15th-19th Nov. 2004

Oxford engineer (Charlie Smith) installs and energizes magnet without any quenches. He topped up

Problems: Helium Meter does not work

Supercon (cryoshims) fell out when removing charging stick

Supercon Values are:

Z1 -16.563 ZX 1.322

Z2 9.394 ZY -3.433

X 1.703 C2 -1.657

Y -0.181 S2 -3.705

NB lots of current in Z1 and Z2

7th Dec 2004

Mayan Tuazon arrives from Varian. Tops up He

10th Dec 2004

Varian engineer had to recycle the cryoshims. Z1 had fallen out! Topped up He once again.

New supercon values are:

Z1 -16.509 ZX 1.237

Z2 13.45 ZY -2.866

X 1.806 C2 -2.371

Y 0.493 S2 -3.704

13^th Dec 2004

Problem with FTS. Iced up. Problem with compressed air? Fixed and working again (manufacturer helped)

14^th Dec 2004

Overnight drift test showed that, magnet is very unstable. Not only was drift very large but the nicely shimmed singlet gradually split after 3 hours!! Check on net indicates Z1 might be unstable still. Engineer suggests that we might have to redo supercon shims!!He Meter still not functioning

For F19 cannot tune to zero ~ 200 is OK!

15^th Dec

Helium port for transfer of He liq iced up!! Removed by blowing He gas over very briefly

After consultation with bosses at Varian and Oxford it was decided to redo super con – Z0 by adding ½ amps. This is to make the field go down (at the moment the field is moving up, and losing shims (Z1 we think).

Still concerned about the He meter and the He boil off. Should be less than 35 cc/hr.

17^th Dec

After cycling Z0 and overnight drift test (stored as 4^th) we see the field drifting in the correct direction. However there is still a splitting. Varian engineer insists it will settle down. Good news that it is refitting in the correct direction. Engineer leaves asks me to install CD r/w and Internet!

19^th Dec

Ran an overnight drift test

20^th Dec

Results of drift test show the magnet has settled – however, still splitting.

21^st Dec

Drift test results again show loss of lineshape and splitting overnight

21^st Dec - 14^th Jan

Drift tests indicate shims being lost

13^th Jan 2005

Oxford Engineer (Ben Hubbard) comes to install new He probe and meter. It now seems to work. Meter seems to have been faulty!

14^th Jan 2005

Helium Fill (by me) did not get to 100%. Need to do this at some point to recalibrate meter. N2 filled

2^nd Feb 2005

Oxford (Ben Hubbard) recycle shims and ramp Z0 due to problems with Z1 (and Z2) stability).

New Supercon values:

Z1 -17.22 ZX 1.237

Z2 13.45 ZY -2.866

X 1.87 C2 -2.371

Y 0.6 S2 -3.704

Note also had to reset lock freq. and change RT values.

After cycling, tried to get RT shims to spec – unfortunately I could not manage it. Might need Varian to remap RT shims

4^th-7^th Feb

Managed to shim D2O down to 4.5 Hz, chloroform to 0.8Hz – still can't get much lower.

Weekend drift test showed splitting, similar to that seen before latest cryoshim cycling.

11-14 Feb

Shimmed chloroform to 0.7Hz – still can't get much lower.

Weekend drift test showed splitting, similar to that seen before latest cryoshim cycling.

18-21^st Feb

Now up on the net. Varian engineer shimmed chloroform down to 0.5 Hz . Drift test showed splitting..

24^th Feb

Varian send bursting disks for replacement. Ask me to replace (I refuse). Varian believe a He leak is responsible for the faulty reading on the manostat (probably through the bursting disks). I place rubber gloves over the disk outlets – no evidence of a He leak. I personally think the manostat might be faulty, but Varian insist it is the disks!

2^nd March

Claude Major (Varian) comes to check for He leaks (since manostat reading never reaches the setting). Turns out manostat is not working properly (so no need to change the bursting disks). Manostat was repaired: setting 1020; reading 1021

7^th - 8^th March

Varian run drift test: splitting seen. Varian (Ed Fadriallan says he will send a weather station to check environment

15^th March

Installed Biopack but gives no signal from HNCO, and other NH-based triple resonance. HSQC's OK. If use TROSY option, lots of signal. Maybe arcing of probe when do simultaneous 15N and 13C pulses (not the case – see later)!!

Note: probably due to 15N – compression was way off expected value of 1.0 (~ 0.7). Not a reliable diagnostic, admittedly.

Varian will send over a temporary “replacement” probe!!

16^th March

Managed to get signal from triple resonance!

Re-ran biopack calibration after rebooting console. Now get signal. Checked for attenuators on 2^nd and 3^rd channels. Couldn't see any! Seems they were not installed!! Varian sending attenuators! Note: with higher powers: 15N compression factor was ~ 0.7, now it is close to 1!

Runs keep crashing. With dsp='r'; get the error message:

Rcvr1: Number of points acquired is 61604 > than np!

Christine Hofstetter at Varian informs me this may be a problem with the DTM board, which has shown similar problems before!

18^th March

Ran 15N -edited NOESY with dsp='i'. Did not get the error message (Rcvr1: Number of points acquired is 61604 > than np!), which suggests the DTM is not the problem, maybe the ADC is not working properly.

However, running dsp='i' does give some corrupted data. Varian have been informed and are trying to work out what it is.

19^th March

During calibration (biopack) the low power 13C pulse profiles are very erratic (jumpy). Indicating the modulators may be faulty. Have informed Varian. They recommend running the auto tests with CH3I. I will do this after running on Laurents sample.

22^nd March

Running HNCA on Laurent's sample, 80 scans

29^th March

Attenuators have arrived! As has the replacement probe and the weather station! Will run tests on old probe first (autotest) and then replace attenuator.

29^th Marc

Did the following with the attenuators:

channel 1: left as was (1dB)

channel 2: replaced 4dB with 6dB

channel 3: replaced 1dB with 6dB

channel 4: replaced 1dB with 6dB

using autotest sample (shimmed ~ 7Hz) ran autotest for channel 1 and channel2. And ZXY gradients (did not do heating and variable temp tests).

30^th March

New pulse widths with attenuators are:

1H (channel 1) 7 @ 59dB

13C (channel 2) ~ 15 @ 59 dB

15N (channel 3) ~ 40 @ 59dB (may use more power if you wish!)

2^nd-4^th April

Ran first 3D – get corrupted data. As seen before with dsp='i'. Obviously, ADC or DTM not working properly. Varian have sent ADC board (29^th March), but cannot get engineer here until week of 11^th April.

5^th April

The 3D experiment stopped saving data about halfway through the experiment (but continued acquiring). The fid's were being recorded (according to celem = 5549) but the number of fids that we could ft were 3858. Eventually killed the experiment.

Started drift test: noticed that during acquisition, the acq light did not illuminate: traced this to the cable to spectrometer which was loose (but it still recorded the fid).

6^th April

Results of drift test show loss of lineshape resolution after less than 12 hours, to eventual splitting (after 12).

Note comparing shim values to march 9^th shows:

Z1 changed by 897 DAC units! This suggests Z1 is still unstable!!

7^th April

Ran another drift test which showed the same splitting within 12 hours!

Still waiting to hear if VARIAN are sending an Engineer to install ADC board. In addition I have informed Varian and Oxford about the drift tests failing!

11^th April

Still getting corrupted data!

Ed Fadriallan will get someone to run a drift test! Show little variation over 12 hours, but after 17 hours the linewidth had doubled.

19^th April

Claude Major arrives from Varian. Installs a new manostat, DTM and ADC boards. Also fixes cable for lock display.

Noticed problem with lock – always seem to be locked “on” even though, we ask it to be locked off. This was seen since the beginning, but shifting Z0, and not getting a change in lock response. I had mentioned this to Varian tec service – they seemed to think it was OK: obviously was not!

22-25^th April

Ran a 13 hour 15N-edited hsqc- noesy. Saw no artefact's!

Ran a 48 hour drift test with lock gain and level set to zero. Saw splitting, and needed to change Z1 from 4306 to 4253 DAC units after run. i.e. Had drifted by 71 units in 48 hours!!!

26^th April (Tuesday)

Mayen (Varian engineer) arrives.

Runs an overnight drift test

MSR board was replaced for lock.

Engineer noticed an extra release valve on the Helium boil off (before the manostat) – could be the problem of the drift tests! No idea who should have noticed it was there (Varian say Oxford, Oxford say Varian)

28^th April

Lock still playing up (Mayen “fixes” this)

29^th April

Mayen leaves, but I noticed lock is still not fixed. Phone Varian. Apparently, Mayen noticed a loose ribbon cable. Varian are sending a new one

2^nd May

Apparently, Mayen noticed a loose ribbon cable, which may be responsible for the lock problem. Varian are sending a new one.

4^th May

Try to install new Biopack version (not just update). Cannot do this! When I inquire with Varian, they say I have been hacked.

5^th May

Back up all data (vnmr1, vnmr).

6^th May

Claude Major arrives to install ribbon cable for lock and reload system software.

Reload solaris version 9, from disks.

After installing cable, Claude notices the lock is still playing up. Ask for lock transceiver board to be sent.

Saw the number point < np error. Claude reinstalled existing DTM and ADC

We are disconnected from the Internet for the time being.

6^th -9^th May

Run a drift test on doped D2O sample. For 60 hours. Results look promising. Will repeat on 10^th May with lineshape sample.

9^th May

Installing new Biopack. Installation OK.

Whole new lock system is sent. Claude Major installs this.

Get number point < nip error. Claude reinstalls the DTM and ADC

Everything seems to be OK with lock and nip error.

Can now run with dsp='r'

15^th May

After a number of drift tests we see splitting once more.

Oxford and Varian are discussing what to do

18^th May

Oxford decide to either recycle all the shims or do a controlled quench

19^th-24^th May

Tests with Biopack – T1 and T2 data acquired, looks OK. Used hdwshim='y' option for any Z1 drift during experiments

31^st May- 2^nd June

Ben Hubbard (Oxford) arrives and does a 53 step shim cycle on 5 cryoshims. Ends up putting in shim values originally mapped by Varian. i.e.:

Z1 -16.509 ZX 1.237

Z2 13.45 ZY -2.866

X 1.806 C2 -2.371

Y 0.493 S2 -3.704

6^th-10^th June

Ming Quan from Varian comes to map cryo and RT shims.

Cryo shims changed on Tuesday 8^th June:

Z1 -16.196 ZX 0.911

Z2 13.737 ZY -3.91

X 1.973 C2 -2.37

Y 0.28 S2 -4.128

10-15^th June

Drift tests show splitting still

21st July

Oxford and Varian agree to replace magnet. Possible delivery date is Mid August

23^rd July

Noticed Biopack given poor signal (as seen before 15-18 th March). Looks like some flaky RF – I managed to get signal back by turning off the system for a couple of hours. Not an ideal case. NEED TO KNOW WHY THIS HAPPENS

9-11 August

John Day – Oxford arrives to de-energise magnet

16^th-18^th August

Mark Ashmore (Oxford) arrives to pre-cool new magnet. Status: He filled, needs to be energised (next week)

22^nd-24^th August

Nick Wilson (Oxford) energises magnet to -119.57 Amps (92.07 MHz 2H) and sets the following superconducting shims:

Z1 -3.357 ZX 2.207

Z2 15.758 ZY 4.044

X 2.29 C2 -5.27

Y 7.057 S2 3.458

Z3 0

Note the extra supercon shim Z3 (doesn't seem to do much as it has no current though)

30^th Aug– 3^rd Sept

Varian Engineer (Kirabel) maps cryo and RT shimms

Sept-November 2005

Drift tests show magnet still has a Z1 drift. Seems to be settling

November 8^th 2005

Unable to tune 2H on RT HCN probe. Sent back to Varian for repair. Installed loaner HCN probe

December 7^th 2005

Agreed with Varian to write off system. Not enitirely happy as still have the flaky problems (e.g. Flip back pulses and triple resonance in Biopack, see 16^th March, 23^rd July) , but only this way will we get the coldprobe installed

December 13^th-14^th 2005

Varian engineer (Deane Alexander?) comes to check system. Agrees there is some weird behavior with the shim system, after ejecting and reinstalling the shims sometimes get worse and sometimes get better, a very strange occurrence. He agrees something is wrong but we decide to let this go!

I Should have mentioned the Biopack problem. But unless it is actually present at the time, it cannot be overcome.

January 10^th-26^th 2006

COLDPROBE installed by Claude Major.

January 29^th 2006

WFG on 3^rd Channel not working – I install a new one. Biopack sequences not working again. Same problem as before (see 15^th March). Calibrations for flipback does not work, and triple resonance only works with TROSY (well gives some signal)

Tried to problem shoot with Varian but not able. In the end a reboot resolved the problem. This problem is still unresolved.

Feb 8^th 2006

Helium filled

Feb 5^th-10^th February 2006

Seeing error Rcvr No. of points acquired less than 2. Varian promise to send a new DAC.

Also find that get ADC overflow, when there is no obvious occurrence in the spectra (using ddff(1)). Sometimes see the recvr overflow light on the remote status come up (and no ADC overflow on the SUN). Will see what happens after replacing DAC.

H2 90 deg pulsewidth seems very long (~ 150 microsec at 61 dB; factory was 121.8 at 58 dB)

Feb 12^th 2006

Numerous problems. No pnts > 32668 continuously occurs. Impossible to run anything. So I reseated the DAC. Saw all sorts of stuff including Lost lock during acquisition, FIFO underflow, unable to find Acode set, cannot find Acq stat . MOST SCARY WAS I TURNED OFF COMPUTER AND CONSOLE FOR 20 MINUTES. WHEN I RESTARTED THE SAMPLE WAS FROZEN..

Also running the cyroburn produced a lot of garbage. Is the ion pump working?

Feb 14^th- 15^th

Tried to run a cryoburn, experiment crashed due to rcvr points error . Acq status was showing it was idle, but remote status indicated it was still pulsing - also got a error on the pfg console.. Resetting with setacq did not alter this or by rebooting console. Every time I type su the same thing happens.. this is crazy!

Managed to reset it with setacq – seems to be random.

Once again sample frozen when console /spectrometer was turned off. And rebooted. Eventhough VT-On is regulated. Seem to need to reset VT, whenever reboot spectrometer.

PFG problem tracked to a blown fuse – most likely happened due to DAC board not operating properly).

Feb 17^th

DAC Board arrived (it was sent to the wrong address). I installed it. Errors wrt overflows, revcr points etc have gone.

However, 15N is still not working properly, on both channels 2 and 3 (tested with benzamidine sample). Could be arcing (might happen if ion pump is not taking the crap out).

Feb 18^th-20^th

Ran a cyroburn with just ion pump. Looked very noisy. So I attached multi-task pump and did cryo-burn over the weekend.

Error with cold probe - “loss of cooling” power, for some reason lost water supply to, all seemed OK when I arrived. Must be a dip in pressure or flow

Feb 21^st 2006

Claude Major (varian Engineer) arrives, agrees with me that 15N is arcing. Varian recommend to send the probe back for repairs.

Feb 22^nd 2006

Replaced Coldprobe with HCN RT probe.

Calibrations with Biopack showed same problems as before (see 16^th March, 23 July, Jan 29th) i.e. No signal for most triple resonance experiments, and flip back calibrations in BIOPACK do not work. Tried reducing power levels by 3dB on 13C and 15N to no avail...

This problem keeps returning and can be removed by rebooting. But I want to get to the cause of this. Maybe it is diagnostic of common falut that has led to the changing of so many boards etc, in the console...

Feb24th 2006

Got another rcvr points error number of points acquired is 2 less than np. Restarted experiment and got the same error (after about 2 days into the experiment)

March14th 2006

Varian assure me coldprobe is OK – must be system

March 20^th

Claude Major arrives and tried to install new digital card cage – no luck as the new one is faulty. The original is put back in.

Checked 15N power level with scope – all seemed OK. No luck with any of this!

March 31^st

Set to install cold probe again. Failed due to loss in cooling power. This was due to filters clogging for cooling water.

Still have not received card cage!

April 1^st 2006

Started cooling cycle for cold probe. Ran a cryoburn over the weekend.

Need to condition on 1H, 15N and 13C first (page 20 coldprobe manual) then do a cryoburn.

April 3^rd - 5^th

Did a cryoburn

Claude major installs yet another card cage. Did not work, experiments would hang after first fid!! Had to go back to UK, so left as is..

April 18^th

Claude Arrives, reinstalls old card cage. Notice thet power levels from PTS is 17dB (4.4V) should be around 11 dB( 2.2 V) . End up using ~ 1.8 V for 3^rd channel (a bit less than 10dB) .

This overcome the problem with arcing on 15N with the cold probe. But DID NOT OVERCOME PROBLEM WITH BIOPACK..

April 19^th

15N pulsewidths were too long after decreasing power from PTS, ended up changing the attenuation after the 3^rd channel amplifier, changed this to 3 dB... (62dB gave a 15N 90 of ~ 33 microsec).

Biopack OK after turning off. This problem is still not resolved

April 25^th 2006

GOT THE ERROR np > 4.... STILL A PROBLEM WITH CARD CAGES/DIGITAL

May 1st

Still waiting for card cages – Varian agree to replace current console with a refurbished INOVA.

May 8^th 2006

FIFO Underflow error – crashed a 3D 15N-NOESY after 8 planes

May 9^th 2006

Deane Alexander from Varian arrives: new card cage installed. Seems to be working OK.

May 16^th 2006

Platform officially opened..

May 18^th 2006

No problems with np errors, etc.. Card cage seems to be working OK.

Filled Helium (was 56%) with 100 Litres

May 31^st 2006

FLOOD in NMR room. ~ 9 pm phone call from security. Glycol leak lost ~ 450 L of glycol. Damage to instrumentation as yet, unknown

June 2^nd 2006

Damage to SUN computer. Repaired over the weekend by microtext (Parc Avenue).

June 5^th 2006

Claude Major comes to check problems with system after the flood. Mini-task seems to have lost it's turbo power. Can;t get a good vacuum on coldprobe. Send away for a new one.

June 7^th 2006

Lost cooling on Cold probe. Hard to know why, but the water pressure was very low (20 psi) the absolute minimum value, This is the most likely problem.

June 8^th 2006

Pressure and flow still low, ask Jp Fortin and R. Chaisson to look into it and hopefully make suggestions to improve the system.

Another unscheduled warm up of coldprobe.. Pressure is low, must be the problem. Not due to the filters as I replaced them yesterday (and back pressure is good)

June 11^th 2006

Another unscheduled warm up of the coldprobe! Happened sometime between 18.30 Saturday (when i set up the experiment) and 9.30 Sunday morning. Again not due to filters

June 12^th 2006

Contacted HASKRIS about a chiller. Forwarded response to JP Fortin. It would appear that we the building guys cannot guarantee a stable supply of glycol that is up to our specs (eventhough they had previously said they could!). Thus our new plan of attack is to use a water chiller

June 17^th 2006

ANOTHER unscheduled coldprobe warm up. Sometime between Friday evening and Saturday morning. I checked glycol pressure etc and everything was in spec..

June 19^th 2006

Another unscheduled cold probe warm up. Luckily I was around to restart cooling before temp reached > 60 K. THIS HAPPENED THREE times this morning.. loss of cooling power was due to low FLOW rate..

June 20^th 2006

Lost cooling power overnight causing another unscheduled cold probe warm up..

Since warm ups are damaging to the probe, I will replace the coldprobe with the RT HCN probe.

June 30^th 2006

Had problems with water suppression in T1 sequence ONLY for relaxT time between 0.3 and 0.7 seconds.. All other values of relaxT are OK..

Finally, I trouble shot this problem to hardware. I noticed that I could not calibrate water flip back pulses in Biopack (The old Biopack problem). So I turned off the console for 1 hour. Upon rebooting the problem had gone. It would appear the old Biopack problem has not been removed by introduction of the new card cage!

Still no news on a chiller.

July 12^th 2006

Exchanged WFG between first and 4^th channels. Overcame problems with Calibration of flipbacks in Biopack and manually

July 13^th

Bugger!! Tried to run T1 and again with relaxT=0.3 seconds I had receiver overflow! SO looks like the changing of the WFG did not help!

July 14^th

Tried using a hard pulse for experiments – still no luck.

September 11^th-22^nd 2006

New console arrives. Installation by Claude Major. Passes all APT tests

September 25^th

Biopack installed. Everything OK

September 26^th

NP greater than errors... after running an overnight T1 experiment.

September 27^th

Track NP error to DTM board using digital diagnostic software.

Replace board with board from old console. Seems OK (at least initially).

October 17^th 2006

Fuses for amps on 3^rd and 4^th channel blown after restarting the system. Changed these, seem to be OK now

October10th - ??

Haskkris chiller has arrived. However, despite knowing for three months, no pump for the housing water supply has been installed.

27^th October new pump installed. BUT. Not enough power (despite having been guaranteed this and the chiller being made to specifications detailed by engineers here, who then cannot meet them - priceless).