What is I squared S?

What is I squared S?

Post by Mike » Sat, 08 Mar 1997 04:00:00

I'm looking for a higher quality codec to interface with my ADSP-2181 on
my EZ-KIT Lite.  I see that a lot of codecs use an I2S (I squared S)
serial communications format.  Will the ADSP interface with that format
easily?  Is it compatible with the 2181's multichannel mode?




What is I squared S?

Post by Chris Ho » Sun, 09 Mar 1997 04:00:00

I sqares S  comes from Philips. Perhaps you can get information on some
Philips homepage?! Crystal devices have the same data format, have a look
on their homepage. I can't help you with adresses, I only have data books.



What is I squared S?

Post by String » Thu, 13 Mar 1997 04:00:00

You can find this informatio in almost any data book related to audio

It is a serial interface for audio converters.  It consists of the
following signals (ony 4 are necessary for either direction):

Master Clock - Typically  the sampling rate x oversmapling.
  e.g. 48kHz sampling x 256 times oversampling will be 12.288MHz
Data Clock - one cycle per data bit in or out.  Alignment and phse
  are important here, see a data book.
L/R signal - identifies if right or left channel information
  is currently being sent/recieved.
Data out - data sent to DAC
Data in - data in from ADC.

There are dozens of variations in use by various people.  The newer
codec's are actually giving you quite a number of options, such as
positive clock or negative clock, data starts with L/R or is delayed by
one cycle, 16, 24, or 32 bit lengths, etc.....


1. AM demodulation using I and Q

I started a thread re: AM demodulation using I and Q a few weeks ago
and I am still having a problem.  According to Frerking, AM should be
A=sqrt(i^2+q^2).  This works if I tune off the AM zero beat frequency
by 1/2 the low pass filter bandwidth.  I don't understand this

A little more detail:

I am using a pair of mixers fed with a sin/cos local osc. to generate
I/Q at baseband.  SSB demod works fine for upper and lower sidebands
with excellent opposite sideband rejection.  I can zero beat the AM
carrier and receive upper, lower, or both sidebands to stereo output.
I am using a DDS LO and I am able to accurately zero beat the AM
carrier to less than 1hz. When I do this while listening to both
sidebands or I/Q in stereo, I can hear the sidebands come into phase
as I tune the radio.

Once I have the signal zero beat, I flip it into the AM algorithm and
I hear a great deal of distortion.  Assuming I use a low pass filter
(actually a bandpass) from 30-4000hz, prior to the sqrt(i^2+Q^2) math,
if I move off zero beat by 2000hz, the AM sounds great.  I can move
either higher or lower by 2000 with the same results. It almost seems
like the filter is acting like a mixer??

I'm missing something here.  When I modeled this receiver using Matlab
Simulink, it works fine at zero beat.  If I get this simple
demodulator working I am going to move onto synch AM.  I actually like
listening to the independent sidebands in stereo since it takes care
of some of the fade.


2. A problem with ali in MH 6.8

3. Am I wrong here?

4. database server reinstall, duplicate accounts in AD

5. AM demodulation help

6. Additional RS232 ports ?

7. TMS320 : I am sure that all of u can help me

8. Locale ARexx Host?

9. I am new to adaptive environment ... help need !!!!

10. McAulay Quatieri, what am I doing wrong???

11. Average power of AM modulated signal

12. I am hopeless ! ! ! !

13. i am a freshman,Good day,everyone!