29 Sep

Collecting environmental data

We were looking to collect some data that then may be the source for the render output of our other projects: vPiP drawing machine, G-code pen plotter and RGB LED Grids.
A few of us have built portable environmental data sensors with GPS tracking that I detail here.
The thought that these units could then be cycled around the local area or left statically for a period to collect the data.

Link to live map https://makebournemouth.com/geo/geo-enviro-2021-09-15.html

The build

RPI-025Raspberry Pi Zero Wshop.pimoroni.com
PIM458Enviro for Raspberry Pi – Enviro + Air Qualityshop.pimoroni.com
COM1707PMS5003 Particulate Matter Sensor with Cablethepihut.com
TEL0138USB GPS Receiver with 2m Extension Cablethepihut.com
102891Ethernet Hub and USB Hub w/ Micro USB OTG Connectorthepihut.com
SD-16GBMicroSD Card (Class 10 A1)thepihut.com
A1263PowerCore 10000uk.anker.com
The separate components

The Links to the software

Python script running on the Raspberry Pi to collect the data https://github.com/MarkJB/enviro_pi_csv

p5.js sketch to draw the GPS track with a render of the data values https://github.com/ch45/geo

18 Dec

Pewter casting in high temperature silicone

Every once in a while we like to try our hand at something different, so this time it was casting metal in high temperatue silicone moulds.

We start the process by mixing the two part silicone and pouring that over the object we want to cast. Where the object sits flat, we can make a one piece mould and pour the molten pewter into the mould from the bottom. For objects that do not have a flat side, we need to make the mould two sided. To do that we mount the object in clay (sulphur free as sulphur can react with the silicone and stop it from curing) and pour silicone on to the exposed side. Then when the silicone has cured, we flip the mould over remove the clay, apply mould release to the cured silicone and with the object still in place, we can pour the second side. Don’t forget to add some mating features in the clay to make it easier to align the two halves of the mould.

With the mould fully cured we can remove the source object and where necessary cut in a pour hole and air holes (to avoid trapped air). For a two part mould secure the halves together with elastic bands or a clamp.

Next we melt the Pewter bars. We used a stainless steel kitchen ladel heated with a blow torch.

When there is enough molten Pewter, skim the oxiditation (dross) off the surface with a spoon. Then pour the metal into the mould.

When the mould has cooled down enough you can demold the cast Peweter parts.

Remove any flashing and clean up with files or a dremel and polish with a cotton wheel if you want a shiny finish.

20 Nov

8×8 RGB LED Grids – the build

We put on a short fireworks display using our LED grids (see 8×8 RGB LED Grids project for more details. This is how we did it:

Here’s the list of the major items for the build

  • WS2812B RGB LEDs strips
  • Fadecandy module
  • 3mm MDF board (laser cut)
  • Frosted Perspex

Here’s the project outcome.., 8 x 8 LED grids connected up as a 2 down 2 across. All LEDs are illumined with the full brightness test pattern from the Fadecandy server web page. You’ll notice that there is a slight impurity in the ‘white’ because of the different performance of the red green and blue LEDs within the WS2812B. In practice multicoloured patterns are shown so this is not noticeable. Frosted Perspex is used to diffuse the light. A future design may secure the Perspex in a channel within the MDF as no amount of hot glue secured these well enough.

Multiple 8 x 8 grid with a ‘white’ RGB LED signal

This is the rear view showing the LEDs zig zagging across the LED mount panel. The MDF panel is on the shoulders of a narrow piece of MDF around the inside of the frame. LEDs in strips of 50 were purchased so an extra strip had groups of 14 unsoldered to make these up. The 64 were then rolled without twists onto an old ribbon cable spool so that is was easy to unroll flat onto the LED panel. Heavier hook up wire was taken to every 16th LED to cure any dimming of the LEDs at the end of the strip.

64 LEDs zig zag and glued to an MDF board

To make the cubes in to which the LEDs shine 3mm MDF was laser cut into narrow slotted strips, 7 horizontal and 7 vertical. The cube size is about 52mm. MDF ‘staples’ were also cut and within the mount panel you can see 10mm diameter holes for the LED and the staple arms to be accepted. Each LED and staple were hot glued.

Laser cut MDF sections to make cubes

A custom loom was made with heavier gauge (16AWG) hook up wire for the ground and +5V – 64 LEDs take about 2 Amps at full brightness. When soldered onto the pins and the crimp folded over this just fits the JST connector. For mobile use power was taken from a 12V car battery with a 5 Amp DC to DC step down buck converter module adjusted for 5V output. One per 8 x 8 grid and again part of a loom – à la spaghetti junction – manageable with only 4.

Using the Fadecandy module and the Processing.org IDE the project was quite code light.

  • Code to write patterns to the LEDs was just a few lines
  • Libraries for images, video and audio are easily available
  • Processing code and Fadecandy’s fcserver could be a computer that we could leave at an installation (e.g. Raspberry Pi)

Sample Processing.org code for this project is available on Github…

12 Nov

8×8 RGB LED Grids

One of our long running projects was to design and build large form factor RGB LED grids for a Processing day involving algorithmic soundscapes with reactive displays.

Unfortunately due to lockdown that didn’t happen, but the development and construction of the grids continued and during a break in lockdown, we did a guerilla art installation displaying ‘8 bit’ style fireworks and fire displays between Halloween and Bonfire night.

This video is the result

08 Jan


A while back hackaday posted about Trammel Husdons Charliewatch (An analog watch that uses 72 tiny ‘charlieplexed‘ leds.

We liked the project so thought we’d have a go at building some, plus it was a good excuse to try our hand at small component (0603) SMT solder paste assembly and pizza oven reflow.

Case printed on an Anycubic Photon in FunToDo Industrial Black.

With a few cosmetic tweaks to the PCB (the original used numbers but Roman Numerals seemed a bit more appropriate for an analog watch) and a bunch of PCBs ordered from JLCPCB, we assembled a couple to test and to design a new case around.

Checking the PCB alignment…
Reflow soldering in a Pizza Oven!
Post reflow.
And a shaky video showing one of the hourly animations.

Future updates will include updates on final case designs and materials.

More details on github

20 Nov

UnitSeven is no more :(

The local maker space ‘UnitSeven’ is now closed.

MakeBmth are still meeting and working on projects.

Join the slack channel, mailing list or drop us an email to find out more about what we are working on.

08 Jun

UnitSeven Maker Space

Thanks to Daizy.io there is now a fledgling Maker Space in Poole and MakeBmth will be meeting there every Thursday and helping fit-out the space and working on projects.

The space needs your help and support, so if this is something you think could be useful for the community, please come along and lend your support!

UnitSeven Maker Space is located at Unit 7 Birch Copse, Technology Road, Poole, BH17 7FH

27 Jul

Green Woodworking

Through our meetup group, Rob organised a Green Woodworking workshop with Toby from the Green Wood Work Shop! After a bunch of emails, we finally found ourselves wandering through the Purbeck woods trying to find the place, only 40 mins late we could hear the banter and see the smoke rising from the charcoal  burners. We were greeted by Toby and his dog, and a couple of the guys that made it, some as far away as South Africa! (I jest, he lived nearer London!)

Fortunately being late wasn’t that much a problem, I quickly followed what the other guys had started and found myself introduced to wooden mallets, lumps of wood (turns out it was Ash) and hand axes, some heavier than others! We had a few options of what we could make, from rounders bats, garden dibbers, to spatulas.

Some images from the day:

The space


The rather rudimental work holder / lathe


The axes had flat faces on one side, and chamfers on the other, this helped them bite into the wood. Some had both sides sharpened, some had short handles, some had long.


Some rolling pins, a dibber and spatula, some having aged better than others.


Forget what this is called – but it makes the stock as smooth / round as possible before heading to the lathe.


Using a gouge to make a mark for the poppet to be located. I thought the poppet (work holder screw thing’ was a funny name, but we still say, ‘pop it in’ – don’t we?


Guide on using the roughing gouge


workin’ it


Tools of the trade


The beginnings of a rounders bat


venting the hardships of motherhood into this bat


This was a tricky tool to get to grips with, it gave a smooth flat finish, but any lapse in concentration and you’d put a lovely gash in it!



Lots of these on the ground – how many dibbers have been made here?


Rushed this – took about 1.5 hours start to finish – broke a sweat!


Unfortunately I didn’t get a snap of everyone’s finished work! Maybe ill update if I can get copies.

Thanks to Rob for organising the day and Toby for hosting (also, to his mum for the food and cake!)

Was a great day out – I think the potters wheel will be next on the list!


20 Jun

An hour or so drawing


This is what happens when your .SVG export is set with incorrect settings. I left it running for a bit – I thought it looked interesting – also, no pen lift (the horn fell off the servo).

I was playing around with some simple ‘tween’ curves in Rhino / Grasshopper – hoping to increase my proficiency in GH to create some interesting vectors.


The same drawing, with a slightly different home / start position. The red is a STABILO point 88, the magenta is a Pilot V5 Hi-Tecpoint Rollerball Pen Extra-Fine, and the blue is a Eye Micro Rollerball UB-150 (didn’t know that till I just hit up their website!) I knew there was a reason as to why I horde stationary.


The green is a STABILO pointball, the nib is a little large and barely fit through the gondola nib guide thingy, so could have done with a larger hole for it to fit through.


I ended up slowing the speed of the motor accel to a whopping 5 (seconds to max speed) and then I changed the max speed to 25 (half original value). A notable difference in both speed (slower) and resolution (straighter lines (but still not perfect)).



26 May

New build

I fancied making my own VPIP based V plotter – and made some very minor alterations to the cad data (plastic parts) in the process.


Just some of the parts required to build the thing – dog optional.



The plastic parts for the build. Printed on my Mini Kossel in el cheapo blue via amazon @0.2mm layer height.


The first job was to deburr the parts that required it – note the colour change depending on light conditions! The below image is a truer representation of the actual parts.


Bashing the pulley thingies onto the motor shafts


Followed by mounting the motors.


And what is potentially the coolest part of the project (IMHO) is a swappable thing that allows for the pen nib to be stabilised. I have a couple of these with different diameter holes for the different sized pen nibs, eg a sharpie. This should keep the pen at 90 degrees to the paper and not move around inside the gondola. It just needs a deburr in the below image.

And as usual, I got carried away building the thing to carry on taking photos…


Firstly I extended the mounting brackets for the motors – this will allow me to position the endstop encoders later down the line.

home endstop home endstop exploded

These are v1 of the endstop thingies – they will protrude fromt eh front of the machine, but its really the only way I could get it to work ATM. It has Z adjustment with the M3 screw and two captive nuts, and Y adjustment of up to 10mm – having not used these before – hopefully this will give plenty of adjustment to the ball string on the machines.

I reduced the diameter of the gondola so it would print on my mini Kossel (160mm). I also added another screw mounting to hold all the components together.



I added a spacing for the bearing housing  – allowing the shaft to move freely as it was being constrained by the enclosed design, as you can see from the lovely image below  – grey bits are the recess!

bearing bit [Converted]

The weights now have their own printed mounting – V1 worked with a little filing, so V2 I’ll try and fix that!


I finally got around to building the new Hat for the Pi, but that has a few teething problems at the mo – so currently its not moving.


This also has its own electronics mount – but I forgot about the SD reader at the bottom of the Pi, so I had to hack away at it a bit! I aim to make a cover for the electronics at a later date.


I added cable sleeve and that makes it look that little bit better – but the cheap DuPont connectors are utter crap and just pull apart without any prompting >:(.

So, hopefully the config for the UART will get sorted reasonably soon, so I can start drawing with this damn thing! Its been a fun build though, I hope the work on the plastic parts can result in some high res / fast drawings to be created – speaking of which, better get cracking with that!

Some of the beginnings of the drawings…