Newport 2107 User manual

USER’S GUIDE
10-MHz Adjustable
Balanced Photoreceivers
Models 2107 & 2117
21x7 10MHz Adj Blncd Rcvr revA.fm Page 1 Friday, January 7, 2005 12:22 PM
These photoreceivers are sensitive to electrostatic
discharges and could be permanently damaged if
subjected even to small discharges. Ground your-
self adequately prior to handling these detectors or
making connections. A ground strap provides the
most effective grounding and minimizes the
likelihood of electrostatic damage
phone: [phone]
e-mail: [e-mail] • www.newport.com

21x7 10MHz Adj Blncd Rcvr revA.fm Page 2 Friday, January 7, 2005 12:22 PM
Warranty
Newport Corporation guarantees its products to be free of defects
for one year from the date of shipment. This is in lieu of all
other guarantees, expressed or implied, and does not cover
incidental or consequential loss. Information in this document is
subject to change without notice.
Copyright 2022, Newport Corporation. All rights reserved.
The New Focus logo and symbol are registered trademarks of
Newport Corporation
Document Number 90099913 Rev. A

Models 2107 & 2117 Contents • 3
Contents
Operation 5
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Using the Photoreceiver. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Checking the Battery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
General Features & Principles 9
Photoreceiver Circuitry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Optical Power and Output Voltage . . . . . . . . . . . . . . . . . . . . . 10
Frequency Response and Noise 13
Measuring Bandwidth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Measuring Noise. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Performance Data for Frequency Response. . . . . . . . . . . . . . 16
Performance Data for Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Common Mode Rejection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Characteristics 21
Physical Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Model 2107 Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Model 2117 Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Customer Service 24
Technical Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
21x7 10MHz Adj Blncd Rcvr revA.fm Page 3 Friday, January 7, 2005 12:22 PM

4 • Contents
21x7 10MHz Adj Blncd Rcvr revA.fm Page 4 Friday, January 7, 2005 12:22 PM
This page has been intentionally left blank

Models 2107 & 2117 Operation • 5
Operation
Introduction
The Model 21X7 is a general-purpose balanced
photoreceiver with adjustable gain and bandwidth.
These receivers can be powered by batteries or by
an external ±15-V power supply. There are two
models available, each based on a different
photodetector. Free-space (FS) and fiber-coupled
(FC) versions are available for each model:
Complete specifications begin on page 21.
The 10-MHz three-stage transimpedance amplifier
includes selectable gain and selectable low- or high-
pass filters for easy signal optimization.
Model Wavelength Diode
Type Active Area
2107-FC 300–1070 nm silicon 0.8 mm2
2107-FS 300–1070 nm silicon 0.8 mm2
2117-FC 900–1700 nm InGaAs 0.0078 mm2
2117-FS 900–1700 nm InGaAs 0.08 mm2
Note:
Note:
21x7 10MHz Adj Blncd Rcvr revA.fm Page 5 Friday, January 7, 2005 12:22 PM

6 • Operation
Figure 1:
Typical
responsivities
of the Model
2107 & 2117
photodiodes
To obtain the value of the “response factor” in V/mW,
divide the photodiode responsivity by 1.5. For more
information on frequency response and noise, see page
13.
800
0.0
0.2
0.4
0.6
0.8
1.0
Wavelength (nm)
400
Responsivity (A/W)
1200 1600
2107
2117
Note:
Note:
21x7 10MHz Adj Blncd Rcvr revA.fm Page 6 Friday, January 7, 2005 12:22 PM

Models 2107 & 2117 Operation • 7
Using the Photoreceiver
1. Mount the photoreceiver. Use the 8-32 thread
(M4 for metric versions) on the bottom of the
casing to mount the photoreceiver to a post or
pedestal.
2. Supply power. Power the Model 21X7 using
either two 9-volt alkaline batteries or a ±15-V low-
noise linear power supply (such as the Newport
Model 0901).
3. Connect the receiver output. Connect your
voltmeter, oscilloscope, or other instrument to the
Output SMA connector on the receiver.
If you wish to connect to a BNC cable, you can
purchase a BNC-to-SMA adapter such as the
Newport Model 1225.
4. Turn on the photoreceiver power. For external
power, use ±15 VDC ON; for battery, use Batt
Mode ON.
5. Align optical beams onto the detectors.
The photodiodes are not very large, so take care
when aligning each beam.
6. Adjust the gain. Use the knob and rocker switch
on the receiver to set the gain. The bandwidths
vary with the gain setting (see table on page 10).
7. Adjust the filters. Select low-pass and high-pass
corner frequencies using the knobs on the receiver.
8. Balance the optical input levels. Alternately block
each diode and observe the signal strength. When
they are approximately equal and opposite, adjust
their relative intensity until the balanced output is
zero volts.
9. Turn off the photoreceiver power. When you are
finished with the receiver, place the power switch
in the ±15 VDC ON position and switch off or
unplug the external power supply.
Note:
Note:
21x7 10MHz Adj Blncd Rcvr revA.fm Page 7 Friday, January 7, 2005 12:22 PM

8 • Operation
Checking the Batteries
The Model 21X7 can be powered by two standard 9-
volt alkaline batteries. Under normal operating
conditions with low light levels and a high impedance
load attached to the BNC connector, the
photoreceiver draws about 20 mA from the batteries,
and the battery lifetime is approximately 24 hours.
To check the condition of the battery:
1. Turn on the photoreceiver using the power
switch.
2. Set the Low Frequency adjustment to DC.
3. Set the Gain to 3x104.
4. Focus at least 1 µW of optical power on the
detector (or place the detector in front of a desk
lamp).
The output should be greater than 7 V. If it is not,
replace the batteries with fresh ones.
Replacing the Batteries
The Model 21X7 is shipped with two fresh 9-V
batteries installed. To avoid confusion due to low
batteries, replace the batteries on a monthly basis
when the receiver is in frequent use, or use an external
linear power supply such as the Newport Model 0901.
1. Turn off the receiver using the power switch.
2. Use a Phillips-head screwdriver to remove the
two screws on the back panel of the
photoreceiver.
3. Remove the back panel.
4. Replace the used 9-V batteries with fresh ones.
5. Replace the back panel and the two screws.
6. Recheck the battery level as described above.
21x7 10MHz Adj Blncd Rcvr revA.fm Page 8 Friday, January 7, 2005 12:22 PM

Models 2107 & 2117 General Features & Principles • 9
General Features & Principles
Photoreceiver Circuitry
The circuitry inside the Model 21X7 consists of two
photodiodes followed by a three-stage transimpedance
amplifier. The gain can be adjusted from 626 V/A to
18.8x106 V/A in 5-dB steps. The low-noise amplifier
design is optimized to maximize bandwidth at each
gain setting. At the higher gain settings, the
bandwidth is limited by amplifier gain-bandwidth
product. The plots of Figure 3 show the typical
frequency responses for the different gain settings.
Figure 2:
Functional
schematic of
the Model 21X7
circuitry
+15 V
-15 V
GND
BATT
9 V
INDEPENDENTLY
ADJUSTABLE 6-dB/OCTAVE
HIGH- AND LOW-PASS FILTERS
+9 V
DETECTOR
HOUSING IS
GROUNDED
SMA
-9 V
ADJUSTABLE-GAIN
STAGE
x3
x1
ADJUSTABLE-GAIN
STAGE
x104
x103
x102
x10
x1
f_L f_H
BATT
9 V
+9 V REG
-9 V REG
+9 V
-9 V
21x7 10MHz Adj Blncd Rcvr revA.fm Page 9 Friday, January 7, 2005 12:22 PM

10 • General Features & Principles
The following table summarizes the bandwidth at each
gain setting. The bandwidth on the 3x settings is
somewhat lower than the 1x settings, and significantly
decreases at the highest gain settings. There is little
difference in frequency response between the visible
(Model 2107) and IR (Model 2117) models. The plots of
Figure 3 show the frequency-response details for each
gain setting.
Optical Power and Output Voltage
The typical operating range for these receivers is from
a few nanowatts up to 2 to 5 mW (depending on the
model and gain setting). Be careful to keep the
differential optical power below the maximum optical
power difference of 10 mW to avoid damaging the
photoreceiver.
To compute the approximate output voltage for a given
input optical power use the relationship
Vout = (P+-P-)·R·G,
Gain
Setting Specification Typical Performance
1x1 10 MHz 12 MHz
3x1 NA 6 MHz
1x10 NA 12 MHz
3x10 NA 6 MHz
1x102NA 8 MHz
3x102NA 6 MHz
1x103NA 700 kHz
3x103NA 700 kHz
1x104NA 250 kHz
3x104150 kHz 250 kHz
21x7 10MHz Adj Blncd Rcvr revA.fm Page 10 Friday, January 7, 2005 12:22 PM

Models 2107 & 2117 General Features & Principles • 11
where P+and P-are the input optical powers in Watts
on the right and left photodiodes respectively, Ris the
photodetector’s response factor in V/mW, and Gis the
amplifier’s gain setting.
Estimate the value of the response factor by dividing the
responsivity shown in Figure 1 by 1.5.
For example, the Model 2107 on the 1x103gain setting
and with 10 µW of optical power at 900 nm on one
photodiode will have an output voltage of
approximately
(0.01 mW)·(0.35 V/mW)·(1x103) = 3.5 V.
The maximum differential optical power that can be
detected by the photoreceiver is determined by the
input optical power at which either stage of the
transimpedance gain saturates. We can calculate the
saturation power at 900 nm for the Model 2107 at its
maximum output voltage of ±7 V with fresh batteries
or operating from an external ±15 VDC power supply.
Using the expression 7 V = Psat·R·G, the Model 2107 has
a differential saturation power of 20 mW for the lowest
gain setting up to 0.7 µW for the highest gain setting.
At other wavelengths where the responsivity is lower,
the saturation power increases inversely with response
factor.
Note:
Note:
21x7 10MHz Adj Blncd Rcvr revA.fm Page 11 Friday, January 7, 2005 12:22 PM

12 • General Features & Principles
21x7 10MHz Adj Blncd Rcvr revA.fm Page 12 Friday, January 7, 2005 12:22 PM
This page has been intentionally left blank

Models 2107 & 2117 Frequency Response and Noise • 13
Frequency Response and Noise
Measuring Bandwidth
The frequency response and noise characteristics of
the photoreceiver depend on the selected gain. The
figures beginning on page 16 give the typical
frequency response and noise behavior for the
photoreceivers at each of the gain settings. The
frequency response of the transimpedance gain is
plotted using the expression
20·log[Gain(ƒ)/Gain(0)],
where ƒis the frequency and Gain(0) is the gain at DC.
The photoreceiver’s bandwidth is defined as the
frequency where the gain has decreased by 3 dB, or a
factor of .
Measuring Noise
The photoreceiver noise is characterized using the
noise equivalent power (NEP), which is a measure of
the weakest optical signal that the photoreceiver can
detect. The NEP is the optical power which will
produce a signal-to-noise ratio of 1 in a 1-Hz
bandwidth. The minimum detectable optical power
can be found using the relationship
Minimum Optical Power = NEP · ,
where BW is the bandwidth. Note that NEP is a
wavelength-dependent quantity that changes with the
photodetector’s responsivity.
2
BW
21x7 10MHz Adj Blncd Rcvr revA.fm Page 13 Friday, January 7, 2005 12:22 PM

14 • Frequency Response and Noise
Another way to characterize the noise is with the
photocurrent noise (In), which is related to NEP by
In= R · NEP,
where Ris the photodetector’s responsivity (in A/W).
The photocurrent noise is independent of wavelength
because it gives the noise of the photoreceiver with the
photodetector’s responsivity factored out.
To characterize the noise of the photoreceiver, the
output electrical noise spectrum is measured with a
spectrum analyzer. This voltage noise spectrum is
converted to an equivalent optical photocurrent noise
by dividing the voltage noise by the transimpedance
gain (V/A). The photocurrent noise, In(ƒ), has units of
pA/ and is plotted in Figure 3 and Figure 4 using
the expression 20·log[In(ƒ)/1 A].
Calculating NEP
The noise equivalent power (NEP) can be calculated by
dividing the photocurrent noise by R, the detector’s
responsivity (see page 6).
From DC to 150 kHz the average photocurrent noise
for the Model 2107 on the high gain setting is about
0.4 pA/ , corresponding to an average NEP at
900 nm of 0.8 pW/ . The integrated noise
equivalent power from DC to 150 kHz is then obtained
by multiplying the average NEP by , the square
root of the bandwidth.
The expression BW = 2πƒ3-dB/4 for a one-pole low-
pass filter is useful for calculating the equivalent noise
bandwidth. Using the high-pass filter set 1 decade
below the low-pass cutoff reduces noise-equivalent
bandwidth by approximately 10 %. For the Model 2107
with a 3-dB bandwidth of 150 kHz, the equivalent
noise bandwidth is 235 kHz. This gives an optical noise
equivalent power of about 390 pW, so the minimum
detectable optical signal at 900 nm (with a signal-to-
noise ratio of 1) for the Model 2107 on the highest gain
Hz
Hz
Hz
BW
21x7 10MHz Adj Blncd Rcvr revA.fm Page 14 Friday, January 7, 2005 12:22 PM

Models 2107 & 2117 Frequency Response and Noise • 15
setting is 390 pW when operating at full detector
bandwidth.
You can further improve your signal-to-noise ratio by
using optical modulators or choppers with lock-in
amplifiers to limit the detection bandwidth. Using
such techniques you can reduce equivalent bandwidth
to 1 Hz or less.
Calculating Output-Voltage Noise
The output-voltage noise can be calculated from
G · R · NEP · ,
where Gis the gain (V/V), Ris the photodiode response
factor (V/mW), NEP is the average noise equivalent
power, and BW is the bandwidth. This gives an output
noise voltage for the Model 2107 on the high gain
setting of
(3x104V/V) · (0.35 V/mW) · (0.8x10-9 mW/ )
·=4mV
rms.
The Johnson noise at the input of a 100-MHz bandwidth
oscilloscope with 1-MΩinput impedance is 1.6 mVrms. This is
often the limiting factor in broadband measurements.
Summary
With the Model 2107 on the highest gain setting the
minimum NEP is 0.8 pW/ , and this yields an
output noise voltage of 4 mVrms. Viewed another way,
for operation at the peak responsivity wavelength of
900 nm and for the high gain setting, you will achieve a
signal-to-noise ratio of unity if the input power is
390 pW.
For the Model 2117 with an InGaAs photodiode, the
NEP at peak response wavelength of 1500 nm is
0.4 pW/ over the 150-kHz bandwidth. The full
BW
Hz
2π
4
------ 150 103Hz×⋅
Hz
Hz
21x7 10MHz Adj Blncd Rcvr revA.fm Page 15 Friday, January 7, 2005 12:22 PM

16 • Frequency Response and Noise
bandwidth signal-to-noise ratio of 1 is achieved
around 200 pW.
Note that this assumes operation without any post-
photoreceiver filtering and with the full photoreceiver
bandwidth. By using the built-in electronic band-pass
filter or an optical chopper and a lock-in amplifier, the
receiver can detect significantly weaker optical signals.
Performance Data for Frequency Response
The 3-dB frequency bandwidth is defined as the
frequency where the photoreceiver’s transimpedance
gain has decreased by a factor of . The typical
frequency responses for the Model 2107 and Model
2117 are shown in the following figures.
Figure 3:
Typical
frequency
response for
Model 21X7 at
each gain
setting
2
Gain Setting=1
-15
-12
-9
-6
-3
0
3
0.01 0.1 1 10 100
Frequency (MHz)
Normalized Gain (dB)
x1
x3
Gain Setting=10
-12
-9
-6
-3
0
3
0.01 0.1 1 10 100
Frequency (MHz)
Normalized Gain (dB)
x1
x3
21x7 10MHz Adj Blncd Rcvr revA.fm Page 16 Friday, January 7, 2005 12:22 PM

Models 2107 & 2117 Frequency Response and Noise • 17
Gain Setting=102
-15
-12
-9
-6
-3
0
3
0.01 0.1 1 10 100
Frequency (MHz)
Normalized Gain (dB)
x1
x3
Gain Setting=103
-6
-3
0
3
0.01 0.1 1
Frequency (MHz)
Normalized Gain (dB)
x1
x3
Gain Setting=104
-6
-3
0
3
0.01 0.1 1
Frequency (MHz)
Normalized Gain (dB)
x1
x3
21x7 10MHz Adj Blncd Rcvr revA.fm Page 17 Friday, January 7, 2005 12:22 PM

18 • Frequency Response and Noise
Performance Data for Noise
Figure 4 shows the typical noise spectrum expressed as
photocurrent noise for Model 21X7 photoreceivers on
the highest gain setting.
To derive the receiver’s Noise Equivalent Power (NEP),
divide the photocurrent noise by the photodiode
responsivity. To convert to output voltage noise (RMS),
multiply the photocurrent noise by the gain setting
from the 21X7 front label, then by 630 V/A (the scaling
factor between the gain setting labels and the actual
amplifier transimpedance gain).
For example, the output voltage noise (RMS) for Model
2117 in the 3x103setting is approximately:
0.4 pA/ x 3 x 103x 630 V/A = 0.75 µVrms/.
For the 700 kHz of amplifier bandwidth in the 3x103
gain setting, the equivalent noise bandwidth is:
( 2 x π/4 ) x 700 x 103Hz = 1.1 MHz,
so the predicted output noise voltage is approximately
0.75 µVrms/ x = 0.8 mVrms.
Because the NEP is listed at the highest gain setting,
some additional considerations add to the NEP at lower
gain settings. First, the noise spectrum (Figure 4) is not
flat, rising at frequencies above 100 kHz. This
contributes an extra 20% to the output noise voltage in
the 3 x 103setting compared to 3 x 104. Also, as the
output noise voltage approaches 1 mVrms, the Johnson
noise limit of your measurement instrument will
become important. Note that the Johnson noise for an
oscilloscope with 100-MHz bandwidth (assuming
perfect roll off) and 1-MΩinput impedance is
1.2 mVrms.
Hz Hz
Hz 1.1 106
×Hz
21x7 10MHz Adj Blncd Rcvr revA.fm Page 18 Friday, January 7, 2005 12:22 PM

Models 2107 & 2117 Frequency Response and Noise • 19
Figure 4:
Typical noise
spectrum for
Model 21X7
Common Mode Rejection
Using the Model 21X7 balanced photoreceivers with
equal signal powers on each photodiode results in an
output with reduced common mode signal. The
common-mode rejection ratio, or CMRR, is a
measurement of the effectiveness of the balanced
subtraction.
Figure 5 shows the CMRR of each model using the
following definition:
CMRR = -20 · log((Vout1 - Vout2)/Vout1).
Because the common mode subtraction occurs before
the first amplifier stage, the only practical bandwidth
limitation on common mode rejection is the
photodiode bandwidth. Thus, as seen in the following
figures, the CMRR for higher gain settings is relatively
flat to frequencies well beyond the useful frequency
response of the gain setting.
21X7 Output Noise Current
0.20
0.25
0.30
0.35
0.40
0 102030405060708090100
Frequency (KHz)
Noise (pA/ Hz)
Spec
Gain = 3 x 104
21x7 10MHz Adj Blncd Rcvr revA.fm Page 19 Friday, January 7, 2005 12:22 PM

20 • Frequency Response and Noise
Figure 5:
Typical CMRR
for
Model 2107 in
each gain
setting
Figure 6:
Typical CMRR
for
Model 2117 in
each gain
setting
20
25
30
35
40
45
50
0.01 0.1 1 10 100
Frequency (MHz)
CMRR (dB)
1x103- 3x103
1 - 3x102
1x104- 3x104
20
25
30
35
40
45
50
0.01 0.1 10 100
Frequency (MHz)
CMRR (dB)
1x103- 3x103
1 - 3x102
1x104- 3x104
1
21x7 10MHz Adj Blncd Rcvr revA.fm Page 20 Friday, January 7, 2005 12:22 PM