Trying to make Diskokugel louder

November 1st, 2011

After my first experiences with the [Diskokugel] circuit building [Audioweb] and [AudioSphere] I tried to optimize the circuit for further use. That means I tried to look for a way to get some head room with the volume, since the current design didn’t quite satisfy me volumewise. I discussed the problem with Christian and we did some experiments using a [TDA7053a]. Not only would the TDA make a higher amplification available. Working on 5V instead of +/- 5v, as the transistor circuit works right now, it would have rid us of one power line that has to be transferred into the turning ball via a sliding contact. Also, soldering two chips on the board would have been much faster than soldering those 16 transistors.

The TDA7053 has a DC volume control input, which made us hope that we would simply connect it directly to the 4094 register to switch it on and off, making the 4066 switch obsolete. Unfortunately that didn’t work. The 7053’s offset voltage rises with the applied control voltage, thus producing a crack at the output when raised rapidly. Ultimately we got the thing working by keeping the control voltage at ~0.7V and switching the audio signal on and off with the 4066. However, since in this version all amps are constantly active, I am afraid the whole system would be prone to interferences and the power consumption would be very high. I had a very hard time testing the whole setup. I don’t know if it was my breadboard or whatever, but something started oscillating regularly and there was never a setup tat was save from that or other interferences. Christian told me however that he got something working.

I am not sure how to proceed yet. I think I will stay with the old circuit for now and see how far it takes me. After all, one result of the experiments with the TDA was that a higher volume makes it harder to locate the source of the signal. That would be sort of counter productive. It also shows that there are possibly limits to the use of this whole technique…

Audioweb 0.1

September 5th, 2011

As a special gimmick for my birthday party i used the elements of [Audiosphere] to build a web o 6×6 speakers on a wall in my room. The sensation is, of course, a different one and since its now totally at hand, I will do some more experimenting in the next weeks.


the cable length of the sphere didn't fit, so it doesn't look perfect yet. I still like it :)

AudioSphere for PDCON

August 14th, 2011

Having introduced the [Diskokugel] project at Berlins biweekly [puredata] [user group], I was asked to build an installation for the finalizing event in Berlins [LEAP] gallery. Being busy doing technical coordination for the [Über Lebenskunst Festival] this was quite a task.

I used the five prototype circuits to build a walk-in sphere of 40 speakers. The speakers were attached to nylon strings stretched from the ceiling to the ground, so you could bend them aside to step in. In an effort I provided a crude interface in PD that provided a joypad for people inside the sphere to select different programs, such a random, sound going in circles, etc.

Contentwise, we had placed a call for content, which resulted in one piece that was handed to me and which worked quite well. I have yet to get the adress of the composer to thank him again ;)


The result was quite to my liking and so was the feedback (apart from nobody understanding the interface, but in the end that didn’t matter much). The sphere was fun to use. I personally found it pretty dizzying standing inside the speakers while the sound came from a different direction every 25 milliseconds. Unfortunately, as of now there are no pictures to show. The  installation stood at LEAP only for two days, in which I had forgotten to bring my camera. Hopefully I will find someone of the PDCON who made good ones.

[UPDATE: pictures of the building process and a picture from a mobile phone (thanks ckp and astring!)]

a picture by astring

in action. the nylon-strings could be bent apart to step inside. (pic by ckp)

arrangement of the circuits

some notes

more notes


July 8th, 2011

Diskokugel is a work in progress, hopefully soon to be finished. For the whole story and latest updates check [here]



I started the project “Diskokugel” in autumn 2009. The vision was to build a ball made of speakers that can be turned on and off individually. This ball would turn while sound would be “reflected” in one constant direction, just like light from a mirrorball. So I went to the [sublab], a hacker space I am co-founder of, and asked for help. I got it galore from Christian, who volunteered to design a circuit that would both be cheap and easy to build.

more… »

A new Gadget 2.0

July 6th, 2011


Behold! I finally upgraded my [bike sound system], adding extra power, a subwoofer and a battery charger for everlasting supply. It doesn’t look as decent as the last one did but it totally rocks. The first test run was a trip to the [Fusion Festival], around 120 kilometers. The dynamo provides enough power to reload the batteries while the system is playing.

I started of with the [TDA7052] 1W amplifier I had already used with the old sound system, only this time I used two, one for bass and one for mid/high frequencies. But the TDA is made for 8Ohm speakers and my bass speaker had 4Ohms and so had the two high/mids in parallel. So I decided to get something bigger. Now I have an [AN7148] delivering 2×2,5W at 12V 4Ohm which is totally sufficient to predominate the occurring head wind and most traffic noise :)

The battery charger was built according to the circuit provided by [] . Documentation is in German only but it’s extraordinary! It’s housing has still to be completed. To be done in time for the Fusion Festival I was a bit in a hurry.

The AN7148 in its housing. Batteries (9xNiMH) sit at the right.

The battery charger

The battery charger


1,2,3 test…(3)

June 26th, 2011

I found the time to go to Leipzig and use the [sublabs] oscilloscope on my [Diskokugel] circuit. Actually Nobody did that, assisted by Christian. It took a while to get it to work and the result was both satisfying and not satisfying. Because there was nothing to see. The crackle we hear when switching a speaker on and of did not show on the oscilloscope, which means the circuit works perfectly, and leaves us clueless as to the source of the problem. I did some more experiments but still have no idea.

Well, I might just leave it at that. Since the crackle is only really a problem when switching extremely fast or when a pure sine is used. I need to get this ball done now…

nothing to see here...

Keep it moving

June 21st, 2011

My DJ-Blog is getting dusted, so I decided to move it here. That won’t change much in the sense of missing content, but there is one less Blog to nurture ;)

However, soon there will be a “Download” section for my mixes and other stuff, such as puredata patches etc. I’ll keep you posted.

1,2,3 test…(2)

June 21st, 2011

I’ve been quite busy in the last weeks:

I presented my project at the [PD user group] in Berlin and got some interesting feedback. They would like to exhibit the Diskokugel or an installation based on it at the final party of this years [pd-con]. I feel honored but I have no idea if I can actually make it until then. Money is also an issue here. Having all the circuit boards made will cost…

I solved half of my clicking problem by getting the synchronization with the sound phase to work. Still, clicking occurs and since there was no oscilloscope at hand, I recorded it to make sure it’s not some kind of acoustic illusion due to differences in speaker loudness. The waveform clearly shows a distortion when the speaker is activated. When it’s turned off, something produces a sudden click and then the speaker swings out.

I will meet with Nobody and Christian (god praise his patience!) at the [sublab] this weekend and we will check the whole setup with an oscilloscope. I hope that we can fix this easily.

The noise occurring when a speaker is activated

When the speaker is turned of, something gives it an extra push.

1,2,3 test… (1)

May 5th, 2011

Today I did the first test run with one of the soldered boards and the Arduino. After some adjusting it works but I am not quite satisfied.

On the picture below you can see the setup: 8 speakers, attached to the switching board. Between the switching board and the Arduino is a board that converts the Arduinos signals (0/+5V) to the current level used by the switchboard (+/-5V). This converter board also houses a TL072 for pre-amplification and an LM311 to detect the audio signal’s null phase. This is necessary to prevent the speakers from clicking when switched on and of. And it doesn’t work jet. Why, I’ll have jet to find out…

Speakers, switchboard and Arduino connected. To the right is a voltage divider, converting 12V to +/-6V and GND

Burning a new bootloader into my Arduino Duemilanove [solved]

April 27th, 2011


I finally made it happen!!!!

After two days of suffering I found a thorough explanation, how to burn the Arduino bootloader on an Atmega328. This is the solution.

The 'self programming' Arduino


I wanted to start testing the freshly soldered boards for the Diskokugel when I had to discover, I had made a big mistake simply using the Atmega328 of my Arduino in my LED lamp. It turned out to be not quite as simple to replace since the Atmega needs a special bootloader to work in the Arduino board.

There are several ways to burn that bootloader into the chip. The easiest would be having an Arduino and programming it to do it. But since mine was without working chip, I had to look for other solutions. I found one that uses the FT232 on the board to program the Arduino itsself without an AVR writer. However, the solution programmed there is for Windows only. There apparently is a way to do it with a Mac but I wasted a whole day on trying to get it to work. It’s really tiring, installing package after package due to dependencies, having to compile it all, etc. I ended up compiling packages and, when I had finally done that, having forgot what other package I actually wanted to install in the first place.

I’ll try again tomorrow, taking the chip out of the lamp and putting it back into the Arduino. If the Arduino-as-AVRISP works, Ill simply burn me a stash of Arduinos…