3D using Blender, my preferred design tool.

This is version twentyfour of the Blender model, and for reference below is a photo of the real thing : Not too far off. The cover frame is only a test fit here.

I am missing out a lot of the details here, trying to give a short report on a very long and drawn out project. This is where I spend my weekends.

More soon in Part Two. Thanks for reading.

Bike Cables and Rusty Bits

 I have already replaced a lot of bolts on the bike with stainless steel ones : it seems that wherever you put a hex driver or wrench, the paint comes off and it starts to rust after a shower or two, and there has been a lot of rain around here lately.

Ferrules are those little bits at the ends of the cable tubes. Those on my E-bike were chromed metal and they were rusting within a week of rain hitting them.

 

“No problem” I thought, and bought a set of replacement corrosion proof ferrules, cables and the tool to cut the tubing.

A little detail on the nature of bike cables: the brake cables and shifter cables are not the same. Yes, they look the same but they are not. The brake cables are thicker and have 5 mm tubes. These tubes have a spiral wound metal tube with plastic covering outside and slick plastic liner inside. The shifter cables run in 4 mm tubes and the metal tube is more like a hollow cable itself: the strands run along the tube which makes a stiffer tube, more resistant to tension. You should not mix them up. The top ends of them have fittings specific to their function and this should help you figure out what they are if you don’t know.

 

Left - Brake cable, Right - Shifter cable

I started out replacing the tubes on the brake cables. The tube cutter worked fine, Then I filed the cut ends to get them neat and used the pointed bit on the tool to make sure the hole in the middle was clear to run the cable through. In a short time the replacements were done. I did not fix the new alloy ferrules to the cable tubes but it does not matter as they aren’t going anywhere. I also lubed the cables with dry lube. 

Cable tube tool

 

The shifter cable was a whole different thing. First, the shifter cable has three tubes on it. I didn’t know about the difference between tubes at that point and went to chop it with the same tool I used for the brake cables – it just made a mess. It distorted the tube and didn’t cut all of the strands. Fortunately I have a "grinder" tool and so I used that to cut the end of the tube off neatly. It took me a couple of tries to work this out as the heat from cutting melted the plastic on my first try – if you want to cut the tubing this way, chop a little at a time and wait for it to cool off before cutting more. This gives a very neat end provided that you are patient. Then you only need something like a piece of wire or a bent paperclip to clear the end of the hole and it is ready to use. 

 

The "grinder" tool with saw blade

This would have made fitting the cable tubes easy and fine except that . . . The cable itself decided to lose a fibre. This rucked up when I put a tube on and made using the tubes impossible. Once it is bent you can’t really unbend it. A frayed thread or fibre can be rewound back into the cable if it is still curled as it was before but a sharp bend in one is another story. All I could do was cut it off high enough that the cut tip did not scrape on the tubes and wind it carefully back into the bundle.

I now have a new shifter cable and will try to fit this soon. Fortunately, the end of the new cable is already fused together. Fine until you have fitted it and then you will probably want to trim off the end so it doesn't hang down – and there, your fused end is gone. Also, once you undo the cable from its clamp at the end it is squashed and this can make it hard to pull the tubes off or get new ones on. Oh, yes – there is more than one end design, depending on the maker of the equipment.

I would have already put a new cable on except that I don't know how to get the top end out of the controller.

I have opened the side of the shifter controller but I can’t see either the end of the cable or where to remove it.

I went to the maker's website and got the service info for the specific model and . . . . it says nothing about how to replace the cable. Nothing. Cable part number, picture of it, yes - how to do it, no. The trouble with flying blind is that if anything goes wrong, for example a spring flying out of the unit as I take it apart (never to be recovered) – the bike is then unridable. My only answer: buy a new item of the same model and wait until it arrives, then take it apart to figure out the process. Once I have that, I can be sure that I can get it back together or replace any broken or missing bits. 

There are a lot of webpages and vids of "How to change the bike cables" but none of them showed the same model as I have or dealt with any of the problems I had, which is why I am posting this here.  They all make it sound so quick and easy - but it isn't.  

That's it for now. I hope any cyclists out there can gain something from my experiences trying to do my own repairs.  I could have taken the bike to a shop and paid plenty to have someone else do it  - but there is a problem: I don't own a car so how do I get it there?  - and really, I want to do it all myself.

Carrier Project Update

 

After some use I learned that the main rails were flexing under load and I wanted a proper carry box as well – one that was watertight and strong . The case I found is one of those sold for camera equipment and has a very good seal around it. It also has holes in it where you can put a padlock.

I then added some aluminium rail brackets underneath so that it slides onto the rails on the carrier.

The carrier needed to be remade to fit the case and when I did I added some reinforcing bars to stiffen it too. I also added a lock so that the case can be locked onto the bike.

I am now very happy with it, and the bike in general. I ride it to work every day and take it on long rides when the weather is good on weekends. 

The last thing I have to do now is to make the bike rust proofed. Some of the nuts and bolts have surface rust on them due to a lot of rain lately so I am replacing them with stainless steel bolts where possible and I have some special paint to deal with other parts.

 

The Carrier Project

Why a carrier?

I want one so that instead of taking my stuff to work in a backpack I can put it on the carrier. It can also be used for shopping. I bought one but it is tiny and not very straight – considering the size of my bike it will probably need a special design anyway.

Time to make my own. I started by designing things in Blender.

Carrier design Mk.1 and 2

Carrier Mk 1

Model 1 was designed using hollow extrusion for the frame. This was fine apart from the need for longer bolts. Then I saw a battery pack designed to fit a rear carrier that had the rating of 36V20AH which is more than my current pack, so I decided to make Model 2 which uses “L” section parts and thus has a hollow inside that fits the new battery pack.

I am still not sure that I will buy it but I have not cut any metal yet either.

Carrier Mk 2

Mk.3 Carrier

I thought some more about the design and figured it might be better to make the frame from one part and bend the corners if possible. This would mean getting a solid tubular object to bend around but there are big gains: a lot less nuts and bolts and hole drilling, smooth rounded corners that won't catch and a weight saving. Provided it can be bent neatly it will be well worth it. I would need to do a test bend of course.

Carrier Mk 3

Mk. 4 Carrier

Carrier Mk 4

I thought about it some more and realised that I will probably want that larger 20Ah battery pack so I redesigned the carrier to suit that option – as well as making it a little easier to fabricate by only having two bends instead of four. A second copy of the frame could then be added on top to make a hollow carrier to allow sliding the big battery pack in while still offering cargo space on top.

Carrier Mk 5

I found some brackets that will fix the carrier to the frame very well – pipe clips that arrived quickly. I had to buy a whole box of various sizes but so what? They were cheap. They are stainless steel strip with holes at the ends in a “b” shape that can be put around a tube or pipe and they have a rubber pad to insulate them. I can probably use them to fix other addons to the bike too.

The Mk.6 Carrier

I realised the best way to make the carrier was to make it from angle with the flat ribs protruding so that a cargo box can be slid onto the rails as required. If the ends of these are rounded then they are safe without needing the complication of curving metal corners.

Carrier Mk 6

The Cargo box could then have a frame of flat metal bars attached to the bottom with slots that will slide onto the rails and lock onto the carrier. The holes are probably going to weaken too much although they look neat.

After a lot of money spent and messing around I gave up on the idea of chopping the extrusion with the power saw (don’t get me started) and cut all of the parts for the carrier with a hand hacksaw. All went quite well. Then I had to go buy a new set of drills since the one I wanted was blunt and I have no sharpening equipment.

Then I had an idea: why not superglue the parts together instead of bolting everything? It is simpler, does away with the problem of drilling aligned holes in parts and is light.

Glued and painted the parts. Painting with spray cans turned out to be harder than I expected though: lots of it went into the air and after I had finished I discovered areas that needed more - well, I am no expert with spray cans. I might need to sand the parts and repaint them.

I will make a box up as a spray booth to cut down on the waste spray too.

Next Morning, the painting looked terrible. I had to wet sand it all to get a smooth surface and lots of it just came off way too easily despite me sanding all of the bare metal with P800 before painting. The “self priming” paint just didn’t quite work out as expected. There are a few reasons why it failed: first, my own incompetence at spray painting with cans. Then there is the conditions: hot and windy out in the backyard so a lot of paint never reached the things being painted, the parts all hung loose from wires which meant they moved about, and finally maybe the surfaces were not as well prepared as I thought.

I began wet sanding it all down with P400 and after a little pressure the glue joins popped apart. (At least it didn’t happen on the bike.) So now I need to drill and bolt all of the holes as originally designed and finalise the parts, then rig the finished parts in a wire frame and repaint them.

It would have been a good idea to drill all of the bolt holes while the parts were glued together: that would have guaranteed that the holes all lined up. Sadly I didn’t do that but I got them right with care and patience. Lots of patience. The parts matched up by glue areas – I could see where some glue stayed on one side and some on the other so by matching glue patterns you could fit it back together. It all bolts together and after some on-bike adjustment it fits on. Now just one small puzzle: identifying the parts. When I go to repaint them (properly this time I hope) I need to mark them so I know which part goes where since the hole patterns are not exactly the same. Some surfaces don’t need paint and where they meet up no paint would be better so in theory I could write the markings there and then cover those areas with tape.

The final fitting of all parts on the bike showed that I could fix the small support bars to the bolts at the front of the carrier rather than drilling more holes. All of the support bars only got their top ends drilled and lengths cut at this point - regardless of how well you might design things in CAD, nothing beats actual fitting.

I wired the carrier parts into a frame for painting. I did this so that they don’t move around as I am spraying them, something that happened in my first paint spraying.

I was also more careful about heat and wind so it worked out right this time.

Assembled everything today. The paintwork is still not great in that it is easily scratched, which I did during assembly. Never mind. It is good enough. It’s strong and solid.

Here are two pics of the carrier before I added the mudguards.

 

It is held together with M4 bolts and nylock nuts and is fixed onto the back of the frame with two pipe clamps under the seat and two M5 bolts to hold the side bars on.

The back mudguard bolts into that hole in the middle of the carrier.

As a bike to get me to work and back it is real good. Yes, it still looks better without the guards but that doesn’t stop the rain coming down.

                                              Bike with carrier and guards

This is the final design done in Blender of the carrier made from aluminium extrusions from the hardware store. The parts are few and I could carry them all in one hand easily.

The main rails, ends and right- angle joiner parts are all cut from 30 x 30 x 1.5 mm “L” section. The main supports are 15 x 5 mm bar.  Everything is held together with M4 stainless bolts and nylock nuts.

Considering that you can’t buy a carrier to suit a bike like this (well, not that I have seen), and how cheap the materials were, it was a pretty good project to spend a weekend or two on.

If you want to build your own, drop me an email and I’ll send you some plans and a parts list.

 

 

THE SHOLLEY PROJECT

I have a shopping trolley. It is made in Germany and quite good, it has lasted me several years of weekly shopping trips during which I would push it to the mall and shop and then push it home full - and over that time I learned that there are problems with the design – and since I could, I started designing my own to replace it.

Original model weaknesses

1. Front wheels are too small. Even a small ditch can stop the cart.
   

2. Bendy. Some would call it a feature but the plastic frame bends a little on uneven ground. I want something a bit stronger. I am concerned that this will eventually result in it breaking.

3. Folding. The current model has a lot of complexity and parts because it folds flat while I don't need that. Yes, it is nice but unnecessary.
   

4. Low ground clearance. When I  take it up steps the bottom gets scraped.

Basic Dimensions in cm:

length minus handle 64

width 54

height 84

ground clearance 8

main wheels 17

front wheels 9

upper bench height 57

handle overhang 23

Handle height 104

Bin internal 50 x 34 x 26

---------------------------------------------------------------------------------------------

DESIGN QUESTIONS

ONE: To fold or not to fold?

This is the big one: a solid frame is a lot simpler to make but not as useful.

I have designed a number of folding models but the existing model has almost never been folded. This, however, does not mean it will continue this way: I could in future move to a smaller place that makes folding necessary – but the complexity also brings weakness.

TWO: Stair Climbing

This means putting tri-wheels on the back - I would need to take these off an existing hand trolley. A useful feature as there are a lot of places where even a small step can cause a lot of trouble. The plastic trolley has its bottom gouged from these – the tri-wheels would not solve this but an aluminium frame will. Better ground clearance will help too.

-------------------------------------------------------------

DESIGN VERSIONS

Version 10

This is based on the mostly plastic folding trolley I already have.

Version 30

Trying to make the lower tray accessible I split the tray. Not bad but I would need to find or probably make the split trays, not easy.

Version 40

Shifted the main wheels back to allow sliding the lower tray out sideways . This not ideal engineering wise however.

Version 63

This is the peak of folding designs – I even worked out a locking system for the front and back verticals. The down side? First, the complexity, second, the bottom tray has no protection against getting hit upward on bumps or steps. This was directly because I was trying to make it fold small and keep it light.

Version 67

At this point I was still thinking of folding it up and loading on the back of my bike so I put the folded design onto a virtual bike to see what it looked like and it looked bad. Clumsy and heavy.

From version 70 or so I decided to simplify and drop the folding aspect as it made things too complex. I also decided to add triwheels to make it easier to get up stairs and on and off public transport.

Version 74

The triwheels have arrived so here is the latest design with them as well as a solid (non-folding) frame.

This will make construction a lot easier and stronger but there is one down side: you can’t easily put brakes on triwheels. Still, this is looking the best yet. It is probably wildly overbuilt but as I don’t have FEA software I don’t care. It is made mostly of rectangle section alu with stainless steel bolts and nylock nuts. Brakes? Who needs them ! All of the parts are easily available and cheap enough - and now in my garage.  all that I need do now is get time and space to actually put it together.    Hopefully this will happen over the holidays. 

The TRX40 Build - Prices

 Purchase Date: November 2020

A major reason I decided to upgrade at this time was warnings from computer geeks that part prices would soon climb and that the shortages of parts would get worse.

This was highlighted by the GPU being $3000 and that was from a reputable supplier which listed several models of 3090 but could only guarantee supplies of one or two models.

I am still happy that I did get it though – it is the first computer that I am really happy with.

AMD TR4    3970X CPU (32Cores)                     $3279

MSI TRX40-PRO WIFI Motherboard                    $559

64 Corsair 3200 DDR4 RAM                                 $418

Corsair HX1200 PSU                                            $439

NH-U14S TR4-SP3                                               $159  (Special TR4 socket cooler)

Define 7 Compact Case                                        $199

Noctua NF-P12 PWM Fan x 3                              $147

Samsung EVO M2 SSD 500GB                           $145

EVGA RTX 3090 GPU                                       $2999

Samsung NVMe M2 2TB                                     $499

NF-A15 PWM Fan x 2                                            $98

TOTAL                                                               $9154

This was also my most expensive computer buy to date. I considered buying the AMD TR4 3990 64 Core processor but apparently Windows 10 does not run on it. Nope. You need Linux for that and I am not going to go there. Oh, and it cost $6000.

Looking at the same store now, the same model RTX 3090s are still available but now they cost $3600. All prices are in AUD.

This is all here for historical reasons: I am curious to see how this looks in future.