The sources of dead time in a time spectrometer employing a TAC and MCA are easily identifiable, although the derivation of the throughput equations is somewhat more complicated. The time-to-amplitude converter is only able to process one pair of start and stop pulses in each conversion. Once a start pulse has been accepted all further start pulses are ignored until the conversion and reset processes are finished. Similarly, the TAC responds to the first stop pulse that arrives after the accepted start pulse, and ignores all subsequent stop pulses until the next valid start pulse has been accepted. As a result, subsequent start pulses find the start input to be dead from the time of acceptance of the last valid start pulse until the end of the TAC reset. Additional stop pulses find the stop input to be dead from the time the first stop pulse is accepted (following a valid start pulse) until the time of acceptance of the next start pulse.

If the multichannel analyzer dead time is longer than the TAC dead time, the MCA can also contribute to the dead time losses, because the MCA will not always be ready to accept the next TAC output. Choosing an MCA conversion time that is less than the minimum TAC dead time eliminates the MCA dead time contribution. If the MCA dead time is longer than the TAC dead time, one can gate off the TAC start input with the MCA busy signal in order to use the throughput equations developed below.
The following throughput equations relate the time spectrum viewed by the detector to the spectrum actually recorded by the TAC and MCA. They can be used for three purposes: a) to predict the distortions caused by dead time losses, b) to determine the counting rate limits that render the distortion negligible or, c) to implement dead time correction algorithms that permit data acquisition at higher counting rates. The four most common cases are summarized below.

Case 1: Periodic Start and Random Stops, T_s > T_d

Case 2: Random Start and Periodic Stop, T_s > T_d

Case 3: Random Start and Periodic Stop, T_s < T_d

Case 4: Random Starts and Random Stops