transmitter attenuation, to be varied in steps of 2 dB, initial value 90 dB
p1	Gaussian shaped ¹H transmitter pulse, 50 ms
d1	5 s
rg	receiver gain for correct ADC input
ns	1

Since it is not allowed to change the pulse-duration in this experiment, you have to determine the signal strength as a function of the transmitter attenuation. This is best done by recording the integrals. Depending on the shape of the selective pulse, there might be some phase drift during the experiment, which should be corrected.

Whenever pulses with different transmitter attenuation are used in a pulse sequence on the same channel, one has to make sure that these pulses have the same excitation phase. For very recent instruments with linear amplifiers this check might not be necessary; on older instruments, however, it is mandatory. This applies to all experiments with selective pulses described in this chapter, but also to spin-lock experiments like TOCSY (Exp. 10.18), ROESY (Exp. 10.21) and others (Exps. 6.17 or 10.15).

Having found the optimum transmitter attenuation for either the selective or the spin-lock pulses, one has to determine the excitation phase difference between the hard pulse and the attenuated one. For this, a spectrum with a hard 90° pulse is recorded, transformed and phase-corrected; the necessary zero order phase correction is noted. Secondly, a spectrum with the attenuated pulse is recorded, transformed and phase-corrected. The zero order phase correction will probably differ from that in the first experiment.

The difference between the zero order phase corrections obtained in these two experiments is then added to the excitation phase of the attenuated pulse in the corresponding pulse program. Recent software also allows this to be treated as an adjustable parameter of the data set. On repetition, both experiments should then give equally phased spectra using an identical phase correction in the processing routine.

5. Processing

Use standard 1D processing as described in Experiment 3.1.

6. Result

The figure shows a typical plot of integral values versus transmitter attenuation obtained on a DRX-400 spectrometer; an attenuator setting of ca. 68 dB corresponds to the 90° pulse. Note that the dB scale is logarithmic, resulting in a compression of the expected sine curve which starts to oscillate rapidly at smaller dB values. In addition, there is often some deviation from the ideal curve. From the equations given in Experiment 2.9 it follows that a change in attenuation by 6 dB doubles the pulse width. Thus, after finding the 180° pulse at 62 dB, the 90° pulse is expected at 68 dB and the 360° pulse at 56 dB.