AVR Synth RE-re-packaged

Went today to the electronics shop to get some cheap stuff for the yusynth and saw this box for 6 euros. From what you can see, I got it. It is relatively small with a panel area of 15x9cm.

So I went on to repackage my avrSynth16 I had build sometime ago.



There was a failed attempt to house the synth inside a FIRA drawer from IKEA but the design had some flaws (And it was way to big to have it around).



The first step was to make a 15x9cm layout in inkscape and make objects that were the size of the underside of the components I would use. So I could layout the knobs and switches knowing that they'll fit. Purple were the knobs, cyan were the switches.



I made the holes with my newly aquired tools that I mentioned in earlier posts (the step drill that is) and had the panel ready for sticking the graphics.



Since the step drill went up to 12mm diameter, the midi socket hole that needed 15mm diameter, got a little help from my old (t)rusty reamer. (I always had that tool! It belonged to my father from what I vaguely remember)




The simple design is always the best. Made in inkscape. (Super drawing program, open source!)


The time of the knob transplant was stressful. A few wires came loose. (The design is not so clean from what you can see)



So it was obvious that it would end like this.. A big spaghetti bundle.

The hardest job was to fit the boards inside the box. It took a lot of careful squeezing and pulling/pushing wires inside the box. This was done while the synth was playing from a sequencer. It kinda reminded me of some brain surgeries you see on TV where the patient is placed on the bed with an open head, and the doctors have electrodes touching his brains to see what is the problem. When the synth stopped responding I stopped pushing!!

You can see that the panel has a slight bulge at the center..



And now the publicity shots!

Side projects

This is an octal RGB clock. I made it so it can be impossible to actually tell the time but nevertheless its beautiful (in its own way).

There is an atmel AtMega8 µCU and 6 RGB leds. Two for hours, two for minutes and two for seconds.

Each led can be Red, Green, Blue, or a combination of them. So you get eight colors, Red, Green , Blue, Magenta, Cyan, Yellow, White and Black (all off)

Lets say now that a led (with all its colours) is a binary number and Red is bit 0 (rightmost), Green is bit 1 and Blue is bit 2. With the above combinations you get a number between 0 and 7 (hence the octal part) . Two leds can count up to 64 this way. So you can count easily up to 60 seconds (and minutes)

eg: the time on the second photo is:

Red, Cyan : Green Cyan : Red Green=
1x8 + 6 : 2x8 + 6 : 1x8 + 2 =
14:22:10

Nice eh?

(wouldn't sell though : )

Making of the UV Oven

To start, I had to make the array permanent and to such dimensions so as to have even lighting all over the surface of a typical photo-resistive PCB. So I laid them as shown, taking into account the distance of the leds which was enough to have overlaping beams of all leds. (so no blind spots where created) .

This was put in a ikea FIRA box, slightly modded for the purpose. (I left only the bottom drawer, didnt put the mid panels and used one of the mid panels as a swing door, attached to the top of the box with safety belt ribbon, glued and stapled to the box and door)



The first test consisted of burning two photoresistive boards with components needed to make MIDIBOX sequencer v3.3 . It was a total success. as it can be clearly be seen.





Before going into burning the midibox sets , I performed some tests. The results were as such.

* With the current setup (35 UV leds (UV =395nm) at about 20mA each from a distance of around 30cm) the board needed around 15 minutes to be developed
* Reasonable results were between 10 and 20 minutes
* The film used was simple inkjet transparency but I would print it TWICE (my printer is accurate enough) or use two printed transparencies one on top of the other


They were then developed using the standard procedures (NaOH in water for removing excess photo resist and then FeCl3 or Sodium persulfate or whatever you are used to)

The results were AMAZING. The Oven was (and still is) one of the most important tools in my "lab". (The timer was dumped, I had a watch and pulled the plug)


PS. The array of leds needed quite a lot of current/voltage and I was lucky to find a hewlett-packard printer transformer with 1A/37V output.. Don't test THAT with your tongue!

First (UV) rays of light

This is a test:) I had to make an array of UV leds and calculate properties. I made a simple excel file (openoffice format) to calculate these things here .

Board oven timer

Not exactly synth-DIY but this is a timer I made for my soon-to-post PCB oven. You just set the time and keeps track of it, switching it off when it must and securing it from accidentaly opening the lid, saving my eyes from UV burn :)

Pseudo-digital oscillator

An old project of mine (6-12-2008) was to have a counter run through a series of voltages and make a waveform out of it. There is a pulse generator and a counter at the bottom part of the breadboard and decoders at the top. It never worked right so it was dropped. I tried to use the TINA simulation software to have a working version at least theoreticaly so if anyone wants it just drop me a note...

Training day (night)

Got a PIC (microchip.com) programmer off ebay (I wonder what I would do without ebay here is its anthem).

It is a el-cheapo clone, (cheapest USB programmer I found) able to programm only 8-bit microchip.com microcontrollers (such as the 16f84 and 16f877 and 18f452 that the good guys at microchip.com sent me as samples). I tried almost every language available for them (like picBasic pro etc) but settled for microchip's own compilers (namely C18) which also comes in an educational edition that has optimizations disabled. No problem there. Since my C background is much better than Assembly, I felt right at home