Detector Powering in the 21st Century: Why stay stuck with the Good old 20th Century methods?

11 Jun 2011, 16:20
20m
Chicago Ballroom 8 (Sheraton Hotel)

Chicago Ballroom 8

Sheraton Hotel

Oral Presentation Experimental Detector Systems Experimental Detector Systems

Speaker

Dr Satish Dhawan (Yale University)

Description

Future detectors are envisioned with large granularity but we have a power delivery problem unless we fill the detector volume with copper conductors. LHC detector electronics is powered by transporting direct current over distances of 30 to 150 meters. This is how Thomas Alva Edison lighted his light bulb. For example, CMS ECAL uses 50 kiloamps at 2.5 volts, supplied over a cable with a transmission efficiency of only 40%. The transmission loss becomes waste heat in the detector that has to be removed. We have been exploring methods to transmit the DC power at higher voltage (low current), reducing to the final low voltage (high current) using DC-DC converters. These converters must operate in high magnetic fields and high radiation levels. This requires rad hard components and non-magnetic (air core) inductors. We have found that in CMOS circuits, the thickness for the gate oxide is the necessary condition for the radiation hardness. Some of the commercial DC-DC converters are radiation tolerant. A 5 nm thick gate oxide can operate at 3 volts; this can be increased to 15 volts by LDMOS process. For the past couple of years, we have been evaluating Gallium Nitride devices and these are very radiation hard. Converters in this technology are capable of higher conversion efficiency. There is work going on to produce higher speed drivers for the GaN power stage. An efficiency of higher then 90% is achievable for 48 v to 1 v converters. We have also developed spiral air core inductors that have been tested to 7T fields. There is much work needed to study trade off between efficiency, frequency and mass.

Author

Dr Satish Dhawan (Yale University)

Presentation materials