Re: [Jack-Devel] Ethernet-based audio interface using (net)jack idea

On Sat, Aug 6, 2011 at 2:15 PM, Dan Swain <[hidden]> wrote:

>
>
> On Sat, Aug 6, 2011 at 10:22 AM, Dan Swain <[hidden]> wrote:
>
>> On Sat, Aug 6, 2011 at 3:17 AM, Jeroen Van den Keybus <
>> [hidden]> wrote:
>>
>>> In my original post, I proposed slave hardware that would synchronize its
>>> local clock to sync packets from a regular Jack server (digital PLL),
>>> advertise a fixed amount of (floating point) and MIDI channels to the
>>> server, get packet/period/... data from the server and start sending
>>> receiving data over UDP continuously. The software needs to emulate a
>>> fixed-configuration NetJack2 client, which can eventually be done without an
>>> IP stack, since the initial communication is done using broadcasts.
>>> Therefore: small software footprint ==> no EEPROM or DRAM on the board
>>> ==> low pin-count FPGA ==> low cost. Since we're trying to digitize
>>> precision analog signals, reduction of power (no high-current
>>> switching voltage regulators) and omission of memory (no noisy digital
>>> memory lines) is also a good thing for circuit design.
>>>
>>
>>>
>> ...
>
>>
>> Jeroen, are the components you listed in your first response still
>> applicable for this method?
>>
>
> I've just been researching the parts that you've listed, and I'll make sure
> to download and read the data sheet for the EP3C10.
>
> As for the Ethernet, I've found two controllers that may be worth looking
> at - either the Microchip ENC424J600 (10/100 Base-T), or the Marvell Yukon
> Ultra II (8E8057, which is 1000 Base-T). The microchip design sounds quite
> promising, as I came across its 10Base-T 'cousin' when looking at the Open
> Open Sound Controller project.
>
> I may be thinking about this incorrectly, but my approach to have the ADC
> talk to JACK on the FPGA was to have some kind of ALSA 'driver' for that
> chip running in the FPGA, passing the data through JACK to the Ethernet
> controller. Alas, the CS5368 chip isn't supported in ALSA, but we could use
> the CMI8788 chip, which is almost exactly the same.
>
> Then again, am I thinking about this incorrectly? If we could pass the
> signals directly to the JACK system (literally give the data it wants), that
> would circumvent another software layer in the FPGA...
>

I apologise, I misread the datasheet for the CS5368, so I change my
viewpoint on that IC, as it would be perfect for this project (which is why
you recommended it!). Actually, one of the circuits I came across for the
ADC section of the device uses that chip:
http://www.beis.de/Elektronik/ADDA24QS/ADDA24QS.html <-- The designer is
aiming for quality rather than cost savings, so again I'm sorry for the
false analysis!

I was also looking more at Altera's tutorial videos on YouTube, and if the
Cyclone III chip has enough space left, maybe we could create a device on
the chip itself, so that the digital audio signals can be sent to the JACK
server on the chip correctly. From what I've seen of this chip's
capabilities, I'm not sure that we'd be stressing it too much (for a
relative example, one of the video capture cards we use at my workplace has
the first Cyclone chip, which can handle 16 channels of 704x480, 16-bit
video at 7fps per channel)!

1312679562.3905_0.ltw:2,a <CABhkv4MXY0ZZP066gptGFCUMEdOOiZbdBjs+pH=htmvVk0nq=Q at mail dot gmail dot com>