Low power HV supply. Base for 10 dynodes PMT In single photon count mode. Reference: P.F.

Transcription

1 Low power HV supply Base for 10 dynodes PMT In single photon count mode Reference: 24-Jun

2 The product 24-Jun

3 PMT proposed form factor <105 dimensions in mm PCB space 24-Jun

4 HV generation Resistor base k d 1 d 2 d 9 d 10 a - High power -Temp. Sensitive - Volume U hv Cockroft Walton base k d 1 d 2 d 9 d 10 a +Low power -Fixed voltage ratio C b Cp U hv ~ 24-Jun

5 Specifications HV circuit, for 10 dynodes PMT Low ripple 706 mv < T( C) < % 8.5 < T( C) < 22.1 Voltage stable at 3 V < Vinput < 3.6 V Vripple < 500 mv/dynode Low power Vinput 3.3 V ma 3v3 at V < 3 mw (commercial available 50 mw) Low RFI dv/dt < 75 mv/µs, RFI khz-10 MHz Adjustable HV cathode voltage -700 _ V Small foot print PCB = 38_42 mmø (depends on PMT layout) Reliability C, stationary use, after tests 24-Jun

6 PMT base block diagram PMT-base 3 inch PMTs Photons I 2 C LVDS Analog Buffer Comparator I 2 C Decoder + ID Control ID (OTP) Preamp Threshold DAC HV DAC 6.5us Pulse generator Current controlled Oscillator Opamp Driver Current Sense Switch + Sense Circuit Cockroft Walton HV Clock enable 3V3 supply Clock Discrete Components Feedback ETEL bases with the chips PROMiS_v2 CoCo_v2 D. Gajanana et al. ET 24-Jun

7 PROMiS_v2 and CoCo_v2 ASICs D. Gajanana et al. ET 24-Jun

8 Silicon Die size : 1.7 mm 1.25mm Packaged in QFN16 Package body size : 4mm 4mm. Lead pitch : 0.65 mm Package height : 0.9 mm D. Gajanana et al. ET 24-Jun

9 CoCo Chip Silicon die size : 1mm 1mm Packaged in DFN8 Package body size : 3mm 3mm. Lead pitch : 0.65 mm Package height : 0.9 mm CF GND CSNS FB_IN SW GND OPAMP OUT VDD D. Gajanana et al. ET 24-Jun

10 Testing ASICs All 170 packaged PROMiS devices were functional. Tested using sockets 166 out of 168 packaged CoCo devices were functional. Tested using sockets. D. Gajanana et al. ET 24-Jun

11 ASICs: Conclusions and Future MPW were successful proof of concepts, performance and the technology. Packaging of chips was successful with satisfactory yield both in silicon and packaging. Proven to comply with full specifications (Electrical, performance, operational & environmental conditions and mechanical specifications) All functional and characterization tests will be performed at the test house with automatic test-setups Lifetime and long-term reliability tests are foreseen at the testhouse. 6 months after the start-date of the production process, the packaged and fully tested parts are expected. D. Gajanana et al. ET 24-Jun

12 PMT-base next steps PCB layout / PMT type Test devices Acceptance test Cleaning, coating device Production External? Internal? Institute Assembling with PMT Soldering Testing Dy4 Dy3 Dy2 Dy1 nc K Dy5 P Anode Dy6 Dy10 Dy7 Dy8 Dy9 Base for ETEL model used in Antares DOM and PPM-DU 24-Jun

13 Layout differences 7,d2 9,d3 10,d4 12,d5 13,d6 15,d7 16,d8 7,d5 6,d4 5,d3 4,d2 8,d6 9,d7 10,d8 5,d4 7,d6 6,d5 9,d8 10,d9 11,d10 8,d7 12,a 14,da 6,d1 4,f2 2,f1 1,k ETEL D783L D782L 21,a 18,d9 19,d10 1,d1 18,k 16,a 17,d10 Hamamatsu R6233 R ,d9 4,g 3,d3 2,d2 1,d1 19,k HZC 16,g P. F.Timmer K da g(f1) d1 d2 d3 d4 d5 d6 d7 d8 d9 d10 A

14 Status PMT Bases The last base design for the ETL PMT-tube has been successfully tested. From this design 10 pcb s were made. Some of these new bases will already be used in the PPM-DU. For the Hamamatsu PMT-tube a new design is ready. For this design 10 bases are now being manufactured. We expect that they can be tested in a few weeks. For the HZC PMT-tube a study for the base has started. As the layout of the pins is not fixed the design cannot be finished yet. The design for the HZC tube is the most complex. The connections of the dynodes are not subsequent and the diameter is smaller because the HZC tube is longer. 24-Jun

15 P. F.Timmer Test procedures Validation tests Storage; Tamb->30min-10 C ->1h->60 C ->1h->60 C ->-10 C ->30min->Tamb Temp. Shock; Tamb->Tamb+50 C ->1h->Tamb+50 C ->0sec->Tamb Influence oscillator frequency on PMT output Acceptance tests (meet specifications) Voltage adjustment Current measurements COCO frequency and pulse width PROMIS output HV control and I2C Cleaning + Drying Ultrasone Coating Polyurethane 15

16 Mechanical overview of tester FPGA board with USB interface to PC HV measurement unit (Mezzanine) Mother board PMT base to be tested PMT Base adapter set with contact pins P. F.Timmer 16

17 Test setup Safety system Voltage Adjustment and Current measurement Digital value PROMIS outputs Control Digital values frequentie ad pulsewidth Boarderichn Digital value for charge adjust Digital value HV control PMT Tail Adapter Test PROMIS outputs Discharge capacitors from CW high voltage circuit PMT base adapter PMT base CoCo test frequentie and pulse width o CoCo Pr mi PROMIS om is CW High Voltage circuit Electronics for charge injection PROMIS Test HV Control And I2C HV barrier HV measurement supply unit 1 HV barrier HV measurement supply unit 14 P. F.Timmer Digital value of inter dynode voltages 17

18 PCB layout of HV measurement units P. F.Timmer 18

19 Reliability 45 C : 1140 FIT Stress factors; Stationairy use: C: % 12% 9% 9% 4% 22% 15% 9% components design manufactering system management ware out induced no defects software Tests evaluation 2% 5% 5% 5% 0% Evaluated value: 250 FIT 5% 5% For entities with mission time of 15 year, 3 % failures by Poisson. 73% 24-Jun

20 Reliability: some experiences of bases in PPM DOMs On these bases the PROMiS-V1 was directly bonded on the PCB. Several bases had problems with the PROMiS-V1. After testing these PMT modules were cleaned and coated. Cleaning was done in an ultrasonic bath. Unfortunately some modules failed when tested again after cleaning. The ultrasonic cleaning caused failure in the bond wires from several PROMiS-V1 chips. The PROMiS-V2 will be packaged and assembled together with the other components. Because of ultrasonic cleaning also on a few bases the clip of the transformer was unlocked. For the transformer another type of clip will be used. This clip has ground pins and will be soldered on the base. For the new transformer errors because of cleaning will not occur. There were two bases which had a bad assembled transformer. 24-Jun

21 P. F.Timmer Cost estimates HV-base, PMT-HV-v8 total 31 /OM base Euro No. / p Euro tot. Source /PT Coco_v Promis_v Nfet PMF370XN Farnell 150V CMLD4448 TR diode Alcom 22nF V X7R Texim 47nF V X7R Texim 100nF V X7R Texim 10pF % 16V NPO Farnell 220pF % 16V NPO Farnell 100pF V X5R KEMET 100pF V X7R Farnell 1uF V X5R Farnell 10uF V X5R Farnell Core EP-CORE-SET EP7-3E Farnell Bobbin CSHS-EP7-1S-6P-Z Farnell Clip CLI/P-EP Farnell transformer assembly costs assembly firm inductor 22uH ELJPA220KF Farnell BEAD Ferrite, SMD 600 Ohm, Farnell Resistor Farnell Resistor Farnell Resistor Farnell Resistor 1k Farnell Resistor 22k Farnell Resistor 1M % 100 PPM Farnell HV resistor 3G Rhopoint Samtec SEM connector calculation 350nm, AMS PCB costs mm², 6 layer, incl. kapton tail, proto's ca. 20Euro /pc Assembly 0.00 guess total parts 21

22 Questions P. F.Timmer 22

23 HV-protection P. F.Timmer Protect neighbour HV from tails neighbour from HV bases 23

24 RoHS directive Restriction of the use of certain Hazardous Substances in electronic equipment. Consequences; Higher solder temperatures (up to 260 C) Favorite solder: SnAgCu alloy (217 C) Chance for more excessive tin whisker growth Tin whiskers, length up to 10mm, diameter up to 10µm. Optimal growth at 50 C Ref: electrical short circuits debris, contamination, metal vapor arc, due to HV. Image Courtesy of T. Riccio (STPNOC) P. F.Timmer Conformal coating like Parylene C tends to buckle the whisker. 24

25 Hamamatsu PMT with ETL_HV base P. F.Timmer 25