A (final?) note on this subject, as the replies are still pouring in. :)
Clemens Anklin (of course!) wins the prize for the best description of what
is going on in my machine, an Avance III with DQD. (Others can tell me if
that is exclusively what is used these days or that we are just lucky to
have that.)
Here's my take so far on this. Probably don't have this quite right
still....
The 20 MHz-wide signal is really offset well to the side of the spectral
range, which in principle would then -- like the old, old days, not require
a complex FT. Today's imaginary part of the FID arises from the
mathematical treatment of this original (massively oversampled) signal,
shifting it by deconvolution. Downsampling improves signal-to-noise and
allows for a smaller, more manageable dataset represented by cosine and
sine terms. This is the fid that is placed in the dataset and transformed
into the real and imaginary components of the spectrum, allowing them to be
phased.
Maybe this discussion will spur someone in this group (hint, hint) to
create a blog post highlighting the evolution of detection technology that
has gotten us to this amazing state of the art. I've always thought that
NMR is (at least in chemistry) one of the absolutely most amazing
combinations of mathematics, engineering, and science. The story should be
told! ;)
Bob
Received on Sat Feb 18 2023 - 06:48:36 MST