More Enigma progress.

June 11th, 2017

Sorry it’s a very badly composed film. It’s a bit tricky getting the angle right! Am slowly making progress on the machine. It’s a slow process because I need to design, print, test then often redesign and re-print the parts. But I am getting there.

I need to thank Paul Zetzmann who is undertaking his own truly amazing Enigma replica build. It makes mine look like a toy in comparison. His site is here:

Paul, like James Grime, was kind enough to explain some details about the machine as well as make a small video for me explaining some of the details. I must also thank him for pointing out my embarrassing mistake of gluing the letters onto the rotor rings in the reverse order (fixed now)!

I also just discovered this site:

They have a replica three rotor Enigma that looks amazing although it isn’t a 100% replica as it doesn’t fully function. It is built to replicate the look and feel of the machine and show some of the machine functions. It has some limitations (they describe it on the site here) as well as a modern computer inside to to do the encryption/decryption. Unfortunately their site doesn’t show the innards of the machine so I can’t see exactly how it works. It looks as though only the fast rotor moves and that you can only encrypt/decrypt 20 characters. I wonder what they are using as the computer? Some sort of Arduino or a Raspberry Pi I imagine? The machine definitely looks the part though!

Am currently somewhat pre-occupied with trying to find a new job. Unfortunately I am not earning enough at the workshop to break even and when there is no work on that problem is of course compounded. Unfortunately other animatronics jobs aren’t exactly easy to come by! So I am busy trying to get back to my old career of software testing/test automation. The good thing will be when I do find something new it should mean more money and more time. No more 10 hours standard work days – more time for projects!

A 3D printed Enigma machine.

May 16th, 2017

This post is a little late as I have already made a film about this project but it’s fairly early days so far. Basically I am trying to 3d print a working Enigma machine using the 3D printer I built.

The film explaining it is here but I will add more details here now.

I have made an Enigma machine before of course. I  made an Enigma machine wristwatch some time ago. There are plenty of posts about that and the Turing Welchman Bombe I made here on the blog. That Enigma was mostly done in software but for a long time I have wanted to build my own working, physical Enigma machine. For years I had it in the back of my mind to make one from Meccano but it would take an awful lot of it and I don’t have any! Lego Technic also sprang to mind but again I don’t have the parts, my 8860 set sadly lacking in any useful Enigma type parts!

But then 6 months or so ago I built my own 3D printer and have been looking for useful things to do with it since. Normally I am a metalworker but being able to model and print plastic parts does have its uses. I have mocked up some parts for my vintage cars with it and have made useful little bits and pieces. I’ve been a bit slack on blogging lately so I haven’t shown all the progress I have made with it.


The machine has been upgraded to a 300 by 300 heated bed. It’s mains powered so heats up in seconds! Recently I have set up Repetier Server running on a Raspberry Pi so I can remotely upload and watch prints from any computer on the Internet. I also have a web camera set up that I can remotely monitor so I can watch the prints in action. So now I can start a print in the morning, go to work then watch the progress during the day and stop things if necessary. Being so close to work (7 minute walk!) I can pop home and start new prints if needed.

Part of the printing process is having something to print. You can download things people have already made from sites like Thingiverse but I wanted to do my own designs and learn about 3D modelling. On the advice of someone at work I started learning Fusion 360, a free for personal use, CAD program. Aside from being Cloud Based, which I don’t like (‘Cloud’ meaning ‘someone elses computer’ when it comes to software), it works very well and is quite powerful. It is also being updated all the time with new features and there is a ton of help and advice for it available online.

To learn how to use it properly I needed a decent project. That’s where the 3D printed Enigma machine came in. I understand how the machine works very well already and knew it shouldn’t be too hard to make a machine that works in the same way. It wasn’t to be an exact replica, that’s not possible when 3D printing because of the way the parts are made, but it is to be a fully functioning machine following the form of the original as much as possible. There will be no electronics and certainly no software or microcontrollers to worry about. Purely mechanical and electrical only. If you want electronic versions there are plenty of them available online including some very nice, modern replicas for a decent amount of money.

When I first started all I had were pictures I found online and in the various reference books I already have. So I started measuring things up to start drawing up plans. I started with the rotor. I would print out the picture then choose a base measurement, say a rotor diameter of 100mmm, then I would measure the picture to see how big the printed dimension was and work out the scale factor I would need to then convert everything into real life measurements. This sent me off on a tangent after getting sick of converting every measurement by hand.


I decided to build a set of calipers that would do it for me! Now this is a whole other project in itself that I need to write up. I bought some (very) cheap calipers online and using and Arduino worked out how to read the data that they spit out and convert that to a measurement. There are lots of examples online of how to do this although I found mine used a different protocol to any example I found so don’t assume they all work the same!


You can see the calipers working in the film above but basically you measure something on the drawing (or model or whatever) then press a button on the device then slide the calipers to whatever the real measurement should be and hit the button again. The device then works out the scaling factor and from that point on anything you measure will be converted to the real measurement. I took the prototype to work and people there (prop makers, model makers, sculptors and so on) all loved it. Turns out the idea is patented already but I will eventually finish mine just as my own useful tool to own.

Soon after that interesting diversion I discovered  that there are plans drawn up from an original Enigma machine online that you can download. They are available here although you have to register to download them. I can’t tell from that page (which appears to be for some kind of university or technical school?) if they ever actually finished the project. There plan seems to be building a number of replicas for teaching purposes. The drawings though seem complete although it’s hard to tell exactly which variant of machine they were drawn up from. I think I have found a few small errors in them too.

But, as I mentioned above, 3D printing a machine doesn’t mean you just copy the plans and print the parts out. It just doesn’t work that way. The problem is the way 3D printing works. Basically it’s an additive process unlike CNC machining where you are taking away. This means I have to modify every part so it can be printed, and be printed on my homemade machine. I did hit a limitation there what I will explain later. I also gave myself the added challenge of printing it using no support. Support, in 3D printing, is when you print a scaffold like structure to hold up other parts of the object you are printing. As the layers are build up from the bottom up any overhang part needs support under it. This site explains it pretty well:

I don’t like using support, well not on visible pieces, since removing it always leave the part looking ugly. You can sand parts but I don’t want to bother with all that so I am trying to use no support in my parts. You can see where the support was in the part below. The face upmost was printed as the bottom layer hence the need for support on the overhang.


Since the part above is bolted into place I printed the actual version as two parts that bolt together, that way I can avoid the support.

I should mention that there is another 3D printed Enigma machine out there. Well, sort off. There is this project where again it seems a school has a project to make a 3D printed Enigma. It was from there I learned of the existence of the plans online. It seems they never finished their project either (probably because they ran out of term time) and it seems to be a somewhat simplified machine but it is a very good start. Some of the parts for that are on Thingiverse here.

A very good technical description of the actual Enigma machine is found here is you are not sure how it works or want to know more of the physical details.

As soon as I had the plans I was able to start modelling my parts with 3D printing in mind. I haven’t yet modelled them all and I am still working out details like how the keys will work but this is what I have so far. I started with the rotors because even if I don’t get the whole machine working a nice set of rotors would be nice to have. I am using pogo test pins as the spring contacts and small pieces of bronze welding rod for the static ones. I also want to have a working ring setting on the rotors so built all that into them too. The ring setting is fairly easy since all you are doing is moving the wiring with respect to the lettering on the ring. I made a prototype rotor.


Hmmm, thank goodness I still have the mind of a tester! Reading what I just wrote I realised I made a slight cock up. My model is wrong and I just had to fix it. Only the wiring moves, not the notch position. That remains locked to the same letter on the ring. On mine the letters moved but not the notch position. Novice Enigma error!

But this shows the beauty of 3D modelling. I was able to fix my model by quickly redesigning how the parts worked and now I am busy 3D printing my updated version. This is what they mean by rapid prototyping!

I confirmed with James what the actual notch positions are on the physical rotor. They don’t match the turnover positions of course as the turnover is based on the letter uppermost on the rotor where as the physical notch is on the ‘back’ of the rotor in the machine.

The so called turn over positions are usually given as follows (what is showing when the next rotor turns over):

Rotor I: R

Rotor II: F

Rotor III: W

Rotor IV: K

Rotor V: A

They say that was remembered at Bletchley Park during the war by the mnemonic “Royal Flags Wave Kings Above”.

The actual, physical notches are not against those letter though. I worked out they need to be at N, M, D, R and H on my rotors so when the turn over happens the right letters are showing in the top of the machine.


The letters were interesting. For a start the typeface used is unique to the rotors as far as I can tell. Annoyingly, to my my mind which likes order and neatness, the typefaces used on the real Enigmas are all different. The rotors, the keyboard and the lamp board all use different typefaces! For the rotors I created my own using an online tool called FontArc, which is free to use. It works brilliantly, you can start with standard fonts then tweak them which is what I did to get my rotor letter correct.

To actually create the little letter panels I tried various things. I printed all the letters as one piece with the little letter blocks sitting on a couple of layers printed underneath them. This was so I could print all the letters at once as one object then cut them out to stick on later. This is a good example of the rapid(ish) aspect of 3D printing.

My first thought for the actual lettering was to print the white squares with the letters in them as a depression in the surface. I could then fill the depression in with black paint. That didn’t work for several reasons. First the plastic is quite rough so paint would get into small cracks in the surface and not be able to be wiped off. And secondly the plastic is actually a little porous and the paint would bleed out of the letter.

IMG_0780_1 IMG_0779_1

My next go was to make the letters raised then paint the tops of them. I tried using a roller, that didn’t work.

IMG_0782_1 IMG_0778_1

So I tried colouring them in with a sharpie.  That worked better but I  still wasn’t totally happy with the result.

What I actually did in the end took a lot of fiddling but the result was so much better it was worth it. I did a two colour print! It took a few goes…

IMG_0783_1 IMG_0784_1

How you do this on a single extruder machine is tricky but basically what I did was print the letters as in my second attempt with the letters themselves as a raised surface. But I modified the g-code the slicer produced to pause the print at the point that it was starting to print the bottom layer of the letters. Repetier allows you to do this using the command:

@pause My Message

I then manually lifted the extruder using Repetier, change the filament from black to white, purged the extruder, the reset the nozzle position. The print then continued using black filament.

The results were excellent after a bit of trial and error.


Since then I have been designing and printing a lot of the parts to make the machine. The rotors are pretty much done. But printing parts, like the letters, takes a lot of trial and error.


I haven’t done the electrical parts yet as I want to make sure the mechanical works first. I have started printing some of the housing and also the mechanisms to do the movement of the rotors. I want to make sure all that works before I start tackling the keyboard. That will take a bit of experimentation. Since the base is a large print I am mocking things up on a wooden base first before I commit to doing such a large print.

This is the mock up so far of the rotors, entry wheel and reflector mounted between the correct housings with the lever to the left hand side.


One thing about trying to do this with 3D printing is how flimsy plastic is compared to metal. I am having to model in strengthening ribs on parts to stop them bending and you can see in the pic above even that isn’t enough. The reflector is sagging because the shaft it is on it only attached to plastic and it isn’t strong enough. I will probably solder a flat piece to the metal shaft (instead of replying on plastic) and attach that directly to the side piece. This is how it was done in the original anyway.

One thing to note is not everything is 3D printed. Shafts, screws, nuts and so on are metal. This is for practical reasons. On trick I have is shortening screws to be the correct length. I tend to buy screws over length then cut them down. This is a trick I use to cut them down to the same lengths. I get a piece of steel and drill and tap it the right thread of the screws I am cutting down. I then thread the screws through the plate and cut off the protruding thread then sand the ends flush on the linisher.

IMG_0768_1 IMG_0769_1 IMG_0772_1

When you unscrew the bolt the thread is cleaned up by the plate and all the threads are the same length. If you need to make the thread longer you simple add nuts under the bolt head before you screw it into the plate.


I am not sure if writing all this up or just doing little films is a better way of showing progress on this project. I don’t seem to have as much time for blogging about things these days. Perhaps small films are a better way to go? This is one I did about making the rotors.

Printing update.

September 15th, 2016

A quick update on my printing efforts the other night. It fell over!

Well, not the print itself. This is what I got up to the point it went wrong.

IMG_0383 IMG_0384

It’s upside down in that first picture. The second one gives a good view of the infill, the pattern inside the printed object so it is not totally solid.

The reason ti fell over was simple. My laptops power settings kicked in and it went to sleep! I am sure I had fixed that. So far my biggest issues 3D printing have been caused by the computer side of things. I forgot to mention I spent hours in the weekend not able to upload new firmware to the printer. It keys getting write errors. The simple blink program would upload fine, but not the 3D printer code.

In the end my friend Mike had the answer, try a different USB cable. That was it. A new cable fixed it.

I redid the print the other night. It took 7 hours! This is how it came out.

IMG_0389 IMG_0390

IMG_0391 IMG_0392

This is the Pink Panther Woman off Thingiverse. I wanted to print something other than a boat and try a longer, and bigger, print.


It is 150mm high so took up 3/4 of the build height I have available. I print at 50% feedrate so my prints take about twice as long as ‘normal’ printers.

I started experimenting with that. I printed another 3D Benchy boat at 70% speed to compare to my best print at 50%.


This is the 50% print on the left and the 70% on the right. You can see how the faster print isn’t quite as good.

IMG_0387 IMG_0388

I still need to play with the speed and acceleration values in the firmware itself. Currently I have them set to 1000 acceleration for X and Y and 100 for Z. Speed (feedrate) is set to 200 for X, Y and Z and Jerk is set to 20. I think my issue is going to be acceleration, there is just a bunch of weight to be getting up to speed (and stopping).

I started putting together a Linux box as a dedicated printing machine. I am running Xubuntu on it on the advice of Dave. I still need to sort out wiring on the machine and a balancing load for the 5 volt rail. I can see if we have any old, high powered resistors in the junk at work.

I also want to start 3D modelling my own designs. I did start with DesignSpark a while back so need to get back into that. I was trying to model up my TARDIS!

Posted in CNC Machine | Top Of Page | Comments Off on Printing update.

Actually printing.

September 13th, 2016

I spent all weekend and nights after work fiddling with the printer. I added the limit switches. The Z axis switch mounts to the motor which is at the positive end of the travel. The X and Y are both on the negative ends (i.e. at 0,0). This is all configurable in software but you do have to be aware of what it means for your home position, the position the machine goes to to know where it is. With the Z on positive it means my home position is (0, 0, whatever I tell it maximum Z is!).

IMG_0309 IMG_0376 IMG_0310

The Z axis switch is triggered by a small bent piece of metal attached to the top of the Z base, currently with masking tape but soon to be screwed in place. The X and Y switches bolt to the frame between one of the guide rails and the lead screw. They are pressed by small discs I machined up in the lathe. They have an off centre hole so the disc can be rotated and operate the switch like a cam. That gives me very fine adjustment of the switching position by rotating the little disc then tightening the screw.

For now all the wiring it temporary and taped in place. Cable management will need to be carefully though out so the wires aren’t caught in any of the moving parts, get in the way fo prints and they also need to be run so they cables aren’t strained. I have ordered connectors so I can make longer leads to the motors and sensors as well as some drag chain to help route it all nicely.

With the cabling in place I was able to test things out. I ended up changing the firmware I was running from Marlin to Repetier as I am also running Repetier as the PC host to talk to the printer. I set everything up with guessed values I figured out from a lot of reading online. The problem is not many people make 3D printers using screws on the X and Y axes. Most seems to be belt drives so the settings are not the same. Where lead screws are used it is usually for the Z axis which tend to be run much slower than the X and Y.

I cobbled together some values that worked, set my limits and tested the switches. They need some careful configuration to get right. I also played about with the microstepping on the motors, eventually settling on 1/16th microsteps. Any bigger and the machine was noisy and any smaller (1/32) sometimes the steppers missed steps.

With all that done I was able to control the machine to move!

So the next thing I did was try it out as a plotter. This takes a little fiddling with configuration but I got that going. You also need something to plot with so I drew up and laser cut this pen holder at work.

IMG_0370 IMG_0371 IMG_0372

It has a sprung attachment that bolts to the Z axis that the pen rides on. This is so I can use a constant pressure on the pen. I used a piece of MDF as a flat plotting base and managed to draw this on it. The screws that hold it to the machine are effectively captive behind the sliding arm. The holes allow an allen key through so the thing can be screwed in place.


My little table made from an old stool base comes in handy for working on the machine. It rotates and you can adjust the height! It makes a very good laptop stand too.

I was able to plot a larger version of the test file I made for my little laser cutter.


With the plotter I was able to fine tune my speed and acceleration values (down) until I could accurately draw the logo multiple times over the exact same lines.

Since I knew all that worked I looked at attaching the plastic extruder. I am using what’s called a J head extruder. First I made the heated base mount because I needed to know the height of that to know what height I would need for the extruder.

The base is another piece of scrap MDF (cut at work now I have been trained am allowed to use the table saw). That attaches to the base plate with 8mm bolts that fit into the mounting holes. The top of the bolts were machined flat and drilled and tapped for 3mm screws.


The holes allow the heated bed to be screwed down to the bolts on the base plate. The heated bed sits on screws so you can adjust the level of it. On top of it goes a flat piece of glass held with binder clips. This is to provide a perfectly flat surface to print on. The surface can be leveled by adjusting the screws. The springs ensure the level stays set. They also allow the bed to move sideways slightly which is important as it expands and contracts slightly as it heats and cools. If you screwed the corners down hard, as the bed heats it would expand and bow.

I am also using a piece of cork tile under the bed as insulation.


The original idea was the bed could be bolted to the Y axis base with 8mm nuts under the plate (remember the cut outs in the frame rails to allow for this) but so far that hasn’t been necessary as there is enough weight to hold things still and the holes are drilled to a close enough tolerance to not allow any movement.

I will make a similar, larger base for plotting/laser cutting on but with a flat, sheet steel top. The steel means you can hold things onto it with magnets.

With the bed done I knew the height I needed for the extruder. The J head uses an aluminium body with a round head that has a slot in it for attaching to the machine. I expect you can buy brackets for these (or 3D print them) but I made my own.

A small piece of L angle has a slot cut into it to match the groove in the extruder. Another small plate had a hole drilled in it the same diameter as the groove. This was then cut in two. One plate is screwed to the angle bracket. The other has over sized holes allowing it to be pushed hard up against the extruder then hold it in place.

IMG_0323 IMG_0321 IMG_0322

This works very well.


With that in place I was able to try the machine again but I hit a snag. The Z axis wasn’t working very well. The stepper was skipping steps!

It turns out the carriage was binding slightly on the guide rails and was too still for the motor to move. That turned out to be because of an ‘improvement” I had made to the machine. I originally used button head screws to attach the plate to the carriages. I then changed to countersunk screws so I counter sunk all the holes.


There is one small problem with doing this. The countersink forces the screw into one position, even if the hole is oversized to allow for some wiggle. Since this machine is hand made my drilling wasn’t precise enough and when countersunk screws were used it forced the carriages into a position that made them still on the rails.


The solution was to go back to the original screws. I just turned the plate over so the screws had a flat surface to sit on. That allows a bit of wiggle in the positioning before the screws are done up hard. This solved the problem. Note the taped on Z axis limit switch actuator!

Steed looks on.


With that fixed I was able to start squirting plastic. The first thing I tried printing was the 3DBenchy boat. This is meant to be a good test for 3D printers. The first couple of attempts didn’t go so well. It seemed the machine was going too fast so I lowered the feed rate each time.

IMG_0357 IMG_0359

At 50% speed things started going better. I got my first actual print! This took about 1h 45 minutes. Apparently this print should normally take about an hour so my printer is slow but then it was never built with speed in mind. The first print was rough but recognisable.


One of the parts of the process of 3D printing is called slicing. This is where you take your 3D model and run some software on it that that works out the commands to actually drive the printer, this is the slicer. It spits out G-code commands which is what the firmware running on the printer used to know how to control the machine. The Repetier software I am using comes with two different slicers, Slic3r and Cura.

I had been using Slic3r (this is what I used to do the plotting too) but I thought I should give Cura a try too. Both pieces of software have many settings to fiddle with but I am using defaults for my setup. In Cura this is a 0.2mm print using a 0.4mm nozzle.

My Cura print came out much nicer.

IMG_0347 IMG_0348 IMG_0368

I was quite pleased with that! Not bad for an untuned home-hand made printer. That print I actually left running by itself while I went to bed. It seems the way with 3D printing that everything takes forever. I have noticed this at work too. It’s quite a slow process, even on fancy printers.

Today I was working from home (I needed quiet to concentrate ) but I couldn’t use the printer as all the software is on my laptop and I was using that for work. I really must get a dedicated machine built and running. I did fire up the rescued thrown out PC from work and that works so I will build that up soon.

This evening, after work, I did some more playing with the printer and settings. I tried speeding things up again and had problems. I also played with the filament temperature. Then it occurred to me I had never calibrated my extruder. You need to tell the software how many steps it must drive the feeder to extrude a given length of plastic. I hadn’t set that up. It was set to 50 steps per mm but after doing some tests I found it should have been 88 steps. I hadn’t been extruding enough plastic!


During the tests the machine printed that by itself.


With the extruder set correctly I printed another boat! The base was immediately better. I also changed the settings to print more of a border. I think this is just to mark around the outline of the print you are doing but it also gets the plastic flowing nicely before you start printing your object.

That boat came out better than the last!

IMG_0377 IMG_0378 IMG_0379

The blue tape, by the way, is there to provide a nice surface for the first layers of the model to stick to. You don’t want your print moving while printing it! Apparently 3D printing enthusiasts spent a great deal of time experimenting with anything they could print on to come up with the best surface. Blue painters tape works very well. Kapton tape also works well apparently but it is much more expensive.

That’s one of the great things about 3D printing. Hobbyists have done so much work and experimentation into making all this work. It makes it much easier for me to come along later and read what they have done and make my own.

Still lots to do. Play with all the settings. Speeds, accelerations, temperatures, nozzles, layer heights, all kinds of things. Best to change one thing at a time though! I need to sort out the wiring and a housing. As well as cooling. I am currently using the print cooling fan to cool the electronics! I need a permanent fan blowing air over the drivers to keep them cool. I also need to look at mounting a small cooling fan on the extruder head to help cool the print although I am not sure how necessary that actually is yet.

My power supply is having regulation issues. The 12 volt rail wobbles about a bit. This is a known issue with PC supplies where the 12 volt regulation is not very good without a 5 volt load. I might add some 6 volt car laps as a load and to provide lighting around the printer. Hopefully this helps. I could also run a small 5 volt cooling fan as well.

I also need to print something other than boats so I started something tonight. Have to wait till morning to see what it is!


Posted in CNC Machine | Top Of Page | Comments Off on Actually printing.

CNC Y axis.

September 5th, 2016

This weekend I finished (mostly) the Y axis and base of the CNC machine. This is again made from L section aluminium extrusion. I really had to scrounge about at work to find enough bits to make it. I cut the angles on the drop saw again one lunchtime. I did the same as on the gantry and used angle pieces and small square plates in the corners to hold it all together.


One thing that is really handy when working with aluminium like this is a very coarse file. I used the file I use when I do lead work on the car. The name of the file totally escapes me but it has a very coarse pattern, one side a curved cut. It works great on aluminium and doesn’t clog like a normal file.


After bolting it together I made the base plate and the two lots of spacers that go between that and the linear bearings. The spacers are needed to lift the base plate above the edges of the extrusion used to make the base. After doing all that I drilled the holes for the rails and screw mounting again using a laser cut card template to mark where the holes went. It was only when I went to assemble things I found the template was wrong! I am not sure what happened but the spacing of the rails didn’t match the spacing of the moving bed.


I suspect the issue happened when I drew the templates. I probably accidentally stretched something. Since the holes were only just off I couldn’t drill new holes and I didn’t want to drill or file the existing ones bigger so in the end I remade the base plate that mounted on the rails so I could use the slightly off rail holes. Instead of remaking the two spacers I just cut them up to make 3 sets of smaller spacers. I had been thinking about doing that anyway to help save weight so that worked out fine.


The base plate is mounted with countersunk screws that sit just below the surface so that the top face of the bed is totally flat. I drilled six 8mm holes in it to be used as mounting points. The idea is to use MDF sheets with mounts that match the 8mm holes as changeable bases depending on if the machine is being used for laser cutting or 3D printing or milling.

The holes match the holes in the corners of the 3D printing heated base so my idea is to use 8mm bolts through a piece of MDF. The bolts go through the 8mm holes and are secured with nuts underneath. The tops of the bolts, where they come though the MDF, will be drilled and tapped for 3mm screws. The heated bed will then sit on springs sitting on top of the bolts with 3mm screws holding it down. That allows for height adjustment to level the printing base. It should become clearer when I make it I hope. I am rather making this up as I go along!

Since the base goes past the edges of the frame I needed to make cutouts to leave space for the nuts that hold my bases on. These were marked, then cut with a hacksaw and the bottom of the slot drilled so a piece could be broken out using pliers. The edges were then filed smooth.

IMG_0292 IMG_0291

I also files grooves in the X axis to allow for the bolts heads there that hold the Z axis in place where they pass the edges of the frame. Having the axes go past the edges of the frame make the machines footprint smaller while still allowing full travel on the lead screws I have.

IMG_0304 IMG_0305

It only occurred to me later that I could have used countersunk head screws there too and removed the need for the grooves!

I also bolted the gantry with the X and Z axes to the base. When I cut the base I deliberately made it fractionally wider than the width between the gantry uprights. This was to allow for any small variations in measurements. I wanted the base to be wider so there was a gap between the base and the gantry. If wider you can add in shims. But if the base had been narrower it would have been very hard to fix!


I made the shims to fill the gap each side from some 1.2mm aluminium sheet.

IMG_0295 IMG_0297

They filled the gaps perfectly. I still need to make another bracket that fits behind the vertical so that the gantry is held to the base in two planes and not just on the side. The extra mounting holes for that are seen above.

I also spent some time this week getting the electronics done. The thermistor for the heated bed arrived so I fitted that and I have had the bed and the extruder successfully heating. I even managed to squirt out some plastic but not very well as the bowden cable was all curled up and the filament was in a huge tangle on the desk. The feeder motor would skip sometimes. But I was able to extrude nice, even spurts of plastic. I also cut some glass to go over the heater. Really I should use borosilicate glass as it has the lowest coefficient of expansion. That’s hard to find here so I am just using ordinary thin glass. At least it is a flat surface to print on and the glass stops any damage to the heater itself.


I need to get more of the metal clips to holds the glass to the heater.

I will need to measure the temperature of the actual glass I think. The bed thermistor is in the middle of the heater so the reading you get there isn’t necessarily the actual temperature of the bed. The temperature of the bed may not also be even, it is likely to be cooler in the corners. When everything is properly set up I will measure actual temperatures so I know how the shown value compares to the real world values. I plan to use a layer of cork under the heater to help insulate the base of it.

On both the bed and the extruder the electronics seem to maintain the shown temperature quite well, to within a degree or so. When it is all set up I can fiddle with the PID values (used by the software to maintain the temperature)  to try to make things as accurate as possible.

I still need to make a small power breakout board for running the various fans needed. The extruder has a fan that runs all the time. As do the electronics. You can also use a fan on the parts you are printing I gather but I am not certain of the details on that yet. It depends on what plastic you are extruding.

I have had all three axes moving but they need some adjustment, both mechanical and electrical (motor currents). I need to extend some of the motor wires too as the ones on them are too short. I bought some cable to do that today. I might see if we have the correct crimp pins at work for the motor connectors then I can make entirely new cables instead of splicing in new cable to the existing ones. It would be much neater. I also ordered some drag chain to help make the wiring neat. It is mainly needed on the Z axis as all the wires to that need to move.


The above shot gives some idea of the size. The heated bed is a standard 200mm by 200mm one. I also drilled four 8mm holes in the base to work as feet and/or mounting points. Currently I have four 8mm bolts through the holes to act as feet but I am wondering if I shouldn’t bolt the whole base to some MDF to increase the stiffness of it. I don’t think that is necessary as there is movement in the bearings/rails way before the base ever twists.

Still to do is the limit/homing switches. I will probably mount them near the motors so I can keep all the wiring neat. This means the Z and X axes will have the limits on the positive end but the Y axis will be on the negative end. I am sure this is all configurable in software (I hope anyway!). I have an idea for making some off centre cams as adjustable microswitch triggers but I need to work out the mechanics of that first. Whatever I do they is plenty of room for mounting things.


The other thing I did today was buy a new Dremel.


These are very cheap at only $60NZ, I think because they only have two speeds rather than the multiple speeds of the other models. I am hoping I can use this as a spindle for light CNC work and PCB milling on my machine. I am not sure it’s going to be rigid enough for that (8mm rods actually flex a lot!) but as my old Dremel was starting to play up it was worth getting anyway. The first thing I 3D print, if this even works, might have to be mounting brackets so I can attach the Dremel to the machine!

It is mains operated but I think I can use a relay to switch the Dremel on and off based on the extruder signal.

All of this will become much clearer when I can make a small film of the machine hopefully working!

Posted in CNC Machine | Top Of Page | Comments Off on CNC Y axis.

Next Page »