For the Hackathon event last weekend, we wanted to build a robot and decided to go for a kiwi-drive.

The kiwi-drive is a three wheel drive configurations that use special wheels to achieve omnidirectional movement.

The robot was designed in Corel Draw (that’s what the laser cutter prefers) during the two weeks leading up to Hackathon and parts were purchased so that everything would be ready when the build started. We also moved the design from Corel Draw, via AutoCAD for cleanup, to SketchUp for a virtual test assembly and rendering to make sure everything would fit together.

The design was named KiwiRay, from kiwi-drive and stingray as the original design kind of looked like a stingray from above. Those design elements were later removed, but the motherboard mounting plate was shaped as a stingray instead.

The idea behind the platform was to create a design that uses only standard, easily available parts, so that anyone can reproduce it. The exception is the wheels from vexrobotics, but they can be replaced by any wheels of the same type. Other than that it uses an Arduino Uno, NEMA 23 motors, M5, M4 and M3 nuts and bolts and the chassis is cut from any 4mm material – plastic/plexiglass, wood or even metal. It can also easily be redisigned to support a wide variety of ball bearings and motors and has enough space to fit virtually any choice of motors, stepper drivers and even gear boxes if necessary

The Arduino provides a simple serial interface and does all of the required drive calculations. It also provides an add-on interface that can be accessed using the same serial link and allows up to over a hundred expansions to easily be added to the system. The first two addons will be a battery voltage monitor and an emoticon display.

The actual build went smoothly, all parts fit together nicely although some nuts turned out to be quite hard to get attached, prolonging the build time significantly. We didn’t get as far as we’d like, but the entire platform was assembled during the weekend and can now be controlled over a usb-connection that will be hooked up to the on-board computer once that is finished.

See our wiki page for the details.

Tagged with:


SIDblaster/USB is a portable PC SID-synthesizer. Hook it up to your USB-port and you’re good to go. It is designed to work with a MOS 6501 SID. If you need to use a 6500 or 6502, the RNM-0512S needs to be replaced by a 5V to 9V DC-DC regulator.


SIDblaster/USB is partially HardSID-compatible (under Windows) in the sense that it can be used in many programs that support the HardSID by the means of a hacked hardsid.dll. Compatiblity has been verified with SidPlay2, SidPlay2/W and GoatTracker.


The decoupling capacitors (the grouped 100nF, 10nF and 4.7nF) should be connected near their respective circuits:

  • 100nF: FT232R
  • 4.7nF: FT232R
  • 10nF: PIC16F86


Two examples of compact pcb solutions (left: hole mount right: surface mount) that can be mounted in a small case. The AUDIO port needs to be wired to a suitable connector, for example a 3.5mm audio jack. The card is intended to be mounted in a case with tracks for the PCB. The height of the card needs to be adjusted to fit different cases. On the left hand side of the PCB is a USB-B port. You’ll need to open up a rectangular hole in the case for this port to slide through when the PCB is inserted.

Compact hole-mount solution. Verified, functional!

Ultra-compact SMT solution, only ~50% larger than the SID-chip. Not verified!

Component List

RefDes  Description                   @ELFA         @ElectroKit     @DigiKey
FERRITE Ferrite bead [1]              1x  N/A[2]    1x  N/A[2]
R1      Resistor 1k <=5% >=1/8W       1x 60-104-90  1x 40810310
C1,C2   Cer. Cap. 2.2nF 2.54mm >=15V  2x 65-180-62  2x  N/A[2]
C3,C7   Cer. Cap. 10nF  2.54mm >=10V  2x 65-181-87  2x 40512410
C4,C6   Cer. Cap. 100nF 2.54mm >=10V  2x 65-183-69  2x 40512510
C5      Cer. Cap. 1µF   2.54mm >=15V  1x 65-184-84  1x  N/A[2]
C8      Cer. Cap. 4.7nF 2.54mm >=10V  1x 65-181-20  1x  N/A[2]
P1      USB type A                    1x 42-707-40  1x 41002289
P2      Header 6p 2x3                 1x 43-717-12  1x 41001167[3]
U1      PIC16F886-I/SP                1x 73-873-52  1x 40361886
U2      FT232RL                       1x 73-206-59  1x 40359232
U3      RNM-0512S                     1x 69-073-94  1x  N/A
U4      MOS 6501 SID                  2x 43-783-60  1x 41003010[4]

  1. Not critical, may be bridged. Used to remove high frequency interferance.
  2. May have suitable replacement.
  3. Is a 1x40 header that should be broken into two 1x3 headers to replace the 2x3 header.
  4. Is a 1x40 header that should be broken into two 1x14 headers that will hold the SID.

Need to know

The PCB layout is missing a restor on the ICSPHDR between the VPP and VDD pins. This resistor MAY be needed for ICSP-programming using a programmer that requires external MCLR (ICD3 for example). The resistor should be between 1k and 10k (see programmer documentation), 4k7 worked well in our case. The resistor is only needed during programming, once programmed you can bridge VPP and VDD using a jumper. If internal MCLR is used, no jumper or resistor is needed.

Several of the capators in electrokits online catalogue are incorrectly marked in regards to pin spacing. All capacators should have a spacing of 2.54mm, allthough 5.08mm spaced capacators can usually be bent so that they’ll fit.

Depending on what version of the PCB-layout you have, it’s possible the the DC-DC regulator will not fit. This has been corrected in later revisions of the layout. The legs of the regulator can be bent so that they can be connected using a piece of wire.

Shock sensitivity

During shipping (using swedish postal service in a padded envelope) of the first SIDblaster from Sweden to Åland the SID-chip became dislodged from it’s socket. Apart from this and the fact that the CD was crushed it was in good shape upon arrival. If your SIDblaster is to be transported or otherwise roughly handled you might want to consider locking the chip down in some way.


If you are interested in the Windows software (hacked hardsid.dll with MinGW source), PIC firmware (.hex or assembler source) or gerber and excellon files for PCB manufacturing – you can contact me at the hotmail address “senseitg”. Either that, or you could drop by the research department next tuesday!

Completed SIDBlasters

STG’s SIDBlaster build notes:

  • Series 1455 case from ELFA
  • LEDs added for CLK and WRITE
  • ESD transient diode added on output
  • Chassis and USB shield grounded

Tagged with:

Tisdagen den 12 januari, klockan 18:30 kommer det vara en RepRap-maskin, version Darwin, på Forskningsavdelningen. Personerna som har byggt 3d-skrivaren kommer att berätta om hur den fungerar, och hur det var att bygga den.

RepRap står för Replicating Rapid-prototyper och är en 3d-skrivare. Maskinen smälter plast och droppar plasten i små lager, tills man får ett stort objekt av det. Det riktigt coola med RepRap, version Darwin, är att mellan 50-60% av alla delar till RepRap:en, kan RepRap:en själv skriva ut! Detta betyder att maskinen delvis är själv-reproducerabar!

Skriv ut själv

Det hade varit väldigt roligt ifall folk tar med egen-ritade 3d-objekt och vill skriva ut dem med RepRap:en. RepRap:en stöjder filformaten .stl. Spara/konvertera ditt 3d-objekt till .stl, ta med filen till Forskningsavdelningen så skriver vi ut den! Vi har grön och blå plast.

  • Här är en lista över objekt man kan skriva ut.
  • Blender är ett öppet 3d-ritprogam som man kan använda för att rita 3d-objekt.

Vi hade idag planerat att tillsammans tillverka kretskorten som behövs till vår RGBike pov. Vi kom fram till att vi skulle börja med att prova med en teknik där man använder sig av ultraviolett ljus.

Vi hade skrivit ut kretskortsritningen på overhead-plast. Vi lade plasten på kopparplattan och lät uv-lampan lysa på det i 8 min, sedan tvättade vi det i en blandning av kaustiksoda och vatten i ungefär 4 minuter. Efter att ha sköljt med vatten så lade vi kretskortet i etsvätskan som värmdes till ca 50 grader i ett vattenbad. Man kunde nu se hur all koppar utom den som inte blivit belyst av uv-lampan sakta försvann. Sedan torkade vi av korten med aceton. Ack vad fint det blev!

Nästa tisdag tänkte vi prova en annan metod som istället går ut på att stryka fast trycksvärtan från en utskrift av en kretskortsritning, mest för att det skulle vara kul att prova olika tekniker.

Att tillverka kretskort var både enkelt och roligt, dyk gärna upp nästa tisdag om du vill prova!


Ledcube 3x3x3

Ledcube 4x4x4

Tagged with:
Set your Twitter account name in your settings to use the TwitterBar Section.
Creeper MediaCreeper