i like this post (click again to cancel)
1
i dont like this post (click again to cancel) remove favorite mark from this question (click again to restore mark)

I'm trying to come up with reasonable tests of probe performance for all channels by simulating conditions at which arcing is most likely to occur.

For those tests I usually read voltage of the incoming wave on the RF line using directional coupler with a 50dB attenuated pickup and oscilloscope after running an "experiment" pulsing at the power and duty cycle that simulates the condition to be tested. If the waveform looks nice and square - there is no arcing.

What experiments should I look into? OK for 1H I'd pick TOCSY with the longest mixing time we use. What about 13C, 15N, 31P?

How do you do this kind of testing for solids probes?

How do you test fitness of gradient coils?

asked Jan 19 '10 at 08:40

Evgeny%20Fadeev's gravatar image

Evgeny Fadeev
5571

updated Jan 22 '10 at 07:34

It isn't usually the length of the pulse that is the problem, but rather the peak voltage applied. Long pulses (TOCSY) create heating, which can also be destructive. Careful, trying to force a probe to arc. Arcing can create a carbon track which means it will arc at much lower voltage in the future - Kirk Marat (Jan 19 '10 at 12:07)


5 Answers:
i like this answer (click again to cancel)
4
i dont like this answer (click again to cancel) the author of the question has selected this answer as correct

Since nobody has answered this yet, I'll venture an answer. Hate to monopolize the forum, however.

There are two ways in which the RF can damage the probe. The first is voltage breakdown (arcing), which is an (almost) instantaneous effect. The length of the pulse doesn't really matter except that arcing during a long pulse will inflict more damage than a short one. The effects are also cumulative, in that arcing during many short pulses can be a severe as one long one. Arcing causes the worst damage when it occurs across plastic materials such as teflon capacitors, plastic MAS stators, capacitor wands, etc. In this case the arc creates a carbon track which will cause the probe to arc at a significantly lower voltage in the future. I've seen an MAS probe that should take hundreds of Watts arc at 25 Watts once this has happened. Severe overvoltage can even punch holes in the ceramic and glass/quartz dielectrics used in some capacitors. Another damage mechanism is the formation of rough edges (burrs) on metal surfaces that can create high fields (voltage gradients) and lower the breakdown voltage in the future. Solids probes in general, and narrow bore solids probe especially, tend to be more prone to arcing because they are really pushing the limits on power handling to get the B1 fields required.

Symptoms of arcing include erratic amplitudes and phases when doing pulse calibrations, or even just an array of one-pulse spectra with identical parameters, or severe t1 noise, or erratic pulse calibrations. Note that it does sometimes does take a pulse length of a few us for the breakdown to occur (the first few points in your calibration array will be fine). You can also check for arcing with a directional coupler and a 'scope, but the directional coupler should be in reflected power mode. A gradual lengthing of your 90 deg. pulse over time is not normally a typical symptom of arcing. If this is your problem, the first place I would check is the actual power being deliverd to the probe and the probe tuning. Also check the S:N if it is specified for that probe channel.

If you operate your probes within the limits specified by the manufacturer, then arcing shouldn't really be much of a concern, especially for liquids probes. Unless, of course, someone forgets to turn off the solids amplifiers! - Been there, done that :-(

Routinely pushing the probe to the point of arcing is neither a recommended nor a necessary procedure.

The other way in which the RF can damage a probe is through overheating caused by long pulses. Your vendor will usually specify a maximum length that the decoupler or spin-lock field can be on for a given power (or B2 field strength) level. They may also specify a maximum duty cycle.

For gradient coils, the manufacturer will also specify maximum gradient pulse length and duty cycle. Some gradient amplifiers have fuses or other protection for the gradient coils. The best performance tests for gradient coils are those probe performance tests (e.g. recovery time, gradient strength, etc.) specified by the manufacturer.

Remember, you vendor has likely already tested the probes to destruction, and knows what any given design should handle.

link

answered Jan 21 '10 at 11:05

Kirk%20Marat's gravatar image

Kirk Marat
664

updated Jan 31 '10 at 14:05

Thanks! Very helpful. That's strange but it does appear in our case that 13C 90 simply became longer at the highest power (dpwr=63) and same attenuator set and cabling. I have tested the amplifier and it seems to output the same power as before. Will check the tuning now and the probe cable. - Evgeny Fadeev (Jan 21 '10 at 12:41)

Kirk, thanks for your help. I don't think there is a huge issue of dominating the forum... But maybe you could invite some people that know you well to join and answer some questions? - Evgeny Fadeev (Jan 21 '10 at 15:30)

Getting back to your origin problem. The trick is to determine whether this is a probe problem or an RF problem. Do you have another probe to try? What is the power actually getting to the probe relative to what is coming out of the amp. 1-2 dB loss through cables, filters, pre-amp etc. is norm. - Kirk Marat (Jan 22 '10 at 06:41)

Yes I need to take a look into that more carefully. My other probe (H{CP}) has shorter 90 (the "problem" one is H{CN}), but by specs H{CN} must have shorter 90. I took 2dB's out of the AMT's input and pulse is back to normal with no arcing patterns on the scope in the reflected wave. - Evgeny Fadeev (Jan 22 '10 at 07:32)

If the 13C pulse on the HCN probe has changed, but it hasn't changed on the H{CP} probe, then the problem is likely the HCN probe. If the pulse has changed on both probes,then the problem is likely the RF path. - Kirk Marat (Jan 22 '10 at 10:11)

i like this answer (click again to cancel)
2
i dont like this answer (click again to cancel)

Can you have a well matched directional coupler (D.C. with 50dB attenuated monitoring port) between the Transmitter output and the Probe input port while the NMR experiment is running and monitor the wave forms real time when the NMR experiment is on?

Then the actual conditions when the arcing occurs during the experiment can be monitored. When there is an arcing there can be transients causing distortion of pulse shpes and these can be appropriately diode detected at the DC monitoring port.

Such experiments have resulted in improving the Power setting conditions some times    S.Aravamudhan


link

answered Jan 23 '10 at 03:20

SankarampadiAravamudhan's gravatar image

SankarampadiAravamudhan
91

updated Jan 23 '10 at 03:35

i like this answer (click again to cancel)
1
i dont like this answer (click again to cancel)

Consider adjusting the adjascent turn interspacings in the coil also optimizing for better Field strength inside the coil for given Power. Other peripheral changes would be effective only if this has been taken care of. Such critical situations arise in the case of line narrowing Multiple Pulse techniques for HR PMR in single crystals and there have been instances when 1-2 microsencond 90 degree pulses have been possible for sequence of about 8 pulses per cycle with a cycle time of 24 microseconds and applied repetitively over 100-200 microseconds of acquisition of the FID. S. Aravamudhan

link

answered Oct 07 '10 at 17:56

SankarampadiAravamudhan's gravatar image

SankarampadiAravamudhan
91

updated Oct 07 '10 at 18:00

Changing coil turn spacing isn't very possible. The maker already found the best shape to optimise the RF field. Modern coils are not helical, but are birdcage shape. Changing turn spacing requires making a whole new coil. Physical changes to the coil will also require changes to the tuning circuit. - Jerry Hirschinger (Oct 12 '10 at 10:28)

i like this answer (click again to cancel)
1
i dont like this answer (click again to cancel)

Hi! Jerry, from my PC terminal any added comment does not get posted to this nmrwiki. Hence I resort to use the 'Your Answer' box instead of adding a comment right below your posted comment. "Changing coil turn spacing isn't very possible. The maker already found the best shape to optimise the RF field" . This remark in your comment makes me seek an additional clarification.

Does it mean you are using a commercial spectrometer and your experiments are getting into such realms where the required conditions for RF power (of RF field strength and homogeneity in the sample coil) are beyond the tested specifications and approved by the maker?

S.Aravamudhan

link

answered Oct 15 '10 at 19:48

SankarampadiAravamudhan's gravatar image

SankarampadiAravamudhan
91

updated Oct 15 '10 at 19:51

I am referring to any commercial NMR probe. Altering the coil is non-trivial, and normally done only for probe development. Testing a probe for arcing threshold is also normally done by the manufacturer only during probe development because it is a destructive test. - Jerry Hirschinger (Oct 19 '10 at 11:31)

i like this answer (click again to cancel)
0
i dont like this answer (click again to cancel)

Hi Jerry,

refer to your posting :"I am referring to any commercial NMR probe. Altering the coil is non-trivial, and normally done only for probe development. Testing a probe for arcing threshold is also normally done by the manufacturer only during probe development because it is a destructive test"

I am at the moment not quite sure why you have not considered composite pulse sequences to simulate a 90 degree pulses. This may not be the same as applying longer pulses at lower powers. The power requirement may be less and 90 degree pulse would be longer but it would be as effective as the short high power pulses.

You can explore the "adiaabtic pulses" instaed of the ususal rectangular pulses at the spectrometer frequency.

Aravamudhan

link

answered Oct 22 '10 at 01:56

SankarampadiAravamudhan's gravatar image

SankarampadiAravamudhan
91

Your answer
Please start posting your answer anonymously - your answer will be saved within the current session and published after you log in or create a new account. Please try to give a good answer, for discussions, please use comments and please do remember to vote (login to vote)
toggle preview

powered by CNPROG