Optical Receiver Sensitivity

Rainer Schwemmer, LHCb Electronics Meeting Dec. 2014

Goal

• Determine the spread of sensitivity of different MINIPod receivers– For the Detector DAQ direction

• Sample of 20 x12 receivers from Marseille – 240 receivers in total – ~ 2-3% of planned installation

2

Methodology

• Unfortunately no x12 tester for sensitivity measurements available

• The receivers need to go onto the new AMC40 prototypes limited time frame

• Fixed transmitter with stable optical power output, change receivers and try to map resulting BER to sensitivity

3

Methodology

4

1.00E-11

1.00E-09

1.00E-07

1.00E-05

1.00E-03

1.00E-01

Channel 1

Channel 2Optical Receive Power

BE

R

• Assumptions:– Curve for different receivers has same shape but is moved to the left

or to the right depending on sensitivity– Transmitter output power is stable between on/off cycles

Methodology

5

1.00E-11

1.00E-09

1.00E-07

1.00E-05

1.00E-03

1.00E-01

Channel 1

Channel 2Optical Receive Power

BE

R

• Inject optical signal at fixed power• Measure BER of different receivers at this power

Fixed Optical Powerat receiver input

Measured BER spread

Sens.difference

Methodology

6

1.00E-11

1.00E-09

1.00E-07

1.00E-05

1.00E-03

1.00E-01

Channel 1

Channel 2Optical Receive Power

BE

R

• If the two curves have the same shape then the two triangles are identical

• Use the BER difference measured on a reference receiver to get the power/sensitivity value of the device under test

Fixed Optical Powerat receiver input

Measured BER spread

Sens.difference

Setup

• One x12 Transmitter of one prototype AMC40 as source– Attenuated to intensity which actually produces errors

• Swap out receivers on the board and measure BER

7

Source Receiver

AMC40 FrontplateConnectors

Attenuator

Prizm Connector Prizm Connector

Setup

• One x12 Transmitter of one prototype AMC40 as source– Attenuated to intensity which actually produces errors

• Swap out receivers on the board and measure BER• Have to power cycle the board for swapping out

receivers though

8

Source

Receiver 1AMC40 Frontplate

Connectors

Attenuator

Prizm Connector Prizm Connector

Receiver 2

Setup

9

Receiver

Transmitter

Attenuator

Results

• Temperature dependency

10

10 min

Results

• Bit error rate vs Channel ID• Individual channels form clusters

– Injected optical power is different on different channels– Attenuator is not same on every channel

11

Results

• Every now and then: put the same receiver back into the socket• Variation is quite large• Prism connector seems to have quite a large spread on re-connects

12

Results

• After conversion of BER into optical power• Distribution is centered around average• No absolute values yet, because we have not measured absolute value of

optical output power13

Conclusions

• Spread of sensitivity is quite large (+- 1 dB)• A large portion of it is probably due to losses

in connector rather than receiver• Need to do calibration of transmitter to get

absolute sensitivity values (2-3 days of measurements in optics lab)

14