Choice of a hardware platform for the development of the SSLMixed Computer

[Artur D'Assumpção]

Hi guys,

I've been investigating for the last couple of days what could be a good platform for us to base our SSLMixed computer development. I think that if we decide to start with an already designed and well supported hardware base platform it would ease up greatly both the hardware and software design, giving us a couple of significant advantages:

          - This would give our hardware designers more time too keep focus on what's really important which is the reverse engineering tasks and the logic design for the I/O interface between the computer and the console;
          - Take advantage of a well established and test proofed platform;
          - Usually these platforms already include other multi-purpose components (Ethernet, USB, HDMI, several I/O, etc.);
          - There are many options out there leaving us the ability to choose the best option to fit our purpose;
          - Most of these platforms are based on pretty standard electronics;
          - Since these platforms are designed for development purposes usually they provide lots of documentation and implementation examples. Also provide a software base, drivers, SDKs and IDEs that speed up greatly the software development;
          - There's a big community backing it up and providing support;
and much more...

Of course we have to think about the future when opting for this alternative, and these development platforms they will eventually become obsolete and give place to others. This is why it's important to choose a platform that uses standard components mainly the CPU (architecture) to maintain our source code intact. This enable us, if we need to upgrade our base, to quickly port to another platform with very low effort, mainly just re-configuring I/O configuration to support our hardware design and not much more. Also another possibility that we can pursuit on a later future is to design our own platform to assure an even longer platform lifespan. This platform would then be based on the same components as these development platforms use enabling once more an easy migration between platforms. In my opinion I believe this is a great added value for this project since we can focus our effort and time into the core of this platform which is the I/O logic and Software design. Trying to reinvent the wheel would be a big setback for us since our time and resources are limited.

I've been digging in the search of several options and I've found one that's called the BeagleBone Black which I think it's very powerful and versatile, give us almost everything we need (to validate this). Please check the following link about it: http://beagleboard.org/products/beaglebone%20black

Don't let yourself fool for it's appearance, it's indeed a very powerful full blown computer on a very small footprint that surpasses the old SSL Computer capacity in 10000000000000000x times!

Bellow you can find a quick reference on the board configuration and features:



Also the following link will give you a more complete reference to the capacity of this little beast: http://circuitco.com/support/index.php?title=BeagleBoneBlack

I'll highlight a couple of things that caught my eye when evaluation this solution:

• CPU Architecture: AM335x 1GHz ARM® Cortex-A8 (we can run Linux OS with full support)
• 512 MB DDR3
• 2GB 8-bit eMMC on-board flash storage
• USB Port
• Ethernet port 10/100
• microSD slot
• miniHDMI (video+audio, maybe we can hook a TV screen to this if we like)
• Big ammount of GP I/O to connect our card (65 available digital IO pins and 7 analog)
• very small form factor
• very cheap solution (45 dollars!!!!)
• big community support (such as Raspberry Pi)
• Provided sofware (several Linux distributions specially prepared for beaglebone, very good documentation and examples, SDKs provided in a few programming languages)

Here also a couple of videos about the beaglebone black if you have the time to watch:

* BeagleBone Black features: https://www.youtube.com/watch?v=ciX08ysl6LE
* Intro do the Beaglebone (the previous version before black): https://www.youtube.com/watch?v=z6b4zlh0IrE
* I/O on the BeagleBone Black: https://www.youtube.com/watch?v=wui_wU1AeQc

General information about the Beaglebone boards here: http://beagleboard.org/

UPDATE:
- If you want to know even more details here's the link for the reference manual: http://www.mouser.com/catalog/specsheets/BBB_SRM.pdf

Enjoy and discuss!

Cheers,

Artur

[pascal.verdet]

Hi Guys,
I'm reading the reference manual.
This seems nice ...
There also a touchscreen interface, may be interesting

Cheers,
Pascal

[Artur D'Assumpção]

Hi Pascal,

Yes there are already several addons  (which are called capes because you can mount them on top of each other) made and commercialized by the community that you can fit into the beagle. Anyway you have to be careful because when you fit these "capes" you'll be using GPIO ports that might be important for our main purpose. Like you said before we first need to start understanding what we'll really need on the side of the SSL and how many ports will our logic card need on the beagle.

Do you see why I was telling this is a solution that will cut our development times greatly without any compromise, on the contrary?

Also the advantage is that Beagle is an open-hardware solution (like open software but on the hardware side). There are many projects like Beagle out there, for instance the Raspberry Pi or the Arduino. This means that we have total access to board schematics and how it's made, so if we wish to make our own board based on one of these designs it will be much easier since 99% of the work is already done!

Cheers,

Artur

[djgout]

Hey Guys,

This project is light years beyond me and any type of design I'd be comfortable handling but besides the SSL specific things (ie total recall/ultimation) how is this box really going to be that different from a DAW? If we're just talking recording the analog voltage levels from the fader onto a disk and replaying them out to the vca later, couldn't we just develop a modification to some cheap audio interfaces to use them for DC instead of AC and shove them into really any DAW? It would be easy enough to update mixes, trim faders, draw in mutes and the like from any software that already exists.  If ya need to sync to smpte, just plug the output from your analog into the daw and go. Maybe I'm over simplifying things though in my mind. I know there's more to it than that, but that's how I've always explained the concept of mix automation to clients in the past.

[Artur D'Assumpção]

Hi,

Well actually you'll be just doing what the computer does nowadays, which is handling projects and recording automation. The idea of DAW integration somehow (although it seems to be a pain in the ass, it should be easier to link both systems) is something that can be evaluated on a later stage, and its supported by the current design. Besides that it's important to state that one of the requisites is to enable it to work solo, independent of any other computers.

Cheers,

Artur