Home ➲ idea update ➲ candidate: VIA EPIA N-Series Nano-ITX Mainboard
thanks to Coocha for the suggestion:
VIA EPIA N-Series Nano-ITX Mainboard
Sphere It. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Sinclair Says:
February 2nd, 2007 at 6:57 pm
Has anyone from Line6 Variax expressed interest in this project? It would be nice to see a linux version of sound libraries for Line6 POD devices also.
Michael Trowbridge Says:
February 2nd, 2007 at 7:25 pm
First, this idea is absolutely brilliant, and I can draw signal flow diagrams to show any audiophile why recordings captured inside the guitar itself can be better quality than even the best studio recordings.
Second, I have worked with ash, and with the amount of cavities you will have to cut to build in the hardware, I think it’s an excellent choice because of its strength. I would recommend avoiding any displays on the neck, because assymetric cavities in the neck (even if they’re just screw-holes) are enough for string tension to cause serious long-term neck warp.
I also have worked with an Epia motherboard, and I think the hardware would be a good match, but poses a couple of challenges. The first one is vibration dampening. Ash transmits these vibrations extremely well, so you’d need some kind of rubber shock dampening system between the motherboard and the ash body. Without one, the vibrations could potentially rattle connectors loose enough for abnormal hard drive or memory performance. Long term vibration stress could rattle the processor heatsink loose if it’s not mechanically secured to the motherboard, so that’s something to look for when you’re motherboard shopping.
Also, pickups might create huge EM fields that could cause inducted current in odd places on the motherboard. If this happens, symptoms would be similar to bad RAM (i.e. every aspect about the system seems to fail, but inconsistently and irregularly).
To defeat this, line the inside of the motherboard cavity with a conductor (foil or wire mesh), and connect one end directly to chassis ground, and the other end as to a dedicated ground cable. The cleanest (electrically) way to do this is to give your guitar two 1/4″ outputs – one carrying the actual audio signal, and the other connecting to an electrical ground. Any power supply grounds or motherboard grounds should be connected to the chassis as far away from the external ground wire and connected so that any current that runs from the power supply or motherboard to ground has to pass through the shielding conductor in the motherboard cavity to get to the ground wire.
I would recommend avoiding Wi-Fi or disabling all current to any built in wifi adapters because the RF transmitters in the wifi adapters are so close to the pickups and audio lines. It would be very easy for pickups and wiring in the guitar to act as antenna, and pick up either the wifi carrier or harmonics of it. That would cause additional system noise that could sound like dirty pickups.
I would recommend not using any amplifiers between the pickup and the A/D converters/samplers, because this system has an extremely high potential for inducted electrical noise. Once that noise hits an amp, it’s game over for recording quality. It’s much safer to do any required amplification after sampling as part of DSP.
Your idea is brilliant, and it has a huge amount of potential. Stay with it!
tommyk Says:
February 3rd, 2007 at 1:26 am
You GOTTA change the background of this website- can’t read the font at all.
ps
put LEDs under the frets that will light up notes being played when you play a song by van halen, for example. shows you what eddie was playing.Kinda like working backwards.
Neil Richardson Says:
February 3rd, 2007 at 1:39 am
Sounds a great idea! Definatly has potential!
I mean building effects into a guitar has already been done, but software effects, the potential is seriously limitless. Especially for the recording factor.
Good luck with the project!
Michael Trowbridge Says:
February 3rd, 2007 at 8:00 am
Correction to my last post: shielding for motherboard compartment should be wired in parallel, not serial. I apologize.
salvarsan Says:
February 3rd, 2007 at 8:02 pm
The EPIA-N is just under 2-1/4″ high with heat sink and needs a 60Watt power supply. This means a thick guitar and chunky battery.
Rather than put a headless laptop in a guitar, I suggest using somewhat lower powered embedded CPU boards (~600MHz, 12-15 Watts). A google on “single board computer” brings up some possibles from http://www.embeddedsys.com and http://www.evalue-tech.com , but there are dozens more.
(edited) url’s
shavenwarthog Says:
February 4th, 2007 at 2:14 am
perhaps a non-Linux microcontroller would do the trick:
This ARM board is only $50! — http://www.luminarymicro.com/products/ekk-lm3s811_evaluation_kit.html
It’s 50 MHz, 64KB flash, USB, display(!), and a 4-channel ADC on a 3.65†x 1.40†x 0.30″ board.
Matthias Martin Says:
February 5th, 2007 at 9:09 am
The EPIA-board is pretty close to an no-no because of its power and space reqirements. Please look into the ARM-9-based SOCs from APLIO – those feature onboard DSPs and are actually designed to handle much of the audio stuff in hardware, and they’re running linux too. But I do realize that it would require much more skills though. It would most likely also involve doing a custom PCB with the required audio cirquitry on it (Humbucker2DAC or so).
Ok, wish you much success with your cool project.
salvarsan Says:
February 5th, 2007 at 2:31 pm
It might be easier to get a linux guitar application suite working on a desktop workstation, then see how much will comfortably move to a 100-500 MIPS embedded platform.
Keep in mind that Gibson has been paying lip service to a digital stream guitar for 10 or so years with their MAGIC system. It is, of course, stupidly expensive but has backing from a few big ones (Intel, HP, Xilinx) and even has Cakewalk embedded within.
The important thing is to create a lean alternative for people with a little more sense than money.
Michael Trowbridge Says:
February 5th, 2007 at 5:14 pm
Another idea for small form factor computers running linux:
http://www.gumstix.com/products.html
Dave Carlson Says:
February 5th, 2007 at 6:20 pm
I would recommend an ARM 9 processor based board, you can acheive a nice small footprint and very low power consumption, well under the 5 watt mark. There’s already alot of Linux kernel development done for this core as well as development tools.
Raver Says:
February 12th, 2007 at 8:55 pm
Take a look at this little computer. It use PPC.
http://www.efika.de/index_en.html
Leon Says:
February 18th, 2007 at 1:57 am
> Also, pickups might create huge EM fields that could cause inducted current in odd places on the motherboard. If this happens, symptoms would be similar to bad RAM (i.e. every aspect about the system seems to fail, but inconsistently and irregularly).
Why won’t you use piezo pickups? They lack all those disadvantages.
Leave a Comment