The 13C experiment without decoupling of attached protons is quite similar to the standard 1H experiment. A short radio-frequency pulse centered on the Larmor precession frequency of 13C nuclei in the static magnetic field created by the cryomagnet causes a net transverse magnetization of the carbon nuclei. This transverse magnetization is monitored as the sample decays exponentially towards its equilibrium state, providing a Free Induction Decay, or FID. The FID contains all of the spectral information in the time domain. Subjecting the FID to a fast Fourier transform produces the spectrum in the frequency domain. Before it can be read, the spectrum must be phased. Subsequent processing allows for integration and peak picking.
Note that both 13C and 1H are spin 1/2 nuclei. We expect to see a signal for each magnetically-unique carbon, split into n+1 peaks, where n is the number of protons attached to the carbon.
Of course, the signal from 13C is much weaker than that of a proton because only 1.1% of all carbons present in a molecule are 13C--the balance are 12C, which does not give an NMR signal. This also explains why we don't see spin-spin coupling from adjacent carbons--the chance of two adjacent carbons both being 13C are 0.012% (pretty small!). Also, the magnetogyric ratio for 13C is only 23% that of the proton. All things considered, this is a difficult and time-consuming experiment to attempt, which explains why we don't do it very often.
Note the triplet at 77ppm due to CDCl3 (solvent) in the spectrum below. This signal is very strong because the sample is mostly CDCl3. The carbon in CDCl3 is split into a triplet because the D attached to it is a spin +1 nucleus: the D can be in any one of three spin states (-1, 0 or +1). Thus, any given carbon in the solvent "sees" one of three distinct environments, depending upon the spin state of its attached D.
Spectrum of Ethyl Benzene:
(Click on the spectrum to obtain a magnified view.)