Hardware Reference
In-Depth Information
Fig. 3.15 A simplified
schematic of the bipolar
implementation of CCI+
proposed by Fabre
et al. [ 14 ]
+V
I 0
Q 1
Q 5
Q 2
y
z
x
Q 3
Q 6
Q 4
−V
which essentially involve the mechanism of providing a current gain between i z and
i x as explained above. As a consequence, the current transfer from X to Z port is,
therefore, given by i z ¼
Gi x .
Implementation of the CCII+ with current gain described above using high
performance ALA 200 transistor arrays from ATT with circuit biased from
5V
DC power supplies and with DC bias current I 0 chosen as 50
ʼ
A, demonstrated that
G could be varied from 0.5 to 5 with the resulting
3 dB bandwidth of the current
gain remaining higher than 146 MHz. The voltage transfer between port Y and X
was found to be 0.99 with
3 dB bandwidth of 182 MHz while the input impedance
at port Y was found to be consisting of resistance R y ¼
160 M
ʩ
in parallel with
capacitance C y ¼
1.1 pF.
3.2.11 Bipolar Implementations of the CCI
A class AB CCI+ devised by Fabre et al. [ 14 ] is shown in Fig. 3.15 whose basic core
consists of two translinear mixed loops in parallel. Due to the operation in class AB
mode, the circuit can process current with magnitudes greater than the DC bias
current I 0 and therefore, generate very low harmonic distortion.
The implementation of this CCI+ using transistor arrays ALA200 from AT&T
which has NPN transistors with f T ¼
3.5Ghz and h FE ¼
140 and f T ¼
2Ghz and
h FE ¼
40 for PNP transistors it has been found that with power supply as low as
A, the circuit exhibits a bandwidth of
120 MHz, dissipates power less than 1 mW with the X-port parasitic input resis-
tance of the order of 17
1.2 V and DC bias current I 0 as 100
ʼ
. The circuit clearly is a good choice for low voltage and
low power applications requiring CCI+ as a building block.
ʩ
 
Search WWH ::




Custom Search