HAWAII-1RG multiplexer readout

Transcription

1 HAWAII-RG multiplexer readout Preliminary results

2 HAWAI -RG data acquisition prototype To HRG bias Digital control Digital board Analog board Sf3 airtight Fisher connector (HRG pixel output) 2

3 Analog Electronic bloc diagram From HRG SF3 Fisher output mode Out7 2 output mode Out5 3 µa 3 µa Voltage reference - + AD797 AD Diff. input Diff. input Internal Gain=9.5 ADC 6 bits SPT8 Internal Gain=9.5 ADC 6 bits SPT8 clk (up to 5 MHz) Oe_out7 ADC data (6 bits) Oe_out5 To Digital board 3 µa AD797 Internal Gain=9.5 Ref Pixel - + Diff. input ADC 6 bits SPT8 Oe_ref 3

4 Digital electronic bloc diagram To HRG Digital control Connector To HRG Digital control Connector 2 Power supplies From analog board ADC data 6 ADC clk OE 3 FPGA ALTERA CYCLONE Bias generators FIFO RefSample RefColBuf Vreset Dsub Data bus (6 bits) Slow control Temp/2 DAC DAC DAC DAC ADC FPGA ALTERA CYCLONE 28K 32bits FIFO Mezzanine board 6Mo SDRAM Memory ETRAX Processor Linux MHz 4Mo Flash Memory RJ45 Ethernet Bias connector To HRG 4

5 Electronic KHz Analog input characteristics (@ KHz) ADC quantum Vref = 2.5 V q = 76.3µV Input range < V detector < 263 mv (@ gain=9.5) Equivalent noise ADC only lsb (G=9.5, q d ~ 8µV) (ADC only + digital average) => ~ 2.2 e - Equivalent noise with HRG cables (ADC + cables + HIROSE and HRG outputs simulated by voltage sources + digital average) (detector capacitance 4fF) 5 lsb (G=9.5) => ~ e - Equivalent noise with Reference substraction 2.2 lsb (G=9.5) => ~ 5 e - 5

6 Readout Setup Power supplies Vacuum proof Fisher connector HRG digital control HRG bias Ethernet Digital board With embbeded processor Hirose connector HRG MUX Kapton ribbon Interface board Analog and ADC board Analog output Metal box (simplified cryostat model) 6

7 Readout setup 7

8 Test room -Pressurized -Air filtered -ESD protected 8

9 Bias setup Vbiasgate = 2.4 V Vnbias =.833 V Vpbias = 2.32 V Vncasc =.77 V Vpcasc =.966 V Vbias_outbuf =.846 V Vbias_power = 3.29 V Refsample =.97 V RefCollBuf =.2 V Dsub = V Vreset = V Drain = V Celldrain = V T= room temperature (25 C) 9

10 Not default value Register setup HoriDirReg HINVDIR / Register setup HINVDIR HINVDIR2 HINVDIR3 / GainReg GAIN CAPEN VTESTEN HTESTEN OutputModeReg MODE MODE2 MODE6 WMOUTEN * Controlled by modectrl /2 OutputBufReg NormalModeReg SFENA VMODE SFENB HMODE NOFSA VINDIR NOSFB REFPIXEN HIGHOHM VEDGE REFMODE FASTBUF HEDGE REFVOLT NOFBUFA RSTEDGE REFROW / NOFBUFB GLOBAL WindowModeReg VMODWM HMODWM VDIRWM HDIRWM VEDGWM HEDGWM REDGWM GLOBWM MiscReg VWMEN HWMEN VRSTB HRSTB FASTEN SAMPCLK CLKMOD * in Window mode * Controlled by V / HRESETB *FASTENDPAD *SAMPLECLK VertStartReg VertStopReg HoriStartReg VertStopReg PowerDownReg OptionsReg PDREF PDCK PDREF PDCKWM PDREF2 IFCTRL * Set in window mode only * Set in window mode only * Set in window mode only * Set in window mode only PDREF3 DOEN PDBUF PDBUF PDBUF2 PDBUF3

11 Output Mode KHz buffered (B) 3µA OutputBufReg SFENB = Fast horizontal readout bus - + B A - + AD797 CTRL6 CTRL5 CTRL4 CTRL3 CTRL2 CTRL Slow horizontal readout bus Output Source follower DRAIN

12 output KHz Baremux only Out#7 Row profile Level gradient near the borders Mux default? 4 rows and columns of reference pixels 2

13 KHz window mode (size 3x3) Left top corner 3x3 Row profile The first reference pixel is always different From the 3 next ones 4 reference pixels 3

14 output KHz Row 64 Zoom on row 64 4

15 Separated reference KHz Reference pixel is sampled synchronously with the active area pixels Histogramme over 24x24 pixels Noise 4 µv 26 e- V=42 µv Level shift on last column pixels Zoom over 2 columns Zoom Problem of alternating level along the row Odd even effect? 5

16 Separated reference KHz Ref Pixel V ~ 4 µv Pixel change at falling edge HCLK 6

17 Separated reference KHz Ref Pixel V ~ 8 µv Power down level? HCLK Last pixel 7

18 Separated reference KHz window mode (size 24x24) Histogramme over a window of 24x24 pixels The problem of atlernating pixel does not exist in window mode! Noise 4 µv 26 e- (24 e- mux only) The last column problem is still present in window mode 8

19 Separated reference KHz window mode (size 24x24) Ref Pixel HCLK No problem of alternating pixel levels in window mode 9

20 Separated reference KHz window mode (size x) The last column problem is independant of the window size Out #7 window x Row profile Last pixel V=8 µv Histogramme over x window * 6525 V=8 µv Factor of consistent with Last column vs matrix size ratio 2

21 Separated reference KHz window mode (size x) First pixel() Last pixel() ~ no pixel level? No pixel level Ref Pixel HCLK 2

22 Row of reference KHz REFROW= Out#7 + REFROW active This pattern seems consistent with the 4 capicitors repeated along the row. But how it works exactly? Reference pixel and active area pixels seems to be mixed! 22

23 Row of reference KHz REFROW= in window mode (x) 23

24 Technical questions summary Separated reference pixel output Difference of behavior between normal and window mode Odd even effect in normal mode only Last column pixel in all mode? Noise µv ~ 26 e- (4fF pixel) is it correct? Pixel output Gradient near the borders We also noticed a spike at the beginning of each row. It disappears when we wait enought time (> ms) before starting reading pixel. How REFROW should be used? What is the expected output of a baremux (open ended inputs)? 24

25 Scheduled and prospects Test baremux in the K Warm test with Lyon s HRG (mid november) Cold test with Lyon s HRG We propose a new status report of the HRG test end november. Depending on these tests Plans for Michigan/Berkeley H2RG No San Diego available at Lyon (HAWAII) Present system can be duplicated (cost ~ 3 euros) Safety requirements for Michigan and Berkeley to accept shipping H2RG in such new conditions? 25