Time flies by rapidly and I’ve been busy making various things happen for the [Diskokugel] project. I am indeed determined to finish this one this year. I need this off my head ;) So what did I do? In short:
- I got fresh PCBs in from Seeedstudio and soldered 14 of them
- then I did a test with 96 speakers, after equipping all of them with crimped cables. Crimp till you cramp!
- I bought and tested a Bluetooth [Audio Module] and a [Bluetooth Serial Module] with Arduino attached. More about that in another article, though.
The PCBs came out well indeed this time and I was very happy to find that [c-base] has a reflow oven and hot air soldering equipment in it’s labs. This way the first half of what is to be Diskokugels intestines was soldered quite fast. While I was waitig there was plenty of time to cut, curl and strip of the isolation of 200 cables and the crimp and solder them onto 100 speakers. I get my lessons in Zen out of these tasks…
After that it was just a few more power and audio cables to crimp and then it was all ready for testing. Main objectives of this run were:
- check how the circuit acts on power-on
Diskokugel will have 200 amplifiers and they can draw _a lot_ of current. To reduce the current needed there will be measures to activate only a certain number at a time. That is however controlled by a controller that needs to boot up first and there is no control whats happening inside the switches of the individual boards in the main time. It turns out that as long as there is no audio signal, the circuits quiescent current is relatively low. That means I will use a relay to switch the audio in from ground to signal once the microcontroller has powered up. Easier than having to switch power on for one board after another :) 2 boards at full volume used about 1A at 12V. That makes about 150W for all Speakers at once. That also have to run through a voltage divider to provide +/- 6V and a virtual ground.
- check how serial communication works with a longer cascade of boards
The serial signal is put through 25 ics after all and there might be interferences. It works, but I had to set the SPI speed to 1/8th of the Arduinos clock speed. The fastest switching speed for 96 speakers was about 40ms this way. Since I had to lay everything out on a table, I had to use longer cables at some points. Once everything is mounted on the [icosahedron], it might work better.
