Category: sidebar


2nd TANR Prototype, workshop ready.

2nd TANR Prototype, workshop ready.

TANR is a tiny AVR based synthesizer. It is not too far from Ljudmaskin, but there are few differences. Most important is that it has more touch contacts. It has three main contacts that are chained with resistors two form additional contacts. I have also added a potentiometer that is used to change modes between different synthesis algorithms.

There are two main influences for my synth. First one is Lumenoise, an audiovisual synthesizer by Niklas Roy. Those of you who have had a chance of trying this thing, know that its synthesis algorithm is a weird mix of control and total lack of control, both at the same time. I wanted to have something that would have the same feel than Lumenoise. This is where the synthesis algorithm kicks in. I have done workshop about algorithmic music players. That workshop was based on concept of ‘one line symphonies’ by Finnish demoscene veteran Viznut.

The circuit diagram of TANR. The circuit consists of Attiny85 and lm386.

The circuit diagram of TANR. The circuit consists of Attiny85 and lm386.

The schematics of TANR show that most of the magic happens in software. How TANR works is that it is using bunch of one line symphonies, switching between them and modulating their parameters based on readings from touch contacts, number of clock cycles etc. I think there is a tiny pinch of astrology in the algorithms as well. With high refresh rates one gets nicely colored and in a way harmonic noises and on some settings effects or even melodies. I’m still in the process of polishing and rewriting the software bit so wont include it here (yet). End result is this:
.

Soothing sounds of TANR.

As in my previous workshops, no PCBs were used. Instead spot board and insulated metal wire was used. New innovation on this workshop was to use a photo with final layout between components and the board to help with placement of the components. I have to say that project like this starts to be at the limits of spotboard technique and might be that after working months with DIMI, I’m fluent Eagle user, could have proper PCBs made in the future iterations. Although I think it is useful to master creating circuit boards from scratch with spot board.

Workshop results!

Workshop participants and couple finished synths!

As always I’m up to giving this workshop few times in the future. This workshop is suitable with people with previous experience on soldering and it takes around 5 hours to finish. Drop me a line if you would be interested in organizing one!

Ljudmaskin. The main components are the speaker, LM386 amplifier, 7805 voltage regulator and Attiny85 for sound synthesis and LED action.

Ljudmaskin. The main components are the speaker, LM386 amplifier, 7805 voltage regulator and Attiny85 for sound synthesis and LED action.

I was invited to give a workshop in Malax by nice people from Stormvind Malax. This was a workshop for children. We spent a busy afternoon building these small noise machines. We had 13 participants in this workshop ages from 10 and up. Arvid van der Rijt and Veera Knuuti were co-directing the workshop with me.

Building of the instrument. This is almost no-soldering workshop, only 9V battery clip needs to be soldered in.

Building of the instrument. This is almost no-soldering workshop, only 9V battery clip needs to be soldered in.

My Ljudmaskin is basically Attiny85 based synth with one touch contact. It is built on tiny breadboard and is run by 9V battery. It has lm386 as amplifier (that I’m now thinking of getting rid of) and a tiny speaker. The main idea comes from drawdio, and the sensing part is done similarly. There is one pin that is programmed to emulate 555 chip. The synthesis is fully done on software. This allows much more complex output textures than what simple squarewave oscillator of drawdio could ever produce. The software is pretty close to what I used in my Synakyna workshop, but at some point I thought that ditching the pencil makes project both simpler and more fun. I have given workshop based fairly similar design few times now and start to be quite happy with the design. But there are still few things I would like to improve for the future iterations.

Ljudmaskin trio performance!

This workshop is suitable from ages 9 and up and takes around 3-5 hours in total. Drop me a line if you would like to book a workshop.

DIMI 1106/1107 sandwhich face side down on my desk.

DIMI 1106/1107 sandwhich face side down on my desk.


This time I will tell you about the first part of DIMI I have finished. This is a unit of two boards sandwhiched together. Top PCB is board 1107 which is the visible touch pad of DIMI. Straight underneath is PCB 1106 which I like to call the bikini board. This is easy to understand by looking at the PCB etch mask. 1107 does not have any electronics in it. Its function is simply to provide interface for the instrument. 1106 has lamp driver ICs for the number plates, pullup resistors for the keyboard, diode matrix for mapping number plates to 4 bit bytes and DIL sockets for three DIP16 cables that will connect the board to main logic board. Drivers for remaining indicator lamps are located in the main logic board. Similarly keyboard tracking is done in the main logic board that is found underneath 1106/1107 boards.

1106 board etching mask.

1106 board etching mask.

DIMI has eight different PCBs (four boards are identical) and I have made EAGLE CAD versions of all of them. I scanned original PCBs, edited images in GIMP, imported them to EAGLE and then did lots of manual tracing. As the original schematics and etching masks are lost, I needed to do everything from scratch. I will probably make a tutorial about this process at some point. Most of the boards I have etched by hand with help from Olli from Olegtron, but all the files are ready to the factory print as well.

Original DIMI-A board from summer 1970. 1106 board is lifted and 1107 backside revealed.

Original DIMI-A board from summer 1970. 1106 board is lifted and 1107 backside revealed.

As seen from this build footage of the original instrument from 1970, I’m proud to say that the end result is almost identical to the original.

Indicator lamps on DIMI are telephone slide lamps. I didn’t have a clue what they were and I still don’t know exactly where and how they are used. I would be happy to add a picture of an actual device where those are used here. But what I do know, is that they are used either in telephones or telephone exchange and that you slide them to some sort of socket. On DIMI they are hard wired straight to the board, so no sliding here. There are many different brands making these, but the company who manufactured lamps installed on original DIMI happens to still exist, so I ordered bunch of these from Taunuslicht.

Installing laboratory hose bits on 1106.

Installing laboratory hose bits on 1106.

Telephone slide lamps radiate light around them so on DIMI Kurenniemi’s design was to use laboratory tubing bits as a shade. I have to admit that it works nicely and gives the light orange color that the original has as well. I found out that you have to use quite tight tube as the lamps are located fairly close to each other. Lamps barely fit inside and I came up with technique of boiling them to soften them up a bit and then using water and electric spray as lubricant. That allowed me to install the hoses without breaking too many lamps in the process.

Boiling laboratory tube bits.

Boiling laboratory tube bits.

Another funny design solution that gives a nice touch is the magnet holding stylus connector on top of the keyboard. Tiny plate on top of the instrument is actually a keeper plate of small magnet hiding underneath the board. Model railway mini banana plugs are installed on keeper. They function as screw-in connector for stylus leads. Most likely Kurenniemi happened to have only one magnet lying around that he decided to use here, as the other DIMI (the one now in Stockholm) had to do without.

Original keeper plate from Helsinki University Studio DIMI-A.

Original keeper plate from Helsinki University Studio DIMI-A.

Sourcing original vintage magnet turned out to be difficult, but I ordered modern replacement from the same company. Keeper plate of contemporary version is much more boring and does not have the cool Eclipse logo the old one has.

magnets

I could easily go on for several pages about the details on this board but I feel it is time to wrap this post up. But at least now it is clear that to actually build a new instrument instead of simply documenting one helps to produce optimal documentation. Already now I have been spending quite a few hours fixing digital files of the bugs and flaws found during the build. If I would build another one, it would be much closer to the original.

To test the board, I hooked it up on Arduino MEGA, did a bit of programming and here you are:

1106+1107 board demo with Arduino MEGA.