the DC B-field voltage is 3.100V. i found that i have have to tune the voltage slowly, to let the CRO signal stabilized.

The NMR setting is 12.8MHz, Level 200, pule length is 5uS on +Y axis. FWD is 49W, BWD is 7W. Gain is 35db, LPF is 100kHz and polarization is 90 degree.

the FID signal is about 200mV peak to peak.

Let me briefly explain the NMR circuit, for reader who want to know the theory. HERE.

Black arrows are any cable, Orange arrows are 12.8MHz cables. The direction of the arrow shows the signal flow.

the NMR pulse is a mixture of 2 signal. the RF generator provides the larmor frequency. and the pulse generator provides a square wave, the duration of the square determine the pulse time, it should be matched the requirement for π/2 pulse. and the period of the pulse should be longer than the T2, longitudinal relaxation time. since we don’t want to either distorted the RF or use 1 more RF generator, the splitter simply split the signal, 1 be a main signal and the other one be a reference signal.

The purple mixer or Double Balanced Mixer (DBM). the function of it is multiple the 2 input signal together. the circuit inside, may be i study later.

after the Mixer, the signal will be amplified and pass a λ/4 cable. i am not sure is it necessary to use a λ/4 cable in here. but anyway, we knew that the phase different from the Amplifier to the crossed-diode is λ/4.

the crossed diode has a feature that only allow signal, which larger then 0.6 V to pass. when the pulse come in, it will not see the diode. however, when the in put signal are off, the signal from the NMR coil will be generated and this signal is smaller then 0.6V, thus, the crossed diode act as a switch or duplexer.

the λ/4 cable from the crossed diode and the NMR tuner + Coil is for impedance matching. even there are little bit mismatch of the tuner, the λ/4 cable is still approximately equal to the impedance matching case. that smoothen the signal.

and i don’t know is it necessary to use  a λ/4 cable from the cross diode to the amplifier.

after the amplifier increase the signal and mixed with the reference signal at the differential mixer. The Free induction decay (FID) signal should be at the Larmor frequency. However, if the input frequency is not exactly, with just little bit of mismatch, the spin can still be rotated. by different the NMR signal with the reference signal ( which is the signal we put in the NMR coil) we can tell the different and calculate back the true Larmor frequency by Fourier transform.

and also, by different the NMR signal by the reference signal, we can more easy to tell the different then by just looking at the NMR signal. since the different usually just few kHz, compare with the Larmor frequency 12.8MHz, less then 1 % deferent.

the different mixer also decompose the signal into sine and cosine waveform. this is because, by only look at the cosine wave, we cannot know the different of the signal is positive of negative. the cos wave is natural real part of the complex wave.

the low pass filter (LPF) can be set the maximum frequency pass. this is for seeing the decay envelop of the FID. anther function is reduce the high frequency noise.