AMMRL- wrong chemical shift on AC300

From: Quincy Teng <teng_at_sunchem.chem.uga.edu>
Date: Thu, 17 Feb 2000 10:34:59 -0500

Thanks to everyone who replied my message. The problem is caused due to
bad ADC board. Does anyone have an aspect3000 computer and no longer need
it? We will pay shipping and $$$ for it.
I started reading ammrl emails early 1994, don't know much about the
history of ammrl. Can someone post a story about it.
Thank you very much.
Quincy

**********************
problem: wrong chemical shift value on Bruker AC300
cause: bad ADC board
confirmed by: 1J(CH) of 13C satellites was also scaled by the same factor.
(I also swapped the board with AC250)

Other suggestions:

What is the sweep width? Are you folding a peak somehow?
---
Perhaps there is a problem with the dwell clock generation in the
computer.  You might try scans with different sweep widths to see if any of
them give the correct result.
---
I have seen a similar problem on a WM-250 where the sampling rate was
controlled by a free-running oscillator of the type used as computer
clocks.  It was not referenced or phase-locked to the main system clock.
You might check your schematics to see if this is the case on your system.
If so, it should occur with all nuclei and sweep widths with a constant
factor.  There is also a divide-by circuit.  Check to see if the problem
occurs at extreme maximum and minimum sweep widths.
___
do you see 13C satellites for CHCl3? if yes, is 1JCH also scaled by
a factor of 8.18/7.27?
___
I once had this happen on an AM-250.  Turns out that someone had set the
base frequency to some random number.  Loading the standard job file did
not help.  Do a "cf" and check the basic spectometer proton frequency.
___
Are you sure the lock solvent is really CDCl3?
CHCl3 chemical shift in acetone-d6 is 8.02
and DMS0-d6 is 8.32 relative to TMS.
____
What I don't understand is how the TMS is still at zero. This sounds like a
digitizer problem to me, with the chnages in sampling rate not being
correctly interpreted during the FT. Try powering the whole thing down and
then starting it again and see what happens, very often these kind of
problems (with AC's)just go away after a power cycle! However if it is the
digitzer board this could be expensive. There are a lot of junked AC/AM's
lying around so I would post an appeal for parts rather than go to Bruker!
____
This may sound off-the-wall, but are you sure there is no DMSO in your
sample ?
Chloroform in DMSO resonates _at_ 8.2 ppm.
Even 1 drop of DMSO in a chloroform solution will make the CHCl3 signal move
a long way.
____
Did you checked the parameters file? Maybe it was altered.
Based on my instruments (DPX-300 and DRX-500) one of the paramters that affects
the chemical shift scale is [sf} which can be found in [edp]. Maybe a user
changed the value and save it with other solvent defined. The value of sf
depends on the solvent used.
____
A possible solution: are you using the EXACT same sample on the 250 and 300?  A
common problem encountered with chloroform is that the chemical shift of
chloroform is highly dependent on the sample matrix and concentration.
Therefore, when different samples are referenced to TMS at 0, you may observe
the CHCl3 in the sample to have a shift of anywhere from 7 to about 8.5 ppm,
depending on the sample concentration and matrix.  Everything is relative.
____
Have you checked SF0 and SY to make sure that they are set to correct
frequency?  Incorrect settings would lead to the problem that you describe.
____
Have you CAREFULLY checked your "processing" parameters?  (Not just the
acquisition
parameters.)  We had experienced something similar and found one of our
processing
parameters was inadvertently altered -- I don't remember which one now, but I
suspect it
was the SF for processing.)  I suspect this is a software issue, since TMS is
unchanged.
____
(A) Software -- I assume that the system software is properly configured,
etc.  In particular there is usually a way to input the nominal
spectrometer frequency as a software variable to calculate correct chemical
shift information.  If this variable is wrong, then you would see a result
as you describe.
(B) Hardware -- An NMR spectrometer is designed to maintain a fixed ratio
between the various frequency sources.  The absolute accuracy of the master
frequency source is not necessarily controlled that closely, but the ratios
are maintained to extremely high accuracy.  The denominator term in these
ratios is the nominal lock frequency.  Typically, the reference frequency
for all synthesizers is a single, high spectral purity, crystal -controlled
master oscillator.  Its absolute frequency may be accurate from anywhere to
1 part in 10E6 of its nominal value to 1 part in 10E9 if the master
oscillator is in a temperature controlled oven.  The idea here is to have a
master oscillator with high spectral purity.  The absolute frequency
accuracy is irrelevant within limits.
Each frequency synthesizer will have the means to generate offset
frequencies.  If an offset synthesizer has failed, then unexpected
frequencies would be produced.  In your case, if the 2-H lock offset
synthesizer is not operating correctly, the system may still manage to
maintain a stable field-frequency relationship between the 2-H lock
transmitter (which is in error) and the 1-H observation transmitter
frequencies (which will also be in error).  The offset error in the lock
frequency (in Hz) will be about 1/6 the error in the 1-H transmitter
frequency.  In your example this would be about 6 MHz producing a 36 MHz
error in the 1-H transmitter frequency.  The effect of operating at the
resultant higher 1-H frequency would be to produce the results that you
describe with respect to chemical shift -- provided that the software
variable was set to a nominal 300 MHz.
I might add that if the above scenario (B) is correct, then the 2-H
frequency was probably in error at the time the system was originally
installed.  The installer ran the field up to the point that he or she
observed a resonance for either 2-H or 1-H in a test sample and never
noticed that the spectrometer was in fact operating at a higher field strength.
Assuming that there is not an error in a software variable (a) check the
2-H frequency source and the 1-H frequency source with a good digital
frequency counter (b) also check the frequency of the NMR master
oscillator.  In the second case (b) if its temperature controlled oven has
failed, it may be in error by several MHz.  BTW, you must locate a place in
the circuitry where the frequencies are available on a CW basis.
Received on Thu Feb 17 2000 - 14:01:43 MST

This archive was generated by hypermail 2.4.0 : Sat Jun 03 2023 - 14:37:30 MST