Philips Semiconductors Product specification
NE5210Transimpedance amplifier (280MHz)
1995 Apr 26
12
V
CC
V
EQ3
V
IN
I
IN
INPUT
I
F
I
B
Q1
Q2
Q3
R2
R3
R4
R
F
R1
I
C1
SD00329
Figure 3. Shunt-Series Input Stage
The operating point of Q1, Figure 2, has been optimized for the
lowest current noise without introducing a second dominant pole in
the pass-band. All poles associated with subsequent stages have
been kept at sufficiently high enough frequencies to yield an overall
single pole response. Although wider bandwidths have been
achieved by using a cascode input stage configuration, the present
solution has the advantage of a very uniform, highly desensitized
frequency response because the Miller effect dominates over the
external photodiode and stray capacitances. For example, assuming
a source capacitance of 1pF, input stage voltage gain of 70, R
IN
=
60Ω then the total input capacitance, C
IN
= (1+7.5) pF which will
lead to only a 12% bandwidth reduction.
NOISE
Most of the currently installed fiber-optic systems use non-coherent
transmission and detect incident optical power. Therefore, receiver
noise performance becomes very important. The input stage
achieves a low input referred noise current (spectral density) of
3.5pA/√Hz
. The transresistance configuration assures that the
external high value bias resistors often required for photodiode
biasing will not contribute to the total noise system noise. The
equivalent input
RMS
noise current is strongly determined by the
quiescent current of Q
1
, the feedback resistor RF, and the
bandwidth; however, it is not dependent upon the internal
Miller-capacitance. The measured wideband noise was 66nA
RMS
in
a 200MHz bandwidth.
DYNAMIC RANGE CALCULATIONS
The electrical dynamic range can be defined as the ratio of
maximum input current to the peak noise current:
Electrical dynamic range, D
E
, in a 200MHz bandwidth assuming
I
INMAX
= 240µA and a wideband noise of IEQ=66nA
RMS
for an
external source capacitance of C
S
= 1pF.
D
E
+ 20log
(Max. input current) (PK)
(Peak noise current) (RMS)@ 2
Ǹ
+ 20log
(240 @ 10
*6
)
( 2
Ǹ
66 10*9)
+ 68dB
In order to calculate the optical dynamic range the incident optical
power must be considered.
For a given wavelength λ; (meters)
Energy of one Photon =
hc
l
watt sec (Joule)
Where h=Planck’s Constant = 6.6 × 10
-34
Joule sec.
c = speed of light = 3 × 10
8
m/sec
c / λ = optical frequency (Hz)
No. of incident photons/sec= where P=optical incident power
No. of incident photons/sec =
P
hs
l
where P = optical incident power
No. of generated electrons/sec =
h @
P
hs
l
where η = quantum efficiency
+
no. of generated electron hole paris
no. of incident photons
N I + h @
P
hs
l
@ e Amps (Coulombsń sec.)
where e = electron charge = 1.6 × 10
-19
Coulombs
Responsivity R =
h @e
hs
l
Amp/watt
I + P @ R
Assuming a data rate of 400 Mbaud (Bandwidth, B=200MHz), the
noise parameter Z may be calculated as:
1
Z +
I
EQ
qB
+
66 @ 10
*9
(1.6@ 10
*19
)(200 @ 106)
+ 2063
where Z is the ratio of
RMS
noise output to the peak response to a
single hole-electron pair. Assuming 100% photodetector quantum
efficiency, half mark/half space digital transmission, 850nm
lightwave and using Gaussian approximation, the minimum required
optical power to achieve 10
-9
BER is:
P
avMIN
+ 12
hc
l
B Z + 12 2.3@ 10
*19
200@ 1062063
+ 1139nW + * 29.4dBm
where h is Planck’s Constant, c is the speed of light, λ is the
wavelength. The minimum input current to the NE5210, at this input
power is:
I
avMIN
+ qP
avMIN
l
hc
+
1139@ 10
*9
@ 1.6 @ 10
*19
2.3@ 10
*19
= 792nA
Choosing the maximum peak overload current of I
avMAX
=240µA, the
maximum mean optical power is:
P
avMAX
+
hcI
avMAX
l q
+
2.3@ 10
*19
1.6@ 10
*19
240 @ 10
*6
Thus the optical dynamic range, DO is:
D
O
= P
avMAX
- P
avMIN
= -4.6 -(-29.4) = 24.8dB.