Teaching
Spectroscopy: Off resonance FID
This is a Free Induction Decay (FID). It is a signal in the "time domain", i.e. the y axis is signal intensity and the x axis is time.
What is the length of the FID?
The length of the FID is order of magnitude in the ms to s range (the range of T2*).
What is the frequency of the signal?
The signal has on the order of 10-100 cycles in this time frame of 100 ms or so. Therefore, this is a relatively low frequency signal, certainly not in the 100MHz range of the signal picked up in the coil from the rotating magnetization of the spins. The signal has been put through a phase sensitive detector, which "beats down" the frequency to a range for low frequency digitization.
Why are there two FIDs on the screen?
There are two phase sensitive detectors, 90 degrees out of phase with each other. The output from both are displayed.
Spectroscopy: On resonance FID
This is a Free Induction Decay (FID) on resonance. It looks like an exponential decay. In this case the frequency of the transmitter is exactly the same as that of the spins, and therefore since the phase sensitive receiver beats down the frequency relative to the transmitter, the signal appears to be a decaying signal with 0 frequency.
How do I get my signal on resonance?
A signal can be set on resonance either by changing the frequency of the transmitter (BF1 or O1), or by changing the "field" such that the Larmor frequency shifts to be equal to that of the transmitter. Note that this is a very small % change of Bo, since the shift will be on the order of Hz relative to the Larmor frequency of MHz.
Spectroscopy: Receiver Clipping
When the receiver gain is set too high, the FID is "clipped". This results in a flat portion at the beginning of the FID. This is similar to a square pulse, and therefore when the FT is taken of the clipped FID, a sinc type baseline artifact is introduced in the spectrum.
Since the number of spins is proportional to the first point in the FID, the receiver gain should be set to avoid this clipping of the early portion of the FID.
Note that when one runs a pulse sequence such as inversion recovery, one needs to make sure that the FID will not be clipped even for the highest signal intensity in the series of FIDs acquired.
Spectroscopy: Double resonance FID
When two resonances are present in the spectrum, the two frequencies in the FID are combined in a "beat" pattern as seen above.
When one of the resonances is on resonance, the FID looks like this.
Imaging: An Echo and Its Transform
The echo is seen on the right hand side of the display, in the acquisition window. This is the time domain signal. The frequency domain signal (FT of time domain) is in the left side, reco window.
The left side display is a projection of the object to be imaged in the read direction.
If the sequence being run is a multislice or multiecho sequence, the echoes from the various slices / echo train will be displayed on one display, as shown here.
Imaging: A Clipped Echo
When the receiver gain is too high in an imaging experiment, the peak of the digitized echo is "clipped", as seen above. Again, as in spectroscopy, note the "sinc" like function that it creates in the frequency domain (projection) signal.
The transformed image will show these
sidebands as "halos" around the image.
Debugging: No Signal
If there is no signal at all, the first thing to note is whether or not there is noise being digitized.
If there is no noise present in the spectrum / echo, the most likely problem is a connection with the preamp / receiver. Check these out first.
If there is noise present, then the receiver is on, and one must check for problems with the signal itself:
- Is the resonant frequency correct?
- Is the sample in the coil?
Debugging: Poor SNR
Is there not enough signal or too much noise?
In general, one should be familiar with the noise level for given experimental conditions. If the noise has increased substantially, there may be a problem with the preamp or pickup from an outside source of rf.
If there is not enough signal:
- Is the coil tuned and matched?
- Is the sample positioned correctly?
- Is the frequency set correctly?
- Is the flip angle appropriate?
