HOME MADE 2025 Fields Of Results

<--- Back to HOME MADE 2025 <--- Back to HOME MADE 2025 Participants

Intro: Homemade Fields of Results 2025

It´s so interesting how productive the Field of Homemade sourced by Orga Team Magic & Crew Power is.

Concerts & Party was hot summer like. What might got missing a bit at the end of the week, is THE SHOW OF THE MECHA STUFF!

So here the call for documentations, so it's not a orga team thing, but a crew thing honoring Homemade Fileds of dynamic & progress.

Please, feel free documenting your projects, errors & slight progresses to make together a bigger picture of the Field of Homemade 2025 visible.

Just few notes doing are precious.

The Fields of Homemade 2025 and Homemade Documentations

[telegram group(2),"Homemade short and sweet"]

This group is made for doku reasons: to sort and bring together in a quick way, short results notes of the group into a documentary channel. So in a quick fun way, hopefully this becomes a public window for interested ones to catch a glim of the resulting dynamic at SGMK/Homemade. Chronology is not an issue.

Drop and forward to here https://t.me/+xeMpIGpLQuQ2ODJk your links, picts, comments... documenting related projects 💧🩷🪩

#top

[name,project]

(please enjoy writing about your results and fields)

#top

[Hugobiwan, Braillerap]

Important : this content is totally assumed by Hugobiwan as an experiment report without any approval of Braillerap.org project.

Look (or hear if you like vintage loops)

The Braillerap machine (Braille + reprap) is an open source Braille embosser able to translate text, vectors, and open street map data into relief dots that can be touched by blind people to read and feel informations.

The whole project is documented, and it is supported by NGI commons, MHK and Climate Change Lab organizations. The main developper of the project, is Stephane Godin.

We brought a Braillerap machine to homemade and first had to repair it as two pieces were missing or broken :

an O-Ring that is being used to catch and push the sheets of A4 paper.
a 3D printed part that is used to keep the sheet of paper on an horizontal plane.

With the help of the great SGMK team, we could quickly find a o-ring in the 2 tons material reserve. Also we had a replacing part from version 1 of the machine.

Repairing lessons

First lesson :

if something mechanical is broken or missing under the paper plate of a braillerap machine, you will probably have to unmount half of the machine to replace it. It will last half a day to be properly made and finished. So take your time. It would be interesting not to glue the paper plate when building the machine, if it can be properly held in place...

Second lesson :

This machine had been made by students of the University of Rennes 2 (honour to them !). We discovered that they replaced an iron bar with an aluminium rod, for the longest one that goes through the whole machine. As this part is blocked by two scroll wheels and screws, it had been crushed at both ends and it was very difficult to take off the enclosure, in order to repair. Conclusion : if it is possible, use ONLY iron rods, or be VERY careful not to crush aluminium rods or you won't be able to maintain or upgrade properly your machine. In fact, using too much force can mean breaking the enclosure. And the enclosure defines the global geometry of the machine (through axis holes...).

Silhouette hypothesis / Urs Gaudenz

At the next table of the Braillerap project were great neighbours : Gaudi labs people and Tara, sewing around. They were using a multi tools digital fabrication machine named the "Silhouette". Urs Gaudenz suggested that it was possible to emboss Braille using this machine. After having looked at the technical manual, no doubt this is possible. The machine takes vectors as inputs, and it is possible to change the toolhead at the condition to adapt it to the machine. Even if we shall lose a lot of the Braillerap softwares functionnalities, generating Braille in vector format (without the gcode step) and then send it to the silhouette equiped with an embosser tool would work.

In july 2025 Babette Sperling and Mario Voigt demonstrated in Fablab Chemnitz a way to make quickly an embosser head by using an electric drill, sand paper and protecting glasses, during our Braillerap mounting workshop. The principle is to sand a grub screw that is turning fast in a sand paper sheet. Then we get an embosser head that could be adapted also for the Silhouette Machine. This is a great reminder of the fact that the Braillerap project was originally born from my idea to transform open source 3D printers (repraps) into braille embossers. The silhouette community is big and could be also interested in accessibility.

Thank you Urs for this input !

Materials and torture prints

Materials

The Braillerap could successfully emboss on aluminium from soda cans, and copper tape. Copper tape is very interesting as it can be taped after having being embossed. The proper way to prepare materials is the following :

Prepare and cut a rectangle of your material (alu can, other)
Use tape to glue your material (on its sides) on an A4 or A5 paper sheet (depending of the size of your Braillerap)
Prepare the content of your page with dekstop Braillerap open source software. You can define an embossing zone in the page to preview the result.
Input your sheet of paper into the machine with me material upside, like a sheet of paper.
Emboss !

Then, AFTER, take out the sheet of paper, take out the material and use it elsewhere (clothes, door to label, place...).

Torture print lessons

Torture print in Desktop Braillerap sofware

We had the luck to present the project in the "yes room" that was full of great electronics and code prototypers. Some sent us a strange vector file that was clearly impossible to translate and emboss in Braille in a A4 paper sheet.

Because it was a complex electronic scheme with all pins labelled in very small typo and the braille must respect a standard physical size
Because multiple layers where mixed everywhere on the page in a whole svg.

But we decided to launch this torture print anyway. Important detail, as we did not have required paper to emboss well relief dots (at least 180 mg/m²) i came to the Kids room where Salome had paper for everybody to make drawings. I took some classical A4 paper sheets.

As the paper wheets where too thin, i decided to take 4 sheets, and to put them like one sheet of paper in the braillerap. Then we launched the torture print with the electronic scheme on the machine.

This was very interesting torture test.

Results :

It seems that in certain conditions (4 sheets of "classical" A4 paper) the braillerap can emboss at least 4 sheets of paper at once. This is important for people that are making transcriptions for students, in order to gain time. The great mechanical Perkins embosser can emboss several sheets at once, but requires a lot of strenght and training to do so. Of course this paper is not enough solid to last. So we could emboss, then glue the sheets on stronger supports (cardboard, strong paper...).

All the labels that were written in the .svg layers for the integrated circuits were automatically translated in Braille as can show the preview of the great desktop Braillerap software... So no need to prepare all you document by separating text from vectors, and then creating them on Desktop Braillerap. You can prepare very good documents directly in Inkscape or a .svg editor, but you must prepare them managing with the Braille size, and using organized layers for your texts, and other layers for your graphics. This shows also the strenght of desktop braillerap software, that makes that easy for anybody.

The intricated texts translated into braille on braille made the machine crazy : the rythm of embossing was totally accelerated (the machine receives Gcode corresponding to braille that was written on braille + parts of vectors). It began embossing, then the current was cut (a feet on a wire :), alignment between tool head and other part as suddenly lost, and ideas came...

We exchanged (with Sammy of Canvas synth) about the mechanical improvement of the machine, because clearly the actual system can't manage such a speed (it was like embossing cloud of dots to create grey levels...).

Hypothesis :

When the tool head that types a dot from the bottom of the machine is pushed up, it is by a strong electromagnet. It can work at important speed.

But when the solenoid comes back in the bottom position, it is just pushed back by a loose spring and gravity. This system is too loose and too slow.

It would be possible to imagine a whole rubber surface at the top of the machine to remove the upper carriage (Sammy hypothesis). In fact things are not so simple because to make a good braille dot you should push your paper into a round shape that is upside. But we could also imagine to make tests or to build a dedicated 3D printed shape exactly calibrated to receive the tool head movements in a kind of matrix.

It would be possible to created a mechanical system that strongly push back the tool head quickly. For example using a H bridge with a polarized electro magnet in order to push it back.

This is one more time a lesson about the interest of being with creative people : embossing torture prints can teach a lot !

#top

If we Listen, Do Microbes Sing our Song - Nora Hauswirth, Kaspar König, Maya Minder

Together for three days, we attended Homemade to prepare a sound piece on the topic of Soil Microbes and One-Health for a participative performance accompanied by a sound installation in the botanical garden of Zurich. We collected the many different voices from Homemade, a total of 17 different languages and countries who were attending homemade this year and made each one recite a sentence from a the Soil Assembly Manifesto. We also recorded extracts from Simon Berz Sound performance and Ralph Schreibers tin can music instruments and mixed it together with sounds from Kaspar and Maya, Nora donated some recordings from the Amazone rainforest.

„If We Listen, Do Microbes Sing Our Stories“ ist eine Soundinstallation und somatische Performance als Annäherung an das Mikrobiom des Bodens. Am Anfang der Performance steht ein gastrosophisches Experiment; eine Vereinigung zwischen dem menschlichen Mikrobiom und die des Boden. Durch die Bewegung im Raum verbinden sich akustische Aufnahmen aus dem Amazonas und verschiedensprachige Stimmen. Die Performance basiert auf Texten der Soil Assembly in Indien und Manaus und erkunden Bodenmikroben über eine kollektive Erfahrung. Sie bringt unterschiedliche Ökosysteme – von tropischen Regenwäldern bis zu städtischen Gärten – sowie verschiedene Stimmen, Sprachen und Perspektiven zusammen. Sie vereint menschliche und nicht-menschliche Stimmen in einem gemeinsamen Klangraum. Der Boden ist Anfang und Ende des Lebens, ein lebendiger Raum, der Geschichten, Kulturen, Beziehungen und Träume trägt und sowohl für menschliches als auch nicht-menschliches Leben essentiell ist. Bodenmikroben nähren die Biodiversität und unsere Kulturen. Ähnlich wie das Wort Agrarwirtschaft im Deutschen verbindet das englische Wort "Agriculture" die Kulturen im Boden mit der Menschlichen. Die Erhaltung von Lebensmittel, Kochen, Sammeln, Dreschen, Fermentieren, Einmachen, dies alles waren kollektive Arbeitsprozess, welche den Menschen evolutionär zum Überleben verhalfen. Der Akt der kollektiven Arbeit ist tief in den Böden eingeschrieben. Unsere Verbindung zur Erde ist sowohl spirituell als auch kollektiv. In einer Welt, die durch Fragmentierung, Trennung und Ausbeutung geprägt ist, ist die Wiedervereinigung mit dem Boden entscheidend für individuelle, kollektive und planetare Gesundheit sowie für territoriale Gerechtigkeit und Widerstand gegen moderne Kolonialisierung. Daher ist der Schutz des Bodens zugleich ein Kampf für Rechte, Vielfalt, ein erfülltes Leben und tiefgreifende gesellschaftliche Veränderungen.
Die Performance und Installation sind Teil der Performance Serie "Botanical Memories: A Performance Series", von WE ARE AIA in Zusammenarbeit mit dem Botanischen Garten der Universität von Zürich. Das Programm erstreckt sich über den Sommer 2025 und umfasst Aufführungen und Workshops, die sowohl unser traditionelles als auch unser sich entwickelndes Wissen über Heilpflanzen erforschen.
Ort: Tropisches Gewächshaus des neuen Botanischen Garten Zürich
Sound-Installation Samstag, 15. und Sonntag, 16. August 2025 (60 Minuten) während den Öffnungszeiten (9.30-16.45 Uhr)
Performance: Samstag, 16. August 2025, 17 Uhr (90 Minuten), Sound, Bewegung, Soil Capsules, Boden-Kaviar molekulare Gastronomie. Anschliessendes Apéro mit Amazonas Kombucha und fermentiertem Gartengemüse.

Idee und Konzept: Maya Minder, Nora Hauswirth, Kaspar König
Sound Design: Kaspar König
Text: Kollektives Manifest Soil Assembly Manaus, Antti Salimen 50 Words of Soil, Nora Hauswirth
Recordings und Sprecher*innen:
Synature: Aufnahmen im Amazonas von Flora und Fauna
Musik: Kaspar König, Simon Bärz, und Ralph Schreiber
Stimme 1: Nora Hauswirth, Stimme 2: Maya Minder, Stimme 3: Lina Lopes, Stimme 4: Gabriela Costa Braga und Patrícia Yusió Weitere Stimmen:
Christian Bobst, Ana Victoria Bruno, Wolfgang Dorningen, Quentin Aurat, Monika Pocirnic, Jena Jang, Hiro Ohara, Chrysa Chouliara, Yair Reshef, Shih Wei Chieh, Oliver Jäggi, Felix Bäntli
Molecular Food Design: Maya Minder
Food culinary: Nora Hauswirth, Maya Minder

Sprachen: Deutsch, Englisch, Französisch, Griechisch, Hebräisch, Japanisch, Koreanisch, Italienisch, Manadrin, Portugiesisch, Schweizerdeutsch, Slovenisch, Spanisch

Performance Information at Botanischer Garten Zurich

#top

[SaLome, projects: installation, workshops, concert]

It's my first wiki page, I set up

Makes appetite to do more pages for what happend in this week about my:

installation
workshops
concert

yes, to describe the learning in this week, is hardly possible, but I try this this way.

#top

[Jonas ↯, Jet Engine Based on Car Turbocharger]

TL;DR Not self sustaining, presumably combustion chamber too small and/or initial air flow too low and/or 30 kW heat input is too low.

Some facts:

Built for demo reasons and tests with compressor bleed air, no (usable) thrust intended
Radial type engine from abusing small automotive turbocharger with test certificate up to 180000 RPM intended for ~120 HP gasoline cars
Initial air by repurposed old vacuum cleaner
Fuel is 4 bar propane gas from gas phase
Burner is 30 kW roofing torch running on that propane, flameout prevented by torch itself
Lubrication with 3 bar (IIRC) gear pump on current-limited 12 V pumping wide range motor oil
High voltage ignition by Mazzilli ZVS + CRT flyback transformer on 24 V / 4 A (extremely reliable <3 )
Tried to minimize welding, welding only required in 2 places xD

Basic continuous media flow: Cool air sucked into compressor inlet of turbocharger ==> air is compressed by compressor wheel of turbocharger ==> compressed air flows from compressor outlet into external combustion chamber ==> the air is mixed and burnt with propane gas in the chamber ==> 1000 °C hot exhaust gases exit combustion chamber and flow into turbine inlet of turbocharger ==> the hot exhaust expands and drives turbine wheel of turbocharger ==> the exhaust gases exit from turbine outlet into surroundings.
Note that media flow is continuous as opposed to discontinuous flow in pulse jets and normal pulse detonation engines. That means compression is also continuous and during operation a steady stream of air is forced through the combustion chamber. The harder the compression the stronger that stream.

The thermodynamic process is called "Brayton process". Work is extracted in the turbine by isentropic expansion of the exhaust, and partially expended in the compressor to perform isentropic compression of fresh air. Between compression and expansion isobaric heat addition takes place. A small part of work can be diverted for external use.

Expected behavior: Initially air is forced through the system by means of the old vacuum cleaner's air outlet. The roofing torch has a pilot flame that requires less air than the vacuum cleaner is able to force through the entire engine. The air is turned up until a point where the pilot flame is expected to survive. The high voltage ignition is enabled which creates a hot plasma arc inside the torch. Such an arc is like that from a welder and keeps burning until the ignition is disabled. It also does not extinguish from strong winds like that from the vacuum cleaner. After the arc is burning, the pilot gas is turned on and the arc will immediately ignite the gas. The pilot flame stays alight without high voltage ignition so ignition is disabled again. Once the pilot flame burns the vacuum cleaner is turned up to maximum power. On the press of the torch handle main gas is added and the pilot flame becomes the main flame until the handle is left. This flame's energy spins the turbine of the turbocharger.
In the turbocharger both the turbine wheel and the compressor wheel sit on the same shaft. The faster the turbine wheel spins the harder the compressor compresses new air into the combustion chamber. For every amount of air only a certain amount of fuel can be burnt and thus converted to thermal power. The more air is compressed into the chamber the more fuel can be converted to heat in a given timespan and the faster the turbine will spin (and the more air the compressor will compress...).
After having heated up sufficiently the process enters the positive feedback region which from my understanding is (nearly?) identical to the self sustaining region. That is the time to remove the vacuum cleaner as from that point the engine will feed itself enough air for the required flame to survive. Initially the gas/air mixture might be rather rich resulting in high flame temperatures and exhaust gases exceeding 1000 °C. Even if the turbocharger is designed for gasoline cars, which are prone to higher (900 to 1000 °) exhaust temperatures compared to Diesel ones it won't tolerate prolonged exposure to >1000° C gases. With more air delivered by the compressor the hot exhaust will cool down more and more until a lean equilibrium is attained. Increasing the fuel flow will increase the gas temperature again. In this operating region every additional gram of fuel delivers more power than what is needed for sustaining the current compressed air flow. That netto power can be harvested as mechanical shaft power, bleed air or thrust. Note that harvesting bleed air will lead to less air in the combustion process thus increasing the flame temperature thus stressing the turbine more

Observed behavior:

Extremely reliable ignition without any deflagrations whatsoever
Serious flame emission from the turbocharger's turbine outlet when on half to full power with red glowing turbine wheel
Small or no flame emission and turbine wheel glowing orange when on partial power
Noticeably attempting to build pressure on partial power but failed attempts to remove initial air (the vacuum cleaner) while running
No positive feedback
Very unreliable lubrication with no damages to turbocharger despite multiple failures in lubrication system

Conclusion: Without afterburner there should be no flame emission into the surroundings. The result indicates that likely the combustion chamber is too small to contain the whole flame that is generated by the roofing torch inside. If in free air such flames typically measure 50 to 70 cm in length and have a diameter of 15 to 20 cm. The small diameter of the chamber may additionally have led to a too high gas speed inside such that the gas hadn't finished burning by the time it had exited the turbine. This should be addressed by a thicker and longer chamber, i.e. a much larger fire extinguisher. If the start persists to fail despite a larger chamber a better alternative needs to be found for the vacuum cleaner. As a last resort liquid fuel needs to be injected into the chamber using a vaporizer coil wound around the chamber. According to rough estimations by ChatGPT IIRC only 10 ml/s ethanol should be necessary to add some 100 kW of heat.

Perhaps this engine has been building substantial pressure for a short period of time because a non-screw but still forced mechanical air joint was found slightly displaced/separated after the first notice of pressure build-up. No further substantial pressure was noticed after fixing the joint.
To be continued !!

Update 2025-10-02: Back in Zürich a second unsuccessful run was performed with an improved air delivery system, added combustion liner and improved oil lubrication. On the test run the turbocharger seemed to pick up roughly double the pilot flame RPM (24 kRPM) tapering off at about 48 kRPM and boost pressure seemed to rise slightly. However, the turbine wheel became incandescent quickly, ultimately leading to some damage to it. Due to the small combustion chamber far from 30 kW heat could be added before combustion would take place outside of the turbocharger. This capped off the usable heat input and RPM, and likely prohibited any self sustain.
Reading several student project's papers and skimming through the Jet And Turbine Owners forum suggests that about everything which could be done wrong has been done so. This includes having overly complicated plumbing with numerous sharp edges inside leading to a lot of flow inefficiencies. In addition both air and exhaust were escaping in various places judging from the aggressive hissing when the pressure rise took place.

#top