I tested the FWHM of a BNC PB-5 pulser signal as a function of shaping time using the first produced ISOL FEE64 card in Daresbury Labs.

Find attached four representative spectra per ASIC taken with 8, 5 and 0.5 us shaping time values. Run times are different.

Also find attached a plot of the FWHM as a function of the shaping time for the first channel in each ASIC. There appears to be a maximum of the FWHM at around 4 us. This may indicate the main source of noise is external, and at ~4 us the band pass lets most of the noise in.

Unlike the FEE cards employed in RIKEN, this card is grounded to the metal base plate through the screws. In order to investigate the effect of this ground on the noise level of the ADCs in the card, the card was disconnected from the metal plate base. An additional layer of insuating material was placed between the base and the FEE64 card. Nylon screws rather than metal screws were employed to affix the card to the base, and the same tests as before were repeated. See plot below. Overall, the same trend is recovered but suprisingly the FWHM is larger than before for ASICs #1,2 and #3. ASIC #4 has a lower FWHM than before. The reason for the change is not obvious, and may need further investigation.

Carlo

BNC PB-5 parameters

Delay 250 ns

Rate 100 Hz

Polarity POS

Pulse Top TAIL

Amplitude 0.5 V