We are seeing bad 13C data on our Mercury-Plus 400, something that just started a few weeks ago.
The interesting thing is that these bad points only happen in the imaginary part of the FID, the part that Varian does not display with the df command. So at first we could not understand why we had huge sinusoidal waves in the spectrum. Then using older software (Felix) I looked at the imaginary FID and saw huge spikes. Here is an example:
25980
-692
-7389
25981
1240
26178
25982
-68825
12172
25983
43404
-45319
25984
-19757
18030
25985
-64645
28127
25986
5474
-30964
25987
-31484
-10322
25988
-46575
33885
25989
-38847
5117
25990
57459
-22061
25991
1982
-12452
25992
-1659
35067
25993
48493
21938
25994
-25257
-14791
25995
71112
47552
25996
-22845
44803
25997
-61452
11573
25998
-34056
-15393
25999
32992
-44518
26000
2768
33562160
26001
43839
39429
26002
19109
-4976
26003
-5377
13840
26004
35741
-26149
26005
8590
14844
26006
446
33556088
26007
21381
10312
26008
86632
-80
26009
2100
-50525
26010
16550
11420
26011
-5781
-30042
26012
-53614
30725
26013
-18545
21258
The first column is point number, the second column is the real data point, the third column is the imaginary. The two bad points are points 26000 and 26006. We usually only see one bad data point in an FID, sometimes we see two.
The problem is sensitive to the receiver gain (gain) setting. At a gain of 12 or lower we essentially never see the problem. As the gain is increased, the frequency of this problem increases. At the maximum gain setting (38), we almost always get a bad data point in a 32 scan spectrum.
I'm wondering if anyone has seen this kind of behavior, or if they have an idea of what causes it. Is this a receiver problem, or a digitizer problem? We have never seen it in a proton spectrum, but this could be because the gain setting is usually low and we don't acquire as many scans for proton.
Any suggestions for how to diagnose and repair this problem will be greatly appreciated.
Neil
Received on Mon Oct 05 2015 - 08:42:27 MST