I have two ongoing problems with our Inova 600 that are driving me crazy.
The first is low output from all three transmitter boards (as I've posted here before). The output is anywhere from 4 to 6 dB below the Agilent specification, with a corresponding lengthening of the 90 degree pulse lengths. The spectral purity of the boards is also bad, with considerable 40 MHz modulation. On the worst board (TX2) the 640 MHz sideband is only down 12 dB from the desired 600 MHz signal, and the 620 MHz synthesizer LO signal is about 18 dB down. Agilent spec is > 20 dB down for both. I'm not sure if I can attach images to these posts, but I can send anyone scope and spectrum analyzer pictures. The DC voltages, measured on the back of the card cage, are all good, and replacing the 80 MHz REFGEN board didn't help. The synthesizer voltages are all good. The TX board output level will be independent of 80 MHz clock signal and relatively insensitive to the synthesizer LO signal (see my discussion in SpinSights). Losses through the SW/ATTEN board (following the TX boards) are nominal at 6 dB.
I've managed to recover some of the pulse length by removing some fixed attenuators that were installed at the input of the AMT amplifiers, and the system is still functional. However, a 50 us 90 at 63 dB for 15 N (RT HCN probe) is still a bit long. I am more concerned that it will continue to get worse.
Has anyone seen similar problems with these TX boards or have any ideas as to what might be wrong? I am suspicious (and am open to being proven wrong) that the problem might lie with the single sideband mixers used to generate the final TX board output. These mixers use plastic package surface-mount mixers from MCL similar, but not identical to, the problem plagued mixers used for detectors on the receiver and lock transceiver boards. Replacing the TX boards is currently out of our budget, but I might try replacing the mixers. This is not easy, as you need access to an SMD rework station to do it properly. Has anyone tried this? The mixers are available directly from MCL at about $7.00 each. There are 4 per board.
The other problem is unstable shim. The shim will oscillate from good to poor back to good again with a period of about 40 +/- 20 sec. This can easily be seen on the lineshape of a peak and as a drop (and recovery) in the lock signal of up to 40% (acetone-d6 lock). Again, I can forward pictures to anyone interested. The following have all been checked:
- It is independent of sample, but is most noticeable/objectionable on samples with narrow lines.
- It is independent of probe.
- It can also be observed when unlocked.
- Disconnecting gradient cable doesn't help.
- Shutting off VT and removing VT hose from the probe doesn't help.
- "Burping" the magnet doesn't help
- The TMC legs seem OK.
- Nothing in the environment (that I am aware of) has changed.
Anyone have any ideas where to look on this one? It is an Oxford AS600 magnet.
Many Thanks!
Kirk
Kirk Marat, Ph. D. NMR Facility Manager
Dept. of Chemistry, University of Manitoba
144 Dysart Road
Winnipeg, Manitoba, R3T 2N2, Canada
Ph 204-474-6259; FAX 204-474-7608; Cell 204-228-6474
Received on Fri Feb 17 2017 - 08:21:33 MST