ATA06211
(see VOUT vs IIN figure) from the TIA and determining
the point of output voltage collapse. Also the input
node virtual ground during heavy AGC is checked
to verify that the linearity (i.e. pulse width distortion)
of the amplifier has not been compromised.
1.9
1.85
1.8
VDD = 5.5 V
VDD = 5.0V
1.75
1.7
Measurement of Input Referred Noise Current
The Input Noise Current is directly related to
sensitivity . It can be defined as the output noise
voltage (Vout), with no input signal, (including a 450
MHz lowpass filter at the output of the TIA) divided by
the AC transresistance.
1.65
1.6
1.55
1.5
VDD = 4.5V
1 0
- 4 0
6 0
Temperature (OC)
7.5
6.5
Figure 9: Input Offset Voltage vs. Temperature
R
f
5.5
4.5
3.5
2.5
1.5
CBY Connection
CT
50
The CBY pad must be connected via a low inductance
path to a surface mount capacitor of at least 56pF
(additional capacitance can be added in parallel with
the 56 pF or 220 pF capacitors to improve low
frequency response and noise performance).
Referring to the equivalent circuit diagram and the
typical bonding diagram, it is critical that the
connection from CBY to the bypass capacitor use
two bond wires for low inductance, since any high
frequency impedance at this node will be fed back to
the open loop amplifier with a resulting loss of
transimpedance bandwidth. Two pads are provided
for this purpose
CT=1.0pF
CT =0.5pF
1000
- 0.1
1
10
100
Frequency (MHz)
Figure 10: Input Referred Noise Spectral Density
Input referred noisetestcircuit
70
VDD = 4.5V
25dB
Sensitivity and Bandwidth
In order to guarantee sensitivity and bandwidth
performance, the TIA is subjected to
comprehensive series of tests at the die sort level
60
400
MHz
TIA
a
0.5pF
58
LPF
o
(100% testing at 25 C) to verify the DC parametric
54
VDD =5.5V
performance and the high frequency performance
(i.e. adequate |S21|) of the amplifier. Acceptably high
|S21| of the internal gain stages will ensure low
amplifier input capacitance and hence low input
referred noise current. Transimpedance sensitivity
and bandwidth are then guaranteed by design and
correlation with RF and DC die sort test results.
50
6500in
R
η(dBm)=10LOG
46
42
-40
0
40
80
Temperature (OC)
FIgure 11: Input Referred Noise vs Temperature
Indirect Measurement of Optical Overload
Optical overload can be defined as the maximum
optical power above which the BER (bit error rate)
increases beyond 1 error in 1010 bits. The
ATA06211D1C is 100% tested at die sort by a DC
measurement which has excellent correlation with
an PRBS optical overload measurement. The
measurement consists of sinking a negative current
6
PRELIMINARY DATA SHEET - Rev 4
08/2001