The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> baseband IQ
https://designers-guide.org/forum/YaBB.pl?num=1182735301

Message started by aaron_do on Jun 24^th, 2007, 6:35pm

Title: baseband IQ
Post by aaron_do on Jun 24^th, 2007, 6:35pm

Hi all,

I noticed that most low-IF transceivers provide both I and Q to the baseband section. Is this necessary if image rejection is already done? Assume i'm using a passive poly phase filter and the image rejection is sufficient for my purposes.

thanks,
Aaron

Title: Re: baseband IQ
Post by mg777 on Jun 25^th, 2007, 7:31am

My gut feeling is the signal dimensionality of each of the I/Q is less than that of the composite.

M.G.Rajan

Title: Re: baseband IQ
Post by carlgrace on Jun 27^th, 2007, 5:33am

Both I and Q are required in a receiver because they employ quadrature modulation. The I and Q are uncorrelated and indepedent data streams until they are demodulated in the digital baseband. The polyphase filter improves the quadrature accuracy but does not do anything to demodulate the data.

Carl

Title: Re: baseband IQ
Post by didac on Jun 29^th, 2007, 1:10pm

Hi,
My two cents on this thread:
1) You can perform the I/Q demodulation directly in the digital domain(it's not needed to provide the DSP with and I/Q path, the transformation it's just a multiplication with a sin() and a cos() easily done with C or whetever use the digital processor).
2)If you have a quadrature modulation and want to the demodulation directly in analog domain(less power consumption) you must provide the I/Q path to implement the analog matched filter( a simple correlator), I've seen this in extremely low power applications.
3)Although that it can seem a bit strange it can be useful to provide a I/Q path in the analog domain while your signal it's something like a FSK, if you don't recover the phase of the incoming RF signal your LO can be orthogonal sometimes to the carrier thus losing all the energy(due to effects of the channel,like a Rayleigh distribution common in f<10GHz or so), if you use a couple of Mixers with LO in quadrature you will have in your worst case energy in the Q path and usually you will have energy at both branches(I've seen this in zero-IF receivers and I think it's usual also in low-IF receivers).
Hope it helps,