DUNE Cold Cable Status J. Kierstead Brookhaven National Laboratory - PowerPoint PPT Presentation

DUNE Cold Cable Status J. Kierstead Brookhaven National Laboratory

Cold Cable DUNE Specification • Cold cable development for DUNE has been Specification Matrix progressing with SBND in parallel for the past few months • Shown is a draft specification matrix for DUNE • For SBND the maximum cable length is about 7 meters which allowed a cable gauge of 30 in a 3M cable with a miniSAS connector • In DUNE the maximum length is 18-25 meters which requires a different solution • Some of the companies which have provided cable, connectors or other information to date are: • 3M • Gore • Amphenol • Glenair • Hitachi • Huber+Suhner • Samtec • ANTA J. Kierstead Brookhaven National October 19 2015 2 Laboratory

Cold Cable DUNE Technical Challenges LENGTH(Signal Degradation) Limited Feedthrough Area • The DUNE twinax data cable will be long • The other challenging part of the cold cable (18meters-25meters) resulting in significant is the limited board area. The baseline at this attenuation and jitter. time is an area of 21mm x 22mm for a • The nominal signal attenuation is considered to header for a 12 twinax pair cable. be about a factor of 3 (~9.5 dB) but this is not • There are 2 connectors that have been the sole variable that must be met. considered for the board that would fit in • To meet this requirement an appropriate this area (but not necessarily the footprint). conductor gauge must be used. In this case a 24 They are the miniSAS connector ( candidate gauge is the smallest conductor size that has for SBND by Molex and others) and the been found to work in a 25 meter length at HSEC8 Samtec connector. room temperature. • However neither of these connectors can • The cable must be perform at both room directly handle 24 gauge conductors. temperature and LAr temperature. Many cables qualify in 25 meter lengths at cryogenic • This implies that an adapter board on the temperature (because of the decrease in cable to the board might have to be used. resistance) and fail at room temperature. • At this point we have not actively pursued an adapter We have discussed this possibility with some vendors. J. Kierstead Brookhaven National October 19 2015 3 Laboratory

Cold Cable DUNE Samples, Candidate • • The data cable for dune has a baseline design Gore Cable Layout which consists of 12 twin-axial pairs. Included with each twinax is a shield which is terminated in a drain wire . If each of these is connected separately that would result in a total of 36 conductors in the board connector. • Several different manufacturers (3M, Amphenol, Gore, Hitachi) provided 25 meter or more of twinaxial cable for testing. All of these samples were tested both at RT and LN 2 temperature. • Shown is a typical layout for a single twinax pair. In this case it is a layout for a Gore 24 gauge twinax cable. • Differential attenuation for 17 meters as measured by Gore is 12.5 dB at 1 GHz. • None of the cable samples except for the Gore cable met the requirements for being measurable at a length of 25 meters at room temperature J. Kierstead Brookhaven National October 19 2015 4 Laboratory

Cold Cable DUNE Measurements Gore cable • The best twinax cable (in a 24 gauge conductor )found to date is from W. L. Gore and Associates • Shown are the eye diagrams for a 25m length of Gore cable at room temperature and at LN temperature. Note that the scales are not the same for the 2 plots. • As stated before, the Gore cable is the only one tested to date that performs adequately in a 25 meter length at both room temperature and at LN 2 temperature. • In a 20m length the eye height drops to 235 mV with • Room Temperature, 25m eye width of ~ 798 pS • Eye Height 150 mV • Eye Width 757 ps • Jitter 243 ps • Room Temperature (0.5 m) • Liquid Nitrogen Temperature, 25m • Eye Height 747mV • Eye Height 484 mV • Eye Width 950 ps • Eye Width 872 ps • Jitter 46 ps • Jitter ps J. Kierstead Brookhaven National October 19 2015 5 Laboratory

Cold Cable DUNE Connectors • • This is an ongoing challenge and HSEC8-DV Without latch it fits within the 21 x 22 mm. there is no clear path yet. • Shown is an HSEC8-DV edge connector with 13 connections per row with and without a latch. To fit the 36 conductors in the cable the next size in the series, with a latch, would be 25 connections/row. • The choice for the SBND is the miniSAS which fits in the footprint and has a latch. • Neither header can be used with the Mini SAS header (Molex) mating connector and requires an adapter • In addition the connector is surface mount (without sufficient mechanical strength) and will require a strain relief structure which is not addressed here. J. Kierstead Brookhaven National October 19 2015 6 Laboratory

Cold Cable DUNE Next Steps • One candidate twinax cable (Gore) with 24 gauge conductors has been identified. Many other cables have been tested without meeting the requirement at 25 meters. Other sample cables are being tried and tested and previous samples will be tested at shorter lengths in the event that the maximum cable length is reduced from 25 meters. • A commercial connector/header for the feed-through end of the cable has not yet been found which will fit within the available area and terminate 24 gauge conductors. It is likely that a custom adapter will be required for this. J. Kierstead Brookhaven National October 19 2015 7 Laboratory

Cold Cable DUNE Status extra slide J. Kierstead Brookhaven National October 19 2015 8 Laboratory

Cold Cable DUNE Setup, Samples and Terminations 25 meter test samples Terminations Eye Diagram Test Setup J. Kierstead Brookhaven National October 19 2015 9 Laboratory