The cold electronics team at Brookhaven (BNL) received the first prototype (P1) cold ADC ASIC in mid-January and tested a preliminary sample at both room and liquid nitrogen temperatures.
“Most new features are now verified,” said Matthew Worcester of BNL. “In particular, the problems with the previous version (V*) that limited the dynamic range of the ADC at cryogenic temperature are gone.”
The differential non-linearity (DNL) from the “stuck codes” — seen in the 35-ton and V* ADC ASICs — is reduced by half in the P1 version, down to a rate of about 10% averaged over the full dynamic range of all of its 16 channels. Collaborators from BNL and Fermilab met early last December to discuss analysis of the ADC data, focusing on how to identify modes of operating the P1 ADC that can further improve the non-linearity. The team has so far found that running the P1 ADC with an external clock and at a 1 MHz sampling frequency can lower the DNL stuck code rate below 10%. They are still studying the impact of this on physics data.
Teams at both labs are testing several prototype Warm Interface Boards (WIBs) that will be used in both the ProtoDUNE-SP and SBND detectors. These boards are designed to collect the data from the 128-channel cold Front-End Motherboards (FEMBs), which are mounted on the APA in the liquid argon, and transmit these APA data to the DAQ system.
At Fermilab, Theresa Shaw (for ProtoDUNE-SP) and Michelle Stancari and Linda Bagby (for SBND) lead a joint effort to study the impact of the grounding and shielding design on the APA readout. They will test the WIB readout in a Faraday screen-room in the D-Zero Assembly Building.
The BNL team has constructed a noise-and-integration test stand with a 1-by-5 square meter “40% APA” prototype. So far they have inserted it into a shielded, cryogenic box designed to submerge part of the APA and the FEMBs in liquid nitrogen. One FEMB has been attached to the APA and read out through a full prototype cold electronics (CE) readout system, including the WIB. Further, physicists from Stony Brook, Fermilab, and BNL have demonstrated that a pulse could be injected on a wire at the far end of the APA and observed in the WIB’s real-time waveform diagnostic readout.
“We are now going to instrument it with up to four FEMBs and read it out with the full prototype readout chain to study the noise of the integrated CE and APA system,” said Worcester. “Each FEMB will be attached to 128 APA wires taking data simultaneously.” The goal of the BNL test stand is to probe for integrated system noise not present in standalone tests of the CE components, especially when operating multiple FEMBs in cryogenic liquid.