audio over USB v. optical - Page 19

Dennis5587 · 02-20-2007, 03:03 PM

...and it has to be asio for all 8 to show up

Felix509 · 02-20-2007, 04:11 PM

In Way Over My Head

Well, I was thinking 8 channels of input as in (4) seperate USB-I2s Chips.. I realize it may not work that way??

I suppose it will be much easier to have multiple channels go out over one USB2 cable to a unit windows recognizes as One devive..

This BeBoB format is another thing i have bookmarked::

BeBoB

I will now let someone that knows more about this stuff do the talking.. I have soaked up some Info, strictly from internet research/forums/etc, but definitely do not have a complete grasp of all of this...

HIPPOhifi · 02-20-2007, 05:56 PM

Quote: Originally Posted by durwood

Although this is an option, would any type of computer software recognize this as multiple outputs, or is that just going to add on outputs like a y-splitter would do?

No, it would be recognised under the OS as having single outputs. It would be like a y splitter either at the digital stage or the analogue stage. Both would have there advantages and disadvantages.

Quote: Originally Posted by durwood

We were thinking it would need to have additional power from 12V to help power the opamps. USB bus power would not be enough to drive to opamps.

The USB standard allows 5V @ 500mA or 2.5watts. The Bloat unit uses near this spec with enough room to add another opamp. This in total makes three dual channel opamps, considering the Bloats uses two for the one output to driver higher impedance headphones.

Quote: Originally Posted by Felix509

In regards to the Multiple outputs, it is equally important to have the same amount of inputs so as to do the processing in the computer..

Correct me if wrong, but we would need a seperate USB to I2S receiver chip for each stereo pair.(highs, mids, midbass, sub) Would (4) of the PCM2707s show up in windows as 4 seperate devices?? Maybe you could test that out???

Yes you are correct. I have tested multiple Bloats and they are recognised as separate audio devices, but I have not tried to output audio to more than one at once. I assume I would require some nifty mixing software, let me see what I can find.

Quote: Originally Posted by Felix509

I am sure their would then be clock issues with the DACs being totally independent from each other. From talking with Doede from the DDDAC1543 fame, I was understanding we could sync all of the receiver chips and all 4 DAC chips to the same High precision clock????? Does that sound right??

The PCM270* series, uses a 12-Mhz clock for both the USB and audio functions. When having multiple Bloats attached, once the USB is "in sync", I would struggle to see where they could be out of sync. That being said, in your application running highs/mids/bass, if the clock were out of sync it would be completely inaudible. It is possible to sync from the same clock but the data sheet for the PCM270* doesn't recommend using the clocking function in this fashion.

An aside question, does anyone have specs on there amplifiers for input impedance. If you do can you post a link to the specifications.

RedGTiVR6 · 02-20-2007, 06:00 PM

75 ohms is normal for most RCAs

Felix509 · 02-20-2007, 06:35 PM

A few examples for the input impedance for various amps, 10-12 Kohm looks average.

Diamond D7 amps have 10Kohm input

McIntosh MCC404 is 12000 ohms

Soundstream Reference is 12Kohm

ALpine PDX is 10k ohms

durwood · 02-20-2007, 06:37 PM

- But actually though the input impedance to an ampilifer is usually between 10k-15K ohms.

HIPPOhifi · 02-20-2007, 06:41 PM

So, I was thinking of HIPPOhifi's next project and how it can be used to assist from the multichannel perspective. Some background information, we have been dying to test a number of rarer op amps, in different configurations powered from different sources on the same board, thus derives a UNIVERSAL OP AMP BOARD. Goals:

Mandatory
a. Maintain high SQ whilst providing flexibility
b. Multi-powered (battery, DC, USB)
c. Interchangeable op amps
d. Class A option
e. Optional ground channel buffering
f. Optional use of buffering components on all channels
e. Size, fit on a board 80mm * 52mm (used in Hammond 1455C801 enclosure)

Desireable
a. Compatibility for both single, dual and quad channel op amps

Provided to customers as:

a. PCB with instructions
b. Kitset with components for basic, nirvana and nirvana plus levels.
c. Completely constructed versions of (b).

Back to our problem, we could "y-split" the output of you source, ie The Bloat or otherwise, and run it into a number of these Universal Op Amp Boards which we can specifically customize for car amplifier application. We are not talking mega $$$s either. Thoughts?

durwood · 02-20-2007, 06:45 PM

Quote:

Back to our problem, we could "y-split" the output of you source, ie The Bloat or otherwise, and run it into a number of these Universal Op Amp Boards which we can specifically customize for car amplifier application. We are not talking mega $$$s either. Thoughts?

Would it show up as multiple channels when using ASIO applications such as Console?

www.console.jp

If it doesn't then it still won't do what we need it to do.

HIPPOhifi · 02-20-2007, 06:45 PM

Thanks for the impedance numbers, I would assume the higher numbers (10k) are with the gain control at minimum and the lower numbers (75) with the gain control at maximum...

HIPPOhifi · 02-20-2007, 06:48 PM

Quote: Originally Posted by durwood

Would it show up as multiple channels when using ASIO applications such as Console?

www.console.jp

If it doesn't then it still won't do what we need it to do.

Ahh, maybe I was misunderstanding what you are trying to achieve. So you basically want to split the frequencies in the software and send it of the PC in separate physical channels?

Dennis5587 · 02-20-2007, 06:48 PM

Quote: Originally Posted by durwood

Would it show up as multiple channels when using ASIO applications such as Console?

www.console.jp

If it doesn't then it still won't do what we need it to do.

I have a feeling this is going to be one of the biggest problems yet...

durwood · 02-20-2007, 06:50 PM

I like your thoughts on design so far.

AS far as dimensions go, a typical car audio processor "brainbox" is about
260mm x 40mm x 160mm (Alpine PXA-H701). So that shoudl give you some more working room.

But smaller is better too

durwood · 02-20-2007, 06:51 PM

Quote: Originally Posted by HIPPOhifi

Ahh, maybe I was misunderstanding what you are trying to achieve. So you basically want to split the frequencies in the software and send it of the PC in separate physical channels?

Yes. Here comes the bad news...right?

turbocad6 · 02-20-2007, 07:01 PM

Quote: Originally Posted by HIPPOhifi

So, I was thinking of HIPPOhifi's next project and how it can be used to assist from the multichannel perspective. Some background information, we have been dying to test a number of rarer op amps, in different configurations powered from different sources on the same board, thus derives a UNIVERSAL OP AMP BOARD. Goals:

Mandatory
a. Maintain high SQ whilst providing flexibility
b. Multi-powered (battery, DC, USB)
c. Interchangeable op amps
d. Class A option
e. Optional ground channel buffering
f. Optional use of buffering components on all channels
e. Size, fit on a board 80mm * 52mm (used in Hammond 1455C801 enclosure)

Desireable
a. Compatibility for both single, dual and quad channel op amps

Provided to customers as:

a. PCB with instructions
b. Kitset with components for basic, nirvana and nirvana plus levels.
c. Completely constructed versions of (b).

Back to our problem, we could "y-split" the output of you source, ie The Bloat or otherwise, and run it in to a number of these Universal Op Amp Boards which we can specifically customize for car amplifier application. We are not talking mega $$$s either. Thoughts?

to me it looks like the problem with this design is this will not give the ability to x-over each driver or delay each driver individually... this is why the need for at least 8 separate channels directly controlable individually is manditory... 10 would be ideal & 12 would be the ultimate for 3 way front, 2 way rear & stereo sub...

I think at least 8 is a definite requirement... y splitting isn't going to do much more than could already be done now with a pre amp I think...

durwood · 02-20-2007, 07:03 PM

Felix-

I missed your post about BeBoB. THat might be what we need, however I still see a DSP chip involved (trying to eliminate that). Maybe there is no way around that. But maybe that is what some of these pro-USB solutions are using. Look at what I found at quick glance:

Quote:

The BridgeCo USB2.0 Audio Driver software enables
high-performance, using a USB2.0 High Speed I/O link.
The driver provides an ASIO and a WDM interface, and
runs on Windows® XP SP1/SP2 Home and Professional
versions.
Round-trip latency is around 10ms.

This might be what we have to do instead of USB to I2S. Any other thoughts anyone?