FW: AMMRL: ToF Glitch with 2 kHz Birds - Summary

From: <don.eldred_at_dowcorning.com>
Date: Wed, 17 Nov 2010 16:19:14 -0500

Thank you for your kind contributions in diagnosing the 2 kHz spikes we were getting off the transmitter offset frequency (tof or O1 depending on instrument manufacturer). I had tinkered around with several of the suggestions prior to sending the email, but as intermittent problems go, it never wanted to misbehave while I was actively trying to diagnose it.

I thought we had gotten ahead of the issue by putting in a new band pass filter on the X-Channel, however, after 7-10 days of good behavior the glitches have returned.

I've captured the responses to the original post below. There are still several approaches that need to be investigated here, but I wanted to gather this quick summary for the group and convenient documentation purposes.

BTW, The producer of the FSY-Microwave filter has changed owners. At the time of this email you can locate the vendor on the web as either Spectrum Microwave or Spectrum Control.



Original Post:

> We have a Varian Inova NMR spectrometer (400 MHz), which has been
> returning intermittent ToF glitches when acquiring 29Si data. This center
> glitch is accompanied by 2 kHz chirps across the entire spectrum.
>
> The trouble is that the errors occur primarily under automation
> conditions. When I sit down at the console to reproduce the glitch the
> instrument works splendidly. It's also interesting that the problem seems
> to be confined to 29Si acquisitions, 13C experiments don't show this same
> problem.




Responses:


2 kHz to me suggests a lock problem, as that is the pulse rate for the lock

---
I'm not sure, Don, but I think the Lock multiplexer runs at 2KHz.  Try turning off the lock to see if your symptoms change.
---
This sounds like it could be 1H decoupling noise "leaking through", esp. WALTZ16 and GARP decoupling often show this kind of thing.  I actually replaced my low-pass filter in line with the Broadband signal-path (between the probe port of the BB-pre-amplifier and the probe) with a much more "robust" one made by FSY-microwave.  To check this, have you tried observing your 29Si w/o 1H decoupling?  Or change the "dmf" and see if the chirps move around.
If this seems to be the problem, check the parameters being used by the 29Si in automation vs. when you do it manually.  If it turns out to be decoupler noise, I minimized this even more (for all broadband nuclei) by adding a low-pass filter (that blocks frequencies above 31P) in the receiver chain between the "output" of the BB-pre-amplifier and the mixer.  This blocks any high frequency noise from getting past the output of the pre-amplifier.
Of course, this might have nothing to do with 1H decoupling, .. in which case... never mind!!  ;o)
(it's always hard to troubleshoot with limited information, but it's easy for you to check whether this is relevant to your situation).
---
        Check the output of your PTS, sometimes the filter caps will begin to fail
(although the normal noise are 60 Hz spikes).
---
... the reason that 15N and 29SI tend to be so glitchy is that the nuclei are not very insensitive, but the relaxation delays required are so long that we're really looking at the baselines of too-rapidly-pulsed data acquisitions and we just don't realize that there's very little signal there and we're looking at glitches that wouldn't be visible with another nucleus.
He did some time doing solids NMR at Argonne and tells me that solid state 29Si required 20 seconds (actually, he might've said 20 minutes) between pulses. Of course solution-state would not be that bad.
---
The 2 Khz glitches sounds like automation is setting the lock power too high?
Make sure the lock power & gain values are set correctly in the probe file for the solvent you're using.
Also, make sure you have the appropriate band bass & lock reject filter in place on the BB preamp.
---
Having worked exclusively with Bruker systems I may be misunderstanding you, but having worked with 29Si I think I have observed the same phenomena, although without the 2 kHz chirps.  Without knowing your power levels and pulse durations, its hard to speculate why the occurrence is only on 29Si.  In my experience, our Broadband Observe (Bruker BBO) probehead (which runs the bulk of our 29Si, 13C, and 1H) has never had this issue.  However our older 29Si selective probehead would sometimes give me an O1 glitch even with DQD filtering (Bruker speak for your issue, I think).  So for instance say the sweep width is 50 to -150, then the spike would be at -50.  It took forever to find the simple solution of increasing the DE (Bruker pre-scan delay in microseconds) which eliminated the spike.  It was hard to solve because our digitizer was in digital mode using DQD filtering, which is main purpose is to filter out O1 transmitter spikes.  Its hard to say why its happening in automation and not manually, especially for me, because while I am very familiar with Bruker automation, I am completely unfamiliar with Varian.
I hope this helps, and if you want to talk through it feel free to call.
---

Received on Wed Nov 17 2010 - 11:19:25 MST