BNC PB-4 - see attachment 1

Amplitude 35,000
Attenuators all OUT
Risetime 50ns
Decay time 1ms
Tail pulse
Polarity +
Frequency 2Hz (min)
Delay min

DSO ch#1 DC/1M - y: 5mV/div x: 100ns, 200ns, 1us, 2us & 10us/div  - attachments 2-6



BNC PB-5 - see attachment 7-9

Amplitude 1.0V
Attenuation x1
Decay time 1ms
Tail pulse
Polarity +
Frequency 2Hz (min)
Delay 250ns (min)

DSO ch#1 DC/1M - y: 5mV/div x: 100ns, 200ns, 1us, 2us & 10us/div  - attachments 10-14



BNC PB-5 - see attachment 15-16

Amplitude 10.0V
Attenuation x10
Decay time 1ms
Tail pulse
Polarity +
Frequency 2Hz (min)
Delay 250ns (min)

DSO ch#1 DC/1M - y: 5mV/div x: 100ns, 200ns, 1us, 2us & 10us/div  - attachments 17-21



Conclusion 

BNC PB-5 significantly noisier than BNC PB-4. Results consistent with GSI tests https://elog.ph.ed.ac.uk/AIDA/886 , i.e. AIDA BNC PB-5 pulser at GSI *not* faulty.
BNC PB-5 operation with 10.0V/10x ( cf. 1.0V/x1 ) shows lower level of noise but still worse than BNC PB-4.

5mV transient at test input ( C_t = 0.7pF ) = 3.5fC = 80keV Si equivalent ( check 1V x 0.7pF = 0.7pC = 16MeV OK ).