[Halld-pid] ST First Tests

[Werner Boeglin]

Hi Fernando,

thank you for the interesting data. It looks to me that the rise time is about the same as what we got with the SiPM array that we currently have. I think I would tend toward the 50um devices as current estimates show that the particle rates in each paddle can get quite high (at 10^8 photons/s we expect about 300KHz of electromagnetic background rate and the hadronic rate is about the same or according to current simulations even higher) The hadronic rate estimates still seem  to have some problems that need to be resolved.

kind regards

Werner



 
On Jul 16, 2012, at 10:34 PM, Fernando J Barbosa wrote:

> Hi Mark, Werner,
> 
> I have been optimizing and testing the ST readout for the 50 um and 100 um SiPMs. There are 4 SiPMs connected together (current sum) to one readout channel on each of two SiPM PCBs (50 um and 100 um). The sensor capacitance is the same in both cases so one optimization is all that is required for fast risetimes for the group of 4 SiPMs.
> 
> The prototype is fully functional but I determined that the micro-coax ribbon cable assembly (connects the readout to the breakout PCB) is not good enough due to cross-talk (primarily due to the connectors on these cable assemblies - a single micro-coax cable assembly provides the power to the electronics, 3 bias supplies, 6 outputs [3 channels w/ 3 for ADCs and 3 for TDCs] and has a thermocouple two-wire output for readout). The readout is temperature compensated via thermistors. The workaround has two coax cables (ADC and TDC), in addition to the micro-coax, which bypass the micro-coax signal connections. So, the final design will change but the current prototype will be useful for tests with ST detector prototype.
> 
> I have attached two pics: ST50 is for the 50 um and ST100 is for the 100 um. These were taken with a laser and a diffuser stimulating all 4 SiPMs in the group. Some comments:
> 
> 1. 50 um - Fast pulses with width of 37 ns (~10%-10%)
> 2. 100 um - Fast pulses with width of 125 ns (~10%-10%), step/kink ~ 5 ns after peak.
> 
> The leading edge slew rate is about the same for both but the 100 um has a slower recovery w/a kink (the quench resistor on each pixel would need to be ~3x smaller for similar recovery but there are other factors at play, i.e. Geiger current) . I will investigate this a little further but these are definitely characteristics of these SiPMs and I will write a GlueX note soon.
> 
> In the meantime, let me know if there is interest in pursuing the 100 um further. Even though the rates are expected to be in the KHz range, the long recovery time will likely lead to pileup.
> 
> Best regards,
> Fernando
> <ST50.png><sT100.png>