4 Jun 2008 Paul Dauncey 1

Crosstalk and laser results

Paul Dauncey, Anne-Marie Magnan, Matt Noy

4 Jun 2008 Paul Dauncey 2

• Trying to find a repeatable pattern…• Study Sensor #1, pixel 61, 28; mask quads 1-3, 10 bunch crossings/bunch train

Crosstalk

Column 61

unmaskedColumns 61-65

unmaskedColumns 61-66

unmasked

Column 61

unmasked

Columns 57-61

unmasked

Columns 56-61

unmasked

4 Jun 2008 Paul Dauncey 3

• Check if in analogue or digital part of circuit• Move pixel 63,129 up by trim=10 to move away from other pedestals

Crosstalk source

• Effect is not quite a step function

• Pedestal position and width are unaffected by extra noise from other pixels

• Crosstalk appears not to affect analogue circuit

Trim=0

Column 63

unmasked

Trim=10

Columns 63-66

unmasked

Trim=10

Column 63

unmasked

Trim=10

Columns 63-67

unmasked

4 Jun 2008 Paul Dauncey 4

Sensor #1

y = 0.2648x2 + 5.1276x + 5.8424

0

20

40

60

80

100

120

140

160

-1 1 3 5 7 9 11 13 15

Trim setting

Pede

stal

(TU

)

Data

Poly. (Data)

• Can measure pedestal peak accurately with short bunch trains

• Scan trim setting and measure pedestal position for single channel• Pixel 61, 28

Trim units

1 trim unit ~ 6TU

1 trim unit ~ 15TU

4 Jun 2008 Paul Dauncey 5

• Easy to focus optically on substrate surface, but…• Epitaxial layer ~300m below this

• Laser wavelength will have different focal point

Laser focus

0

50

100

150

200

250

-300 -200 -100 0 100 200 300 400

Adjustment from substrate focus

La

se

r s

ign

al s

ize

• Focus on substrate and adjust by focus dial gradations• We think these are in

units of m

• Positive value moves focus down towards epitaxial layer

• Measure laser signal above pedestal at each setting• Work with +60m

4 Jun 2008 Paul Dauncey 6

• Unmask 3x3 array of pixels in bulk• Take a threshold scan with laser disabled

• Take threshold scans at each of the 21 “Giulio” simulation points

• Use upper edge of threshold scan to define signal• Analogue measurement of signal size in 3x3 array

Charge sharing

Pixel centre

Pixel corner

5m

4 Jun 2008 Paul Dauncey 7

PedestalsLaser disabled

4 Jun 2008 Paul Dauncey 8

Laser signals

Laser in corner

of four pixels

= point 20

4 Jun 2008 Paul Dauncey 9

Fitting for upper edge

Laser in corner

of four pixels

= point 20 Count events with hit,

not number of hits

Fit to rising and falling erfs

4 Jun 2008 Paul Dauncey 10

Expected total sum of signalDiode

4 Jun 2008 Paul Dauncey 11

Measured total sum of signal

With deep p-well Without deep p-well

Total simulation signal ~ 1300e−

Guess “total” signal ~ 530TU

0

50

100

150

200

250

300

350

400

450

500

0 2 4 6 8 10 12 14 16 18 20

Position

To

tal

lase

r si

gn

al (

TU

)

0

50

100

150

200

250

300

350

400

450

500

0 2 4 6 8 10 12 14 16 18 20

Position

To

tal

lase

r si

gn

al (

TU

)

4 Jun 2008 Paul Dauncey 12

Deep p-well signal per pixel

Giulio simulation Data

Cell numbering

4 Jun 2008 Paul Dauncey 13

Non-deep p-well signal per pixel

Giulio simulation Data

Cell numbering

4 Jun 2008 Paul Dauncey 14

Non-deep p-well signal per pixel (cont)

Diffusion simulation Data

Cell numbering

4 Jun 2008 Paul Dauncey 15

• Crosstalk is pickup from other pixels firing

• Seems to be quite pixel dependent

• No obvious pattern seen yet

• No effect on analogue circuit performance

• Trim is not exactly linear

• Trim unit in TU depends on trim setting

• Laser focus makes a big difference to signal size

• Charge spread can be measured in the bulk

• The spread is qualitatively similar to the simulation…

• …but does not agree quantitatively

• Must check laser uniformity in time and alignment

• Not yet at precision level

Conclusions