Website update

I uploaded a bunch of new documents to the website this morning. I wanted to make note of a few of these.

#1) Sarah and I will begin implementing a new training protocol in the fall. See the training page for details. New users will have to study the document "Things to know before operating an NMR spectrometer" and pass a test before training. Then Sarah will walk through how to record a 1D 1H spectrum. She'll create their account but not activate it. Then, when they have a sample, they'll do a follow up training with me.

I'm trying to combine the best aspects of user training that I've observed by looking at NMR facility sites at different universities. Most of our user training procedure comes from Dave Vander Velde - formerly at KU, now at CalTech. I got the "Thing to know ..." idea from Jeff Simpson at MIT. I got the follow-up training idea from Rich Shoemaker - one of my mentors - at CU-Boulder. I got other good ideas many, many other facilities (too many to list).

#2) We're going to add checklists for every open-access instrument. I got this idea from Atul Gawande, a surgeon at Harvard Medical School. I'll add an essay I wrote about the philosophy of checklists tonight or tomorrow. Check our documents page for the checklists and keeps your eyes open for these on the instruments. Hopefully these help us with "avoidable failure".

J

update on BigSur workstation upgrade

We've had some issue with upgrading the workstation for BigSur. I don't want this to turn into a whiny blog, so I'll try to stick to the facts.

#1) I was impressed with the feature of Topspin 3.0 when I saw the demo at the ENC and I thought that this software would be very convenient for users of BigSur. When faced with the need to upgrade the workstation, I saw an opportunity to expand our capabilities.

#2) Topspin 3.0 requires a 64-bit computer and CentOS linux. We purchased a nice Dell Precision T3500 64-bit workstation and I installed CentOS. I'd configure the workstation and was ready to install Topspin 3.0 and connect to the spectrometer.

#3) Then my good friend Asokan mentioned that he didn't think Topspin 3.0 would work with older consoles. Older consoles - I thought maybe it wouldn't work with DRXs, but BigSur is an AV, purchased in 2004. Turns out he was right. Topspin 3.0 requires hardware that BigSur does not have (a digital receiver unit rather than a RX22 receiver). Hence, even though this console is only 6 years old, it is not eligible for the latest software. NMR technology is advancing rapidly!

I'm a bit frustrated because I told my sales rep my BH number before placing the order for TS3. Clearly, I should have asked more questions, but it doesn't seem too much to ask for Bruker to not sell me things I don't need. Lot of negatives in that last sentence - sorry.

#4) I reinstalled RHEL4 and am configuring the workstation. Strange issues with this too, because the NIC on the motherboard isn't compatible with my kernel.

http://www.linuxcompatible.org/RHSA-20100394-01_Important_kernel_security_bug_fix_and_enhancement_update_p144094.html

Fortunately I have another NIC, but I get to go through the fun of setting up one NIC in Proteaus to upgrade the kernel, then reset Proteaus with the second nic.

#5) We're still waiting for license for TS2.1.

Hope that wasn't too whiny. Cowboy up.

BigSur Workstation

This afternoon I shutdown the workstation for BigSur. Sarah had noticed that the computer was making an horrible sound and I wanted to troubleshoot it. The CPU fan was the culprit. Since fans are cheap, it would make sense to replace it, however, the fan is a very unusual size (65mm x 65mm x 10mm). I scoured the internet and went to three different stores in town and couldn't find anything!

Because of the high demand from the CMLD and SCC, I didn't feel like I could keep this spectrometer down. The fan seems to be functioning properly, so I decided to reattach the workstation to the spectrometer, even with the noise. Then I read that you can fix a noisy fan with a few drops of oil (http://www.zdnet.co.uk/news/processors/2003/06/03/fix-noisy-computer-fans-with-a-drop-of-oil-2135514/). I was skeptical, but also desperate. Maybe it is wishful thinking on my part, but it seems a bit quieter. Tom Coombs said it sounds like trickle of water. Think of the NMR lab as your zen rock garden and come to mediate to the soothing sounds of a gentle mountain stream.

This issue does add some urgency to updating the workstation for BigSur.

Justin

Swapped Probes in Paris

This morning I swapped the TXI probe in Paris for the BBO. The BBO probe should give better results for 13C-detected experiments such as C13CPD and C13DEPT135. 1H-detected experiments will suffer a bit, though I doubt anyone will be able to tell. The other advantage of the BBO is that it is better suited for VT-NMR and we can do "exotic" nuclei. In fact, I'm going to do some 14N this afternoon.

The BBO is a bit shy of lineshape and 1H and 13C sensitivity specification, but we'll get that worked out, soon. Another important task is to make 3D shimmaps of this probe. With the shimmaps in place, we'll (hopefully) be able to shim the probe up faster after swapping it in.

Justin

BACs on BigSur back up

Last night the BACs autosampler failed to run the queued samples. This is the 2nd night in a row that the BACs failed. I'm not sure what caused these failures or if they are related. My hypothesis is that someone interfered with the BACs while it was in motion loading or unloading a sample. I have observed that this confuses the software and stops the automation.

I sent an email to all bigsur users and put a sign on the BACs reminding people not to interfere with it when it is in motion.

Justin

Not making lineshape spec of Yosemite

Unfortunately, we can't make lineshape spec on Yosemite (spinning hw = 0.38 hz; hump = 6.1/14 Hz). Fortunately, we are close enough that most users will not even notice. The problem is high-order off-axis shims, which can be tricky to dial in. We'll keep working on it and hopefully soon be able to report that we are making spec or at least that we are getting closer.

Justin

Lineshape on BigSur

While troubleshooting another issue, I'll see if I can shim to lineshape spec on BigSur. For non-spinning I get hw = 0.42 Hz and hump = 6.1/11.3 Hz, which is close enough to spec, that I'm thrilled. In fact, these are the best non-spinning results we have recorded since we began keeping records in October 2009. For spinning, I get hw = 0.35 Hz and hump = 3.3/8 Hz, which makes spec. I like these shims so much, that I update the shim file.

The KU NMR Lab Blog

Over the last couple of months I have been updating the website (http://www.msg.ku.edu/nmr/home.html). One issue that concerns me is communication with our clients. There is a listproc mailing list, but I only like to post highly critical messages to that list. It occurred to me that we needed a less formal way to keep our users abreast of non-critical developments in the lab, including things like
*Continued developments on the website
*New protocol
*Instrument specifications

So I decided to start a blog. Let's see how long this lasts. I'm guessing about a month. Hopefully, I can keep it going a bit longer than that, because this blog may give me a good place to chart the progress of the lab.

Justin