Raspi based synthesizer with prynth

Some time ago I discovered „The Mitt“ by Ivan Franko. I liked the idea of a flexible software synthesizer combined with a precise hardware controlled parameter modulation.

The idea is to create different concepts of modulation or forms of synthesis in general. The prynth framework itself is based on the legendary „supercollider“ and is served via a webapp running on a rapsberry pi. That way it is very easy to modify with a few clicks.

I had quiete some different synths and other hardware instruments and always came to a point where I really wanted to modify a tiny bit. I know, being limited at some point can help to develop new styles and they teach you how to efficiently use your ressources but this time we have full control over the bits and sounds. FINALLY

I took part in a competition by Ivan in which he gave away a set of prynth pcbs that are used to feed the hardware signals into the raspi. Looks like I´ve won, together with 2 other participiants and for the first time I received a letter from canada!

 The main pcb has a teensy on it reading the voltage on the voltage dividers made up by the potentiometers. Up to ten variable resistors are supported by the muxi control mainboard. For bigger projects featuring more control inputs the smaller mux pcbs add a maximum of 80 (!) inputs.

Currently the „hack and make Jena“ group is working on creating a mostly practical synth with some additional features to modulate it fluent and intuitive.

Solder fume extractor

A definetely important tool that I´ve delayed way too long. Micha remembered me of the importance of these devices and as he needed one as well I had even more motivation to finally build one.

It is a simple design with a 24V 180 m³/h fan sucking the air through a 130 x 130 x 10 mm activated charcoal mat. The protoytpe on the pictures was designed with two fans in mind but the single arrangement was sufficient enough. The single fan design also reduced the printing time of course, so I went with that as standard.

I added some legs that increased the range of the system so you don´t have to be too close to the „defumer“. The working range is usually around 40 cm. The closer the better.

Material:

5,6mm female plug – Example

120mm Fan 24 V – Example 

24V power source- Example

Flip switch – Example

Activated charcoal mat – Example

MPCNC – New KRESS FME800q tool-holder with dust shoe assembly

First render of the new tool holder with dust shoe, connected at the bottom. The goal was to stiffen the spindle for aluminum milling and the transition to a more comfortable suction system as I don´t like standing next to it for an hour.

The clamp itself has already been printed and the vaccuum attachment is on its way. I will see how this turns out and if I need some kind of brush to bridge the gap to the workpiece or if the pure airflow is strong enough to keep the chips and dust under control. The spindle is closer to the gantry and sits firmly clamped by two M4 screws with washers and nylocks.

Sorry for the bad print quality, it is printed in PETG and I didn´t dialed in the parameters carefully as this is a plain mechanical part with no optical ambitions 😛

Those are my preparations for the nex aluminum milling test runs and the upcoming bigger project that might cause lots of dust.

MPCNC – Long-term testing and material decisions

1. Long-Term observations on the XYZ part

I found a small crack in the XYZ gantry today. This is the second time that I´ve seen this. I milled some aluminum as you may know and I had a mild crash when setting up the endstops but I thought I was fast enough switching it off before it took damage. I printed some parts in the meantime and re tightened the XYZ assembly.

This is obviously a weak spot in the construction, at least if you over tighten the gantry. This is hard to measure as it will break sometime after tightening even without use. It was printed with the suggested high infill and outer wall settings but with a line infill instead of grid. That also could have influenced the partstrength.

The first time it happened on the upper XYZ part. I printed the replacement part out of PETG. It seems much stronger and surprisingly it is not bending the tiniest amount. I expected PETG to be tougher but more flexible but in direct comparison with the PLA part underneath in a tightened state on the machine this is way stiffer. (grey)

2. Conclusion

I´m reprinting the bottom XYZ part with PETG.
Thomas Sanladerer tested the same PETG in his filaween series and he confirmed the toughness of the material. The flexibility of PETG can be neglected in my opinion.

MPCNC filament spool holder

My Hobbyking Fabrikator prints just nice but there are times when the build volume is just not big enough. So why not use the capabilitys of the MPCNC for that? I used the MK8 Extruder and mount from Ryan and designed a compact spool holder that uses leftover parts I had from the CNC build. In my case those are two M8 metric screws with a length of 70mm, a nut and two 608 bearings. Combined it looked like this:Spool holder postion

MPCNC spool holder

The clamp for 25mm tubes is fastened with four M3 *16 and nylock nuts. It can be rotated to aim at the printhead. For now it is available for center holes of 53mm and a width of 53 mm spools.

It can be downloaded here: http://www.thingiverse.com/thing:2002909

 

Desklamp „century“ part 2

I did finish the first „century“ lamp just in time for christmas. The base is very close to the CAD design, thanks to the excellent work of a friend and his lathe. The apple wood turned quite dark, alltough I used a clear finish. The lampshade printed from natural PLA gave it an almost glass like appearance. I´m currently refitting the MPCNC with an extruder so I may be able to print the more complex lampshades soon as my Hobbyking Fabrikator is just a bit too small for the task. I´m pretty happy with the result so far. Now I have to dive deeper into aluminum milling to get it a little more stiff.

 

MPCNC – Quickstart in DE

1. Gedanken zur Materialbeschaffung

Grundsätzlich ist es von Vorteil die gedruckten Teile sowie Fräser und Zubehör bei Ryan (Vicious1.com) zu bestellen. In die Konstruktion und den Service fließen viele Arbeitsstunden. Man sollte zumindest eine Spende in Erwägung ziehen um das Projekt zu unterstützen.

2. Druckteile

Die gedruckten Teile benötigen je nachdem was für Drucker man zur Verfügung hat und der dafür freien Zeit 1- 2 Wochen. Es ist aber durchaus schaffbar und macht einen Teil der Faszination für diesen Router aus. Die Teile sind gut konstruiert und lassen sich auf einem kalibrierten Drucker sauber herstellen, ohne große Nachbearbeitung. Gedruckt habe ich sie mit den von Ryan empfohlenen Einstellungen. Aufgrund der langen Druckzeit empfehle ich bei einer Schichthöhe von Oberhalb 0,2 mm zu bleiben. Die Genauigkeit ist vollkommen ausreichend und die Druckzeit bleibt erträglich. Verwendet hab ich PLA und PETG von „Das Filament“ genaueres dann in der Teileliste. PETG ist zwar zäher aber leider auch flexibler deswegen ist die PLA variante zu bevorzugen, wie ich später feststellte.

3. Mechanische Komponenten

Je nachdem wieviel man ausgeben möchte kann man hier natürlich auch hochwertigere Komponenten verwenden. Ich habe mich für Edelstahl entschlossen da ich auch die Bearbeitung von Aluminium im Sinn hatte.

Die größe des Routers spielt eine entscheidende Rolle. Möchte man Aluminium oder Komposite verarbeiten sollte man unter einer Gesamtgröße von etwa 70 x 70 cm bleiben. Umso kleiner umso stabiler wird das ganze natürlich. Auch die Höhe wird vom angestrebten Verwendungszweck mitbestimmt. Möchte man die MPCNC als 3D Drucker verwenden ist es hierbei jedoch sinnvoll mindestens 10 cm finale Bauhöhe zu bekommen. Ich habe mich für eine nutzbare höhe von 15 cm entschieden. Um Aluminium und GFK fräsen zu können hab ich mir einen erhöhten Arbeitstisch gefertigt, der dann die Höhe zum Mittelpunkt der sich kreuzenden Streben minimiert. Die Standfüße sind natürlich dann trotzdem lang und geben etwas nach aber zumindest das Spiel im Werkzeugkopf und somit die Abweichung am Fräser ist minimiert und die Vielseitigkeit bleibt erhalten.

Die Rohrlängen können mithilfe dieses kleinen Tools ermittelt werden. Simple MPCNC Calc

In meinem Fall hab ich also bei einer Gesamtgröße von 70 x 70 cm eine Bearbeitungsfläche von 40 x 40 cm und eine Bauhöhe von 15 cm. Meine Tests bestätigen dabei eine gute Fräsleistung sowie genug Raum um das ganze als 3D Drucker zu betreiben. Die Werte für die Rohre können natürlich gerundet werden um den Schneidvorgang zu vereinfachen.

4. Elektronische Komponenten

Ich habe die empfohlenen Steppermotoren verwendet. Da hier zwei Stepper pro Achse werkeln bringen diese ausreichend Drehmoment für alle bisher getesteten Materialien und Geschwindigkeiten mit.

Als Steuerung kommt ein RAMPS 1.4 board und ein originaler Arduino Mega 2560 zum Einsatz. Zusammen mit den DRVs ist das auch die empfohlene Konfiguration. Alternativ könnte man ein Smoothieboard verwenden welches ein wenig intelligenter beim Fräsen vorgeht und feinere Abstufungen bei runden Formen ermöglicht. Dafür gibt es allerdings keine vorkonfigurierte Firmware und auch der Preis ist logischerweise etwas höher. http://smoothieware.org

Die Kress FME 800 ist eine gute Wahl als Frässpindel. Genügend Leistung, geringeres Gewicht als bei Beispielsweise Suhner und präzise im Rundlauf. Ich hatte bisher noch nicht beobachten können das die Leistung der 800 Watt Version nicht ausreicht. Selbst bei schnellen Aluminium Schnitten ist die Drehzahl nicht eingebrochen zumindest mit den bisher verwendeten 3mm Fräsern. Die Lautstärke ist recht hoch und liegt etwas Oberhalb des üblichen Staubsaugers aber da hilft wohl nur eine HF Spindel. Zu dem Thema hatte ich mich kurz informiert und was qualitatives und halbwegs bezahlbares konnte ich nicht finden, bis auf die dubiosen „Chinaspindeln“. Dennoch scheint es eine ganze Menge Leute zu geben die mit dieser Alternative zufrieden sind.

Bei der Stromversorgung hab ich mich für die 30 A Version entschieden. Das ist prinzipiell nur nötig wenn man später 3D Drucken möchte und grössere Heizbetten verwendet. Für gewöhnlich reicht ein Netzteil mit um die 10 A (12V) locker aus.

5. Kleinteile und Zubehör

Optional aber hilfreich sind Sachen wie Endstops für Werkzeugwechsel, Energieketten für die sichere Führung der Kabel und ein LCD mit SD Karten slot für PC freien Betrieb.

6. Teileliste mit Link

Hier ein Beispiel zu den verwendeten Komponenten und wo man sie bekommt. Ich habe gute Erfahrungen mit diesen Lieferanten gemacht daher liste ich sie auf. Bitte trotzdem selbst prüfen ob die Teile geeignet sind. Dafür kann ich keine Garantie geben.

Desklamp „century“

It has been a while since the last post and the reason for that is the design and manufacturing of a desk lamp. The initial design differs from the final product as the whole process has been simplified a little. I gathered materials for three of them so the first one will serve as a production test to find out if everything fits. Once finished I am going to include the finer details.

Centuries desk lamp

The pedestal is a 3D cnc piece made from walnut here. It has been changed to a more simple round apple wood base with a 10mm phase cut with a bandsaw. Since the apple wood has a real nice grain I like it even more by now. I made three of them and one is probably going to be machined on a lathe by a friend to get closer to the intended design.

Desk lamp parts

The switch case mounted in front is going to be replaced by an inline switch, this also eases up the amount of needed prints. The joint for the arms above has been redesigned and is now clamping instead of a bolt through design with three possible positions. I have not tested yet if the clamp produces enough force to hold the arms and the lampshade with all the components. I hope this works out well.

desk lmap arm clamp and joint

The arms should be made out of aluminium and I found some nice bars I can make them of. The first step will be to mill them out of some old wood planks I got. That is a good test to see if the dimensions of the parts work together well from a design perspective. The final arms are then milled as intended when the design is validated.

The lampshade is 3D printed and adapted to a regular E14 lamp socket. The threads came out usable on first try and I will keep those first parts.

desk lamp lampshade

The project was started in july and was delayed because I had to build the MPCNC first but I´m almost finished by now. Pedestal and arms need to be cnced and some of the knobs are left to be printed.

Building a MPCNC part 3- cutting

In order to find a suitable endmill I had to understand the principles involved first. That led me to feeds and speeds. The difficult part was to determine the limits of my cnc. I already had a look on that topic when I chose my Spindle. It has a minimum RPM of 10000 and a maimum of 29000 RPM so I decided to go for a 2 flute endmill with a diameter of 3 mm from cnc-plus.de. More on the calculation later..

After testing the gcode and finding out if the steps/mm of the steppers are dialed in correctly I used a simple ruler to check. Josef Prusas stepcalculator gave me 160 steps/mm and that resulted in the correct distance of movement with my 20 tooth pulleys. As I couldn´t  set the parameters from the graphic display I had to reflash the Marlin firmware.

I used a M8 threaded rod like suggested in the original plans. It is not meant for machine operations and so It loves to do what it is designed for. It binds pretty often, resetting the z axis position and making sudden, unexpected deep cuts as shown in the video below. It is the straightest rod I could find and it is well greased but it was necessary to lower the acceleration and speed. I only had this one piece of hardwood and I wanted this to suceed. I´m trying to get the speed back up again because the milling took ages (1h15min) because of the adaptive clearing and the many retractions of the tool head.

The 3D printed rigid couplers didn´t worked to well for me so I ordered a jaw coupler also from http://cnc-plus.de to make it more reliable.

In conclusion : I liked the results of the cuts and they were somehow thrilling. The accidentially super deep cuts showed that I can go much deeper than 3mm in one pass with hardwood. The method used here was also a test for the ultimate goal for cutting alumium sheet and composites.

 

Building a MPCNC part2

So I spent some hours on weekends to advance further into my CNC adventure and it came out pretty satisfying.

img_20160916_170838

Simple CNC stand

Before wiring everything together, I saw the need of a good platform, so I don´t have to wire it again once I found a suitable platform. It took around 6 hours to build this from standard „Baumarkt“ supplys.

img_20160922_1536352

I guided all six cables per axis down to the controller board. I wasn´t sure If I might change the controller board and thus the method of actuation later. I used 8 pole shielded datacable from Lapp Kabel. I added some endstops after I ordered the cables so I have to wire them seperetaly.

img_20160922_153656

Z Axis with Kress FME800 and a pink solid coupler

The M8 threaded rod wasn´t straight and wobbles badly so the Z Axis binds from time to time when running. I´m going to change that for a trapezoidal screw soon. The printed rigid coupler also starts to crack. I need a better part for this too.

Everything else works good so far. The Stepper drivers are getting a little hot but I got a small fan as a leftover from my Fabrikator mod that will be mounted on a 3d printed case or similar.

img_20160922_153710

Complete setup with vacuum and cablework done

Getting endmills and doing the first cut is the next target,