I don't think the IQ imbalance is a problem at all in  DDC designs 
... both the I and Q channels  are subject to the same rounding 
errors so in general you get some level of noise added to the signal, 
but not a consistent imbalance as you do in the QSD case where you've 
got two analog paths.   You can also mitigate the rounding errors in 
the math and
cordic by using enough bits in the DSP chain in an 
FPGA.   OpenHPSDR.org's Mercury board just uses the wide enough 
cordic and math approach -  have never found a need to do anything 
beyond that to keep good IQ balance.

Cheers,

Bill (kd5tfd)


At 07:24 PM 5/12/2010, you wrote:
>Alberto
>
>Thanks for a great reply. I'm familiar with the analysis that you 
>detailed.  I should have mentioned that I'm interested in SSB and PSK.
>
>You mentioned that 86 db might not be enough for HF.  What drives 
>the quest for dynamic range?  Is it the ability to dig out a weak 
>station next to a powerful one?
>
>When I last operated, I used a not so great commercial analog 
>transceiver and some "home built" NE602 based direct conversion 
>stuff (shudder).
>
>Also, does the DDC approach suffer from I/Q imbalance problems like 
>the QSD? There must be amplitude and phase errors resulting from the 
>finite precision used for sin/cos lookup tables or CORDIC.  My guess 
>is that these problems are much more trackable in the digital domain 
>despite attempts at restoring I/Q balance in software.
>
>Thanks for the reply...
>
>Wayne


Reply via email to