Transcript 20160229_FOX_Presentation_LArDPS_to_FEX_Modules_Tests.pptx

Characterizing FOX
Demonstrator Test Setups for
LArDPS to FEX Modules at CERN
RuthAnn Gregory
1
Outline
•
•
•
•
•
FOX demonstrator introduction
Review conclusions from tests at MSU
Introduction
Testing the ‘zero’ reference points for insertion loss measurement procedure
Measuring the insertion loss for the systems of optical components to be
used in the L1Calo system:
LArDPS to eFEX test configuration without and with splitters
LArDPS to j/gFEX test configuration without and with splitters
• More Tests:
Fluke meter vs CMX board
• Conclusions
2
FOX Demonstrator
•
•
•
Fiber Optic eXchange (FOX): Refers to an
upgrade for the L1Calo system.
FOX demonstrator was used to test the
optical components
Maximum light loss tolerated by the
optical components is ~ 10 dB
3
Review of tests done at MSU
• Tests were done to better understand the behavior and reproducibility
of the components (MiniPods, cables, attenuators, splitter, etc.) that
the FOX and L1Calo will be using.
• The light output and input measurements show similar reproducibility
and stability under different conditions (different temperatures,
before/after cleaning, etc.)
• Receiver optical power variations:
0.2 dB for disconnecting and reconnecting MTP connectors
0.2 dB for a temperature increase from 36.4 to 38.2°C
0.9 dB from one MiniPod channel to another
.4 dB from disconnecting, cleaning, and reconnecting
4
Review of tests done at MSU
• The attenuator measurements are not very reproducible.
Manufacturer specifications, and measurements done with Fluke and CMX
yield different results
• The splitter behaves as expected:
Measures about 3dB loss for each output
Outputs are split about 50%
• Next Step (in progress): Bit Error Tests
Empirical measurement of the “light power budget”
Characterization of the steepness of that empirical limit.
Synchronization is lost at around 10 dB of attenuation
5
FOX Demonstrator Tests
• Motivation:
To measure the insertion loss of the systems of optical components that
will be used in the L1Calo system.
• Purpose of measurements of test setups for LArDPS to FEX
modules:
To fully characterize the FOX demonstrator.
• Purpose of tests with Fluke meter:
To explore the stability and reproducibility of the system using a different
measuring device.
6
FOX Demonstrator Tests at CERN
• In order to make insertion loss measurements a reference or ‘zero’
offset measurement needs to be measured.
Actual setup for LArDPS to eFEX FOX demonstrator
Diagram for LArDPS to eFEX FOX
demonstrator (details to be explained later)
7
Testing the ‘zero’ reference points for insertion loss
• These references were used in the light loss calculations.
• Two configurations of the ‘zero’ reference points were used.
• A defective fiber was found on one of the breakout cables. This
defect is in the breakout cable and is not part of the FOX.
Configuration 1: Two 12 fiber
breakout cables with direct LC-LC
connections
Configuration 2: Configuration 1
with two additional 48 fiber
breakout cables and one
additional 48 fiber trunk cable.
8
Testing the ‘zero’ reference points for
insertion loss
• Plot shows very little light
loss through two 48-fiber
MTP connectors + one LC
connector
• Sets the scale for the light
loss to expect from two 48
fiber MTP connections.
Difference in Average Light
Input(dB)
• One of the 48 fiber breakout cables used in this setup had bad fiber on a
breakout cable with significantly more light loss than the rest of the fibers
Difference of 'zero' method 1 and method 2 tests
• This fiber is omitted from
0.7
the plot shown here.
0.6
0.5
0.4
0.3
0.2
0.1
0
0
1
2
Blue(fibers 48-37)
Rose(fibers 24-13)
3
4 5 6 7 8 9 10 11
MiniPod Channel
Beige(fibers 36-25)
Aqua(fibers 12-1)
9
LArDPS to eFEX test configuration without
splitters
• The diagram below shows the test configuration used including
specific cables and connectors.
10
LArDPS to eFEX test configuration without
splitters
• The outlier from the bad
fiber was omitted from this
graph.
• This is compatible with and
reinforces the results from
slide 9.
Average Loss(dB)
• A very small amount of light
loss (about 0.2-1.4 dB) was
measured for this setup.
Average loss for LArDPS to eFEX test configuration
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
0
1
2
3
4
5
6
7
8
9 10
MiniPod Channel
Aqua (fibers 12-1)
Rose (fibers 24-13)
Beige (fibers 36-25)
Blue (fibers 48-37)
11
11
LArDPS to eFEX test configuration with splitters
• The configuration with splitters uses a 48 fiber breakout cable with
three built-in passive splitters as part of one group of 12 fibers
• Three built-in passive
splitters plus three
discrete splitters were
used to split the fibers
for one group of 12
fibers in the special 48
fiber breakout cable.
12
LArDPS to eFEX test configuration with splitters
Average Loss (dB)
• Plot shows light loss through LArDPS to eFEX test setup plus loss
through built-in splitters.
Average Loss of through built-in passive splitters in the LArDPS
• Expected loss for
to eFEX test configuration
splitters is about 3dB per
5.5
output
5
• The amount of light loss
was similar between the
4.5
two outputs
4
• Individual discrepancies for 3.5
splitters is expected. Example:
± 0.3 dB corresponds to a
53/47% split.
Black(chan 0, fiber 24) Rose(chan 1, fiber 23) Aqua(chan 2, fiber 22)
output 1
output 2
13
LArDPS to eFEX test configuration with splitters
• Insertion loss from three discrete splitters was measured in different positions.
Insertion Loss (dB)
• First, second and third
group of three
positions were
measured by discrete
splitters 1, 2, and 3,
respectively.
• Discrete splitters
add roughly 3 dB of
light loss.
• Measurements can
vary with position by
up to about 0.3 dB.
Light loss for LArDPS to eFEX test configuration through all discrete
splitters as measured in different positions
4
3.9
3.8
3.7
3.6
3.5
3.4
3.3
3.2
3.1
(chan 3, (chan 6, (chan 9, (chan 4, (chan 7, (chan 10, (chan 5, (chan 8, (chan 11,
fiber 21) fiber 18) fiber 15) fiber 20) fiber 17) fiber 14) fiber 19) fiber 16) fiber 13)
Discrete Splitter 1
output 1
Discrete Splitter 2
output 2
Discrete Splitter 3
Loss without splitters + 3 dB
14
LArDPS to j/gFEX test configuration without
splitters
• This configuration makes use of 72 fiber cables.
15
LArDPS to j/gFEX test configuration without
splitters
• Most fibers in this setup continue to measure very little light loss.
Average Loss for LArDPS to j/gFEX test setup without splitters
Average Loss (dB)
• Some fibers near one
side of the connector
measure a larger
amount of light loss
4
3
2
1
0
0
1
2
3
4
5
6
7
MiniPod Channel
8
9
10
Aqua(fibers 12-1)
Rose(fibers 24-13)
Beige(fibers 36-25)
Blue(fibers 48-37)
11
16
LArDPS to j/gFEX test configuration with splitters
• As before, the 48 fiber breakout cable with built-in passive
splitters and discrete splitters were inserted into the light path.
17
LArDPS to j/gFEX test configuration with
splitters
Average Loss (dB)
• As before, the light loss through the built-in passive splitters was
significantly higher than the light loss without splitters.
Average Loss of through built-in passive splitters in the LArDPS
• The Black and Rose builtto j/gFEX test configuration
in passive splitters
6.5
6
measure similar light loss
5.5
for both outputs.
5
4.5
• Output 2 on the Aqua built4
3.5
in splitter measured about
3
2.5
2dB less light loss than
Black(chan 0, fiber 24) Rose(chan 1, fiber 23) Aqua(chan 2, fiber 22)
output 1 (even though it is
Passive Splitter
the same splitter as on
Output 1
Output 2
18
slide 13)
LArDPS to j/gFEX test configuration with splitters
• The discrete splitters typically add about 3 dB of insertion loss as expected.
Light Loss through Discrete Splitters in the LArDPS to gFEX test
configuration as measured in different positions
5.5
Average Loss (dB)
• Channel 3
shows higher light
loss than the rest of
the channels, which
is attributable to
the higher light loss
measured on
channel 3 without
splitters shown on
slide 16.
5
4.5
4
3.5
3
(chan 3, (chan 6, (chan 9, (chan 4, (chan 7, (chan 10, (chan 5, (chan 8, (chan 11,
fiber 21) fiber 18) fiber 15) fiber 20) fiber 17) fiber 14) fiber 19) fiber 16) fiber 13)
Discrete Splitter 1
Output 1
Output 2
Discrete Splitter 2
Discrete Splitter 3
Loss without discrete splitter + 3 dB
19
Fluke laser light source and light meter
20
Fluke Meter Tests
2.5
Insertion loss for Aqua(1-12) bunch of 12 fibers
without splitting as measured by Fluke and CMX
2
Insertion Loss (dB)
Insertion Loss (dB)
• Fluke meter light loss tests were done for the LArDPS to eFEX
configuration without splitters and for the LArDPS to gFEX
configuration with splitters using the ‘zero’ method 1 as the reference.
• The light loss measurements with the Fluke meter were consistently
higher than the measurements with the CMX module.
1.5
1
0.5
0
Chan Chan Chan Chan Chan Chan Chan Chan Chan Chan Chan Chan
0
1
2
3
4
5
6
7
8
9
10 11
MiniPod Channel
Fluke
CMX(using method 1 as zero reference)
8
7
6
5
4
Insertion loss through output 1 of built-in passive
splitters for the LArDPS to j/gFEX test
configuration as measured by Fluke and CMX
Black chan 0, fiber 24 Rose chan 1, fiber 23 Aqua chan 2, fiber 22
Discrete Splitter 1
Discrete Splitter 2
Fluke
CMX
Discrete Splitter 3
21
Conclusions
• Measurements were made with FOX demonstrator that mimic the
systems of fibers and connectors to be used to map optical signals
in the L1Calo system.
• The insertion loss through the FOX demonstrator was very small,
most fibers had between about .2 and 1.5 dB of insertion loss
(with the exception of two defective fibers).
• The splitters behaved as expected with about 3 dB light loss for
each output.
• The Fluke meter consistently measured about 1 dB more light loss
than the CMX module. (similar to the results of the tests done at
MSU).
22
Backup Slides
23
FOX Demonstrator Tests Notation
• In the 48 and 72 fiber cables, four groups of 12 fibers were used.
• These groups are referred to by color and relative position:
Aqua group of 12 fibers, contains fibers in positions 1-12.
Beige group of 12 fibers, contains fibers in positions 13-24.
Rose group of 12 fibers, contains fibers in positions 25-36.
Blue group of 12 fibers, contain fibers in positions 37-48.
Fiber naming convention example:
First Letter: Indicates type of fiber
(T for Trunk and B for Breakout)
TF48_n1(2)
Second Letter: Indicates
type of connector (M for
Male and F for Female)
Number of fibers
For Trunk cables: number in parenthesis
indicates length of cable in meters.
Fiber identification
number
24
LArDPS to eFEX test configuration without
splitters and ‘zero’ method 1
25
Bit Error Tests
Difference in Just lost synch and Just re-gained synch
Difference in Attenuation (dB)
2.5
2
1.5
1
0.5
0
Chan Chan Chan Chan Chan Chan Chan Chan Chan Chan Chan Chan
0
1
2
3
4
5
6
7
8
9
10
11
Fluke
CMX
27
Discrete Attenuators
Average Attenuation (dBm)
Average Attenuation measured by Fluke meter, CMX software, and Spec Attenuation
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
Average Attenuation measured by Fluke meter
Average Attenuation Measured by CMX software (diff in input without and with attenuators)
Spec Attenuation
25
Disconnecting and Reconnecting MTP Connectors
• Each MTP connector was disconnected and reconnected for two trials
in the LArDPS to eFEX configuration without splitters.
• Two measurements were taken after each disconnection/
reconnection
• There was little change in the average light output after
disconnecting and reconnecting the MTP connectors
28
Disconnecting and Reconnecting MTP Connectors
1.4
1.2
1
0.8
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
-0.8
-1
-1.2
BF12_n2
BM48_n3 TF48_n3(10)B
TF48_n3(10)A
TM48_n1(.65)A
TM48_n1(.65)B
TF48_n2(2)A
TF48_n2(2)B
BM48_n5
BM48_n4
TF48_n1(2)B
Chan 0 RX (dBm)
Chan 1 RX (dBm)
Chan 2 RX (dBm)
Chan 3 RX (dBm)
Chan 4 RX (dBm)
Chan 5 RX (dBm)
Chan 6 RX (dBm)
Chan 7 RX (dBm)
Chan 8 RX (dBm)
Chan 9 RX (dBm)
Chan 10 RX (dBm)
Chan 11 RX (dBm)
TF48_n1(2)A
BM48_n1
2nd removal
1st removal
2nd removal
1st removal
2nd removal
1st removal
2nd removal
1st removal
2nd removal
1st removal
2nd removal
1st removal
2nd removal
1st removal
2nd removal
1st removal
2nd removal
1st removal
2nd removal
1st removal
2nd removal
1st removal
2nd removal
1st removal
BM48_n3
2nd removal
1st removal
-1.4
BF12_n1
29
Variable Attenuators
Average Attenuation (dB)
5.5
5
4.5
4
3.5
3
1
2
3
4
5
6
Variable Attenutor
Fluke
CMX
30