30 May

Mini Kossel – it continues!

Will be posting this a bit retroactively as its very very very near completion!

That is, until something goes a bit wrong with the next thing that gets implemented!

I spent the bank holiday weekend putting in the hours – I loaded up a couple of different build guides to assist me – but most of it was pretty straight forward! The printed parts all fit, the only real headache was the sawing of the aluminium extrusions. I would strongly suggest either taking extra care when cutting by hand to ensure that the extrusions are cut to the correct length – or just get a purpose cut  set via ebay for a similar price than straight lengths. as a friend of mine said – you cannot put a price on ball ache!


Assembling the top frame mounts – very minimal clean-up required. The Ultimaker did a great job at printing these parts for me! I didn’t buy the flanged bearings for the belt guides – no idea why – so instead I am using bearings and washers. They are a bit noisy – so will have to gauge how annoying they are when I am printing to warrant replacing them with flanged bearings in the future.


A little bit further along and now the parts are assembled, at least for a dry fit. The colours look great IMHO.


I chose to add some stepper dampers to the build – its typical to get a song played at you by the printer whilst its doing its thing – but it gets old fast. These were about £5 each and more of an indulgence then really necessary. There are other out there made of lasercut cork sheet – but their cost isn’t that much less than these beasts.


Another shot of the top and bottom frame – with motors installed.


To create the rods for the delta I had to assemble the traxxis mounts and then bond them into the carbon rods. Wasn’t too much faff, I brought pretty decent araldite so that it would set quickly – turns out that I didn’t have any time left in the day and had to leave it overnight anyway! The other thing that is crucial is how long they are – I didn’t really focus on that – so will have a bit extra work to do in the calibration….


End of day one – about 4 hours in total as we spent a good portion of the morning faffing about in the car, followed by the eurovision party which stated at 6. The weather was great on the Saturday so started the build outside – I think there are still bits blowing around to be honest :s.


Start of day two – and I realise that the rails I brought only take M3 Cap screws and my extrusion starts from M4 – but really wants M5. So I had to improvise! Lots of M3 caps later with larger washers over a small washers – and then jamming it into the extrusion. I didn’t have much clearance and the nuts were on by a quarter turn, so a couple fell off during install – I must admit – I just carried on…hopefully it wont rattle too much when doing travel moves! I had to use a power drill to spin and tighten the nuts on the fasteners, as there as no purchase or bite on the smooth washers to the inside of the rail – most of them are tight now though! I possibly should have used MGN 15 – but that would have required remodelling the carriages…


The effector for the delta – pretty straight forward to assemble. Once again i’m using slightly longer bolts for no reason other than that’s what I have…


Mounted and installed with the timing belts – all standard. Except I got to use my matching cable ties! There is something to be said here though – and had I known this before (possibly through lack of research!) but I installed the rails where I was instructed – yet it starts several CM from the top frame,  I have at least 2CM extra space there now – surely I could have used that space for extra build volume?

Secondly – I spent a fortune on these rails – 400mm long. But due to the design – I dont have that length to build with.. So either A, I could have brought rails 250mm and saved some cash (they don’t exist as off the shelf purchase) or to extended the extrusion to 900 high, and mount the rails accordingly to get around 350 / full 400mm of build. In theroy its a simple swap – in reality – its a day to do that…so possibly wont bother!


here I am getting the wiring sorted – I brought the wrong size connectors from RS – so that was a bunch of money wasted! I opted to use female headers and solder my wires to those – a much more affordable method (and available to me without any extra purchases as I just raided the bits box).

The means I can slide them onto the pins on Ramps without any hassle!


The motors all wired up (X, Y, Z) – looks like doc oc from spiderman! What I didn’t realise was that now I started with the cable management – I had to continue – and that’s pretty much where the rest of Sunday and all of Monday went!


Its current form – my desk got really messy – and my usable workspace got less and less. I really should go tidy up but that’s not going to happen any time yet! its one of those things though, by front loading this, I wont have to re visit it later down the line, and when that involves a lot of soldering and connecting, re connecting, it worked out better I think in the long run.


One stupid oversight was the lack of mounting for my hot end – I was in a bit of a panic about it really as I don’t have any real tools, let alone material at home to fabricate anything! But, I do have my not yet calibrated Huxley on hand, and I quickly found a part on thingiverse and printed him out. Its a bit on the slide, but fits a treat! Its the white part in the above image. I have since taken it off, as I needed the space for the fan mounts and also, it pulled the hot end into the effector so well it created a very tight interference fit.


This was now Tuesday evening – meaning its eating it my evenings… I had hoped to have gotten it all sorted within the bank holiday – but that  went out the window with the amount of time it took to sleeve all the cables!

The above image is part way through the calibration stages – to ensure the machine knows where the bed is or some other stuff, like how long its arms are or something. The instructions looked very daunting – but I read through them slowly a couple of times and I think its pretty simple – very different to setting up a cartesian printer, also easier as I’m just following someone else’s hard work!

I did think though that it was a waste of time to continue with the calibration as I didn’t have the heated aluminium bed yet – so it would be fruitless – also it was late! I see a lot of videos on youTube that only print on glass – but I am going for the heated bed option. Whilst I don’t have much experience in printing ABS, I do have a couple of rolls of it, so that might save me a few quid in the future. Also, a headed bed on PLA is quite common now.



So I got onto fabbing one at work – I was planning on using the rotary table on the bridport mill – but was advised to use a jigsaw and the linisher to nibble and smooth it out. It didn’t take all that long really and surprised myself how close to the edge I managed to cut! There are only a few flats on the circumference, good enough for my needs that for sure :).


Just a shot to show it all really very close to finishing! I also made a plywood insulator for below the bed – so in all its a 10mm thick bed! I had to do something about that as I want as much build volume as I can squeeze out of it. Also in the above image was a bed holder I d/l and printed – but it didn’t fit my 170mm build area! Also the tongue that supports the bed was about 5mm so that just would not do!

I promptly measured around and figured I could drop the bed into the base of the printer – as the bed is only 170 across.

eds kossel bed

So they clip to the frame and sink below on the inside – and the glass bed will be a mm or so proud of that.They’re installed and look pretty good!

ed 1

I Dropped into an electrical shop on the way home today and grabbed some terminal blocks, this allowed me to get on with the Z probe – something that allured me to the kossel in the first place.  It took a while to figure out how the thing worked, and how long to make the probe – but its all installed now and really cannot wait to see it bopping around the bed to get my prints looking sexy!

Now I just have to find a day to go through the lengthy calibration process – I hope the build went well enough now to attain the finishes that made me want to build the machine in the first place! Maybe I’l do that Sunday… 🙂



Last Updated on 14 June 2015 by Mark

25 May

Watering system repair

Some time ago I picked up a Draper automatic watering timer for the greenhouse. The idea was to connect it to the water butt and a micro drip watering system, basically a set of small plastic fittings connected with tubes. There are different types of fittings but the ones that do the business have very small holes, the idea is to place these close to each plant so the water goes to the roots and is not wasted elsewhere and left to evaporate. The timer opens or and closes a valve at preset times. I managed to get this working reasonably well, although I wouldn’t recommend this type of watering system unless you have a decent water pressure, it clogs far to easily with low pressure.

Watering Timer

Watering Timer

Setting it up again for year 2, the timer didn’t work, displays and settings were all okay but the valve didn’t move. I took it apart to do some fault finding. I could get the valve motor to run by hooking it directly to the supply, the micro switch used as a position sensor was fine, so it was down to the electronics. Most of the work was done by an embedded micro controller, with just a handful of supporting components. It fairly easy to trace the circuit and figure out what was going on. the motor was switched by a pair of bipolar transistors, the main ‘power’ transistor was faulty. This was a surface mount pcb and the transistors were SO23 packages, very small devices, especially when I need to use glasses. In my spares box I had a selection of through hole mounted transistors, better for my eyesight but not so easy to use to replace a surface mount component. However I struggled on with an oversize soldering iron and a cheap magnifying glass and managed to get the thing fitted. All good for a couple of years, until now. Once again the timer had failed, although this time the motor was on permanently. I stripped it down again and my replacement transistor was blown. I decided to check the datasheet this time and realized that my replacement, despite being physically much bigger than the original, was rated at 100mA collector current, a cheap DC motor on 5V was likely to exceed that, so I’d been lucky it had lasted so long. I’ve now replaced it with a more suitable device, in an even bigger package. Hoping that will be the last time I need to do this but a good reminder to read the specs when selecting components.

Watering Timer

The top was glued in place but came away with a little light persuasion

Watering Timer replacement transistor

Replacement transistor, somewhat larger than the original.

Last Updated on 3 June 2015 by Mark

16 May

Kossel Mini WIP

A few too many days off at Christmas and too many videos watched on you tube!

I started to put together the list – some key parts I would salvage off my older printer, such as the steppers (£50 for 5) and a power supply (£30). Other than that, will be a new beast!

BOM here:

I miss having an extrusion machine at home – so chose one that not only looks cool, but is small enough to fit a particular space on my desk, so the mini kossel is it!

I planned on taking my time with this project – buying some bits every month to distribute the costs, I don’t think its that expensive, but still – pacing myself 🙂


The BOM suggests that the machine will cost £300.

I did fancy using spectra line (cheaper than belts) and also using a fancy E3D Chimera hot end, you know, the one that dual extrudes with a very tidy footprint. Thing is, that was a £100, and would require me to get another motor and extruder set-up, which I don’t want to do just yet – also, printers are a faff to get calibrated, so I opted to  go for single extrusion, get that working(!) and then go for possibly upgrade when time and budget permits. I would really like to print with PVA support material!

For that reason, I decided to go for a ramps setup. I could have used my sanguinolu from my previous build – but will be fireproofing mself with the additional motor controller. Also, the thing was crazy adorable from ebay so just did it. I also brought the LCD panel but that thing hasnt turned up – its been like 4 months! The seller has also sent two replacements and they never appeared either – ah well!

I cheapened out on the Hot End and brought the E3DLite V6, so inc tax and postage it was about £35 – which is a steal for something as reputable as it is.. The extruder is a geared wades, suggested by RichRap as the best he has used, still. So wont be arguing about that.

I started printing the parts on my works ultimaker 2, and then students got in the way so I have yet to finish printing all the parts – almost there though! I brought the machine home with me this weekend so will be continuing to print through the night.


I am using Orange Fizz from Faberdashery, its printing really well at the moment! Its orange to match the Form1 it will be sat next to – so along with the extrusions they should be a good couple!

I have some white acrylic so one its all built might make some kind of bracket to put two panels on the machine, and maybe get an orange translucent door as well – will see how much I can be bothered. My argument for would be to enclose the machine, good for heat, maybe dull the sound a little….

I also went for the expensive Linear Rails, as I’ve always had a penchant for them.


They cost £75 for 3, so certainly the most expensive part of the build. You can go for an open beam runner style thing, but they look a little ugly and well, would potentially get in the way of the side panels should they get installed ever.

Thing is as of late, whilst waiting for the printer to become available – I got a bit carried away buying tiny cable ties and cool cable braiding – presuming I actually do a decent job of it, it should be the most coordinated kossel out there!



I still have to get some wire – was thinking of getting the wire that is twisted together, like an orange and a white, grey and white etc… not sure that that’s called? Also, I need to get them with female headers on as that’s how I’ll be connecting things to the ramps board (endstops etc)..anyone know what type of wire that is?!



Last Updated on 26 May 2015 by Mark

16 May


A while ago I tried building a Polargraph at a hacker event. It kinda worked… but I ran out of time before I was able to get my head around the calibration. Since then the Ardunio Uno, motor driver shield and motors and sat in the bottom of a drawer, so a few months ago I decided it was time to finish this project properly.IMG_20150516_105012

After a lot of faffing with a related project (Eggbot/Shperebot) and the same electronics setup, I decided to bin the Arduino shield and switch to two stepstick motor drivers and assembled them on a mini breadboard stuck on top of the Arduino Uno.

This new setup was a massive improvement so I decided to design a PCB for it.


Version 1 works but has a few mistakes and gotchas that require a few hacks. I’ll fix these issues in the v2 PCB. The PCB is sort of a mini motherboard that takes two stepper drivers (DRV8255 in the picture) and an Arduino Nano clone (using the CH430G USB to serial chip) and has connectors for the stepper motors, end stops, a servo and power to make wiring up as easy as possible.

With that done, I was able to concentrate on putting together a Polargraph.

There are plenty of Polargraph components on Thingiverse (including my first attempt at the hacker event), but most of them didn’t fit the bill or seem a bit heavy on plastics, so I designed my own motor mounts, pulley wheels, Gondola and feet to suit the parts available from Homebase (Chip board, ball chain and a 25mm dia broom handle cut up to make the feet).



I’ll upload the models to thingiverse once I’m happy with them. Will also upload the spherebot and polargraph firmwares (tweaked for my PCB) to github.




It has taken a while to figure out the shading calibration, but it turns out it is reasonably simple. The Polargraph controller software has a “Test pen wdiths” feature that draws 11 pixels of varying density. In my case i’m using a 0.4mm pen, so I told the controller to start as if it had a pen 0.1mm thick and increase the ‘thickness’ by 0.1mm for each pixel. This has the effect of drawing lighter and lighter pixels. I then picked the darkest pixel that wasn’t completely solid (0.3mm), told the Polargraph software that my pen wdith was 0.3mm and the square wave render of Salvador Dali is the result. (The really dark section top left is where I started this print with the pen width of 0.1mm. It wasn’t showing any difference between the dark and the really dark pixels so I canceled the print and ran this print over the top).


Last Updated on 26 May 2015 by Mark