Shielding and Grounding

Improper shielding and grounding of system components is the most frequent reason for poor time resolution, poor efficiency and poor differential linearity in TCSPC systems. 

RF noise piekup from radio and television transmitters in the deteetor and reference lines reduees the signal-to-noise ratio of the signals. The result is poor timing aeeuraey, i.e. broadening of the IRF of the TCSPC system. Moreover, noise ean make it impossible to use a sufficiently low CFD threshold to record all single-photon pulses within the pulse distribution of the deteetor. The result is a loss in eounting effieieney. 

RF piekup from the exeitation source or crosstalk between the deteetor and the referenee lines eauses systematie timing errors. The result is a modulation of the effeetive CFD threshold and the effective CFD zero cross point. Modulation of the threshold modulates the deteetion efficiency. Modulation of the zero cross point warps the time axis. In either ease the result is a ripple in the recorded curves. 

Improper grounding ean even transform the line frequeney into the input signals. Beeause the line frequeney is usually not synehronous with the signal repetition rate, the apparent effeet for long acquisition times is the same as for RF piekup. The differenee shows up if sequences of curves faster than the line fre- 

Like all high-speed electronics, TCSPC uses impedance-matched cables for the signal eonneetions. These connections are relatively immune from capacitive noise piekup. Therefore, inductive coupling is the dominating effect that introduces noise into the system. 

---

For example, one needs shielded and grounded cables, for such tiny currents may be exceeded by stray currents from nearby electrical machinery. 

The power obtained for the cells had induced a considerable problem with 60-cycle pickup in the measurement system. Shielding and grounding at various critical positions were necessary to eliminate this problem. 

Another major source of noise at low frequencies is mains pickup. This may amount to hundreds of millivolts superimposed on the test signal unless major efforts are made at shielding and ground loop suppression. Usually, unless an adequate notch filter is used in addition, the experimentalist must be satisfied with reduced precision at frequencies below about 100 Hz. 

Shield and ground all low voltage DC (signal) wires and cable to prevent noise interference. 

---