Guillermo;
The best solution would be to know exactly the source of the noise and
eliminate it. Do they show up in other samples?
Anyway, have you tried adjusting Receiver Gain manually to a lower level
than optimum, maybe 1/4 to 1/8 of clipping and then increasing the
experiment time accordingly to get the S/N desired for peaks of
interest? Because of non-linearity of most amps and other
eccentricities it might work better.
Another solution would be still another post-processing solution using a
mathematical subtraction instead. Regular peaks should not be too tough
to simulate mathematically if they were close to ideal but you said they
were ugly. If the are close to perfect little peaks exactly spaced, it
will work using a fourier series to the nth place. Keep increasing n to
see what it does.
Something I have not yet seen or read about in any production NMR lab is
complex electronic (or analog) filtering which is available in other
research settings (Physics, Astronomy). Pricey! Transients would need
to be run through extra-fast computerized processors (or custom analog
circuits) removing previously ID'd signatures before they are added to
the FID's. One day;)
///--)\\\
R. Harker
UF, NMR Labs, Gainesville Fl 32611-7200
rharker_at_chem.ufl.edu
352-392-4650
-----Original Message-----
> From: Zhou, Zhe (ZZ) [mailto:ZZhou_at_dow.com]
> Sent: Monday, May 10, 2004 9:16 AM
> To: 'Guillermo Moyna'; ammrl_at_chemnmr.colorado.edu
> Subject: RE: Eliminating spikes caused by VERY strong signals...
Try to use sw less than 25,000Hz and use DQD.
Adjust RG manually, may be it should be around 1K, do not use RGA.
Good luck.
ZZ
________________________________
Zhe Zhou, Ph.D.
Research Specialist (NMR)
Molecular Structure Group
Analytical Sciences, Dow Chemical Company
2301 N. Brazosport Blvd., B-1219
Freeport, TX 77541
Tel: 979-238-1387
Fax: 979-238-0752
Email: zzhou_at_dow.com
-----Original Message-----
> From: Guillermo Moyna [mailto:g.moyna_at_usip.edu]
> Sent: Friday, May 07, 2004 8:18 AM
> To: ammrl_at_chemnmr.colorado.edu
> Subject: Eliminating spikes caused by VERY strong signals...
Hey gang,
I am wondering if someone has encountered this problem before and got a
solution for it. We are taking 13C spectra of carbohydrates disolved in
non-deuterated solvents. The solvent peaks are not in the range of the
sugar peaks, but since the ratio of sugar peaks vs. solvent peaks is
1:100, the artifacts (spikes) originating from the solvent peaks start
showing at regular intervals in the sugar region after extended
accumulation. They are small enough not to be a major problem, and we
can identify them because they are always there no matter which sugar
you are analyzing, but they don't look all that pretty.
Is there any post-processing trick to remove these things? The data is
being collected on a Bruker AVANCE 400, but due to the chemical shift
range of the solvent we cannot use DQD. My solution so far is to record
a spectrum of the neat solvent with the same number of scans and
substract this from the spectra of the sugars. This works, but I'd like
to avoid doing it if there's an alternative solution.
TIA,
Guillermo
+==================-------------- --- -- - - - -
Guillermo Moyna, PhD
Assistant Professor of Chemistry
Department of Chemistry & Biochemistry
University of the Sciences in Philadelphia
600 South 43rd Street
Philadelphia, PA 19104-4495
"The only existing things are atoms and empty space.
All else is mere opinion" - Democritus, 370 B.C.
Office: Griffith Hall 360
Phone: (215) 596-8526
Fax: (215) 596-8543
e-mail: g.moyna_at_usip.edu
WWW: http://tonga.usip.edu/gmoyna/index.html
http://www.usip.edu/chemistry/faculty/moyna.asp
- - - - -- --- -----------=================+
Received on Tue May 11 2004 - 18:59:22 MST