[linux-audio-user] Re: [linux-audio-dev] Re: [Alsa-devel] Firewire Audio Card Support

Mark Knecht markknecht at gmail.com
Thu Nov 18 19:31:03 EST 2004


On Fri, 19 Nov 2004 00:52:48 +0100, Florian Schmidt <mista.tapas at gmx.net> wrote:
> On Thu, 18 Nov 2004 14:54:27 -0800
> Mark Knecht <markknecht at gmail.com> wrote:
> 
> > I've had this sort of project on my mind for a year or two. (Since my
> > first run in with the HDSP 9652 under Linux in early 2003.) I am,
> > unlike I think many people here, a hardware designer by trade. Mostly
> > chips, but my engineers had done boards for our chips.
> >
> > Coincedentally I was laid off yesterday and am sitting at home
> > thinking about what to do with myself during my current delimma...
> 
> Hi,
> 
> maybe you can tell us then how much the components for such a simple pcm
> only soundcard would cost? What's nessecary? And i mean really simple. Just
> a single full duplex stereo pcm device. No mixer (who needs a mixer? Real
> men have their mixer sitting in a rack :))
> 
> - 2 AD's and 2 DA's
> - a dsp (is it really nessecary?)
> - some memory for the buffers
> - a pci board
> - pci logic (raising irq's, doing the transfers)
> - some "glue"
> 
> i may be naive though as i really have no idea about the hw side of things.
> Especially it gets tricky when allowing all kinds of different sample
> rates/buffer sizes, etc. And of course i suppose the price is heavily
> dependent on the quality of the components. But what would be lower and
> upper bounds for the components alone? And what kind of work is nessecary
> (how many manhours)
> 
> - to design the thing
> 
> - to build the thing
> 
> ?
> 
> Florian Schmidt
> 

Florian,
   Sure, I can take a sort of wild swag at some numbers. For that sort
of card I don't think you're going to like them though. They're likely
to be far, far higher than what you can get in the open market from
Creative Labs or M-Audio.

   First, let's look at the board itself. Technically speaking it will
need to be a 4-layer board for sure. Analog stuff requires good ground
and power planes. Here in Silicon Valley I used a company called
Sierra Proto Express, but there are hundreds if not thousands of
companies like this around the world.

https://www.2justforyou.com/NASApp/sierraproject/jsp/tabs_welcome_home_.jsp

They build the basic card in a 'no touch' setup for about $50. However
when you go for more production the price moves down. The issue is, in
a group like this, how many people want to pay to get this done at the
same time, etc., and do we do it all here or do we give out the Gerber
files and just let people build their own. (I think that's not
practical.)

Stereo D/A & A/D are anywhere from $5 each to maybe $50 each,
depending on what sort of quality we go for. Are these 44.1 only?
44.1/48? 96K? 192K?

The PCI interface is best done, in my opinion, in an FPGA ala the way
RME does their cards. That can be expensive, from the $5-$10 range up
to $1000 that I used on a recent design at my last company. Other
onboard components won't cost much. $5-$10 probably.

Adding it all up raw material cost for this sort of board is anywhere
from $75 to $230, with an opportunity to go much more expensive
depending on what the board wants to do.

However, then you have to assemble it. This can run (at a small
assembly house) on the order of $100-$200 for each board so at this
point we're talking about anywhere from $175-$400.

And this is for a board that might replace an AP2496. Not much
technology, but we do have control over the design and the software
and it's a completely open design, which is cool.

Pretty bad, 'eh?

Circuit design time would likely be anywhere from 2-4 weeks, what with
discussions and timezones, etc., and assuming that there are some
predesigned PCI interfaces we can get somewhere. Longer if we have to
do that from scratch. (Ever looked at the shape of the Linux Verilog
simulator? And we need FPGA programming tools, etc.)

Usually these things don't work perfectly the first time, so assume a
spin or two on the board and assembly as we debug it. The money starts
adding up...

Anyway, before I depress you too much, I'll stop going there.

>From my POV a more interesting idea would be to do an external sound
device, probably 1394 based since that will work for more people.
Please remember that a PCI card is almost useless for laptop users
unless we're trying to put this into a cardbus formal also. That adds
money.

If it was 1394 based then you can put a 1394 adapter in your PC for
$20 and then everyone uses the same audio unit. We control how it
works, so we can follow specs or do it in our own standard. You get
the advantage of probably more channels and better SNR, but you do
have to package the unit in a box or some type to be of general use.

Anyway, those are some ideas for you to chew on. Hope I haven't poked
a balloon with a pin here...

- Mark



More information about the Linux-audio-user mailing list