13
LT6301
sn6301 6301f
of n. To analyze this circuit, first ground the input. As RBT =
RL/n, and assuming RP2>>RL we require that:
VA = VO (1 – 1/n) to increase the effective value of
RBT by n.
VP = VO (1 – 1/n)/(1 + RF/RG)
VO = VP (1 + RP2/RP1)
Eliminating VP, we get the following:
(1 + RP2/RP1) = (1 + RF/RG)/(1 – 1/n)
For example, reducing RBT by a factor of n = 4, and with an
amplifer gain of (1 + RF/RG) = 10 requires that RP2/RP1
= 12.3.
Note that the overall gain is increased:
V
RR
R
nR R
R
O
I
PP
P
FG
P
=
+
()
+
() +
()
[]+
()
[]
22
1
12
1
11
1
/
//
/
A simpler method of using positive feedback to reduce the
back-termination is shown in Figure 14. In this case, the
drivers are driven differentially and provide complemen-
tary outputs. Grounding the inputs, we see there is invert-
ing gain of –RF/RP from –VO to VA
VA = VO (RF/RP)
and assuming RP >> RL, we require
VA = VO (1 – 1/n)
solving
RF/RP = 1 – 1/n
So to reduce the back-termination by a factor of 3 choose
RF/RP = 2/3. Note that the overall gain is increased to:
VO/VI = (1 + RF/RG + RF/RP)/[2(1 – RF/RP)]
Using positive feedback is often referred to as active
termination.
Figure 16 shows a full-rate ADSL line driver incorporating
positive feedback to reduce the power lost in the back
termination resistors by 40% yet still maintains the proper
impedance match to the100
characteristic line imped-
ance. This circuit also reduces the transformer turns ratio
over the standard line driving approach resulting in lower
peak current requirements. With lower current and less
power loss in the back termination resistors, this driver
dissipates only 1W of power, a 30% reduction.
While the power savings of positive feedback are attractive
there is one important system consideration to be ad-
dressed, received signal sensitivity. The signal received
APPLICATIO S I FOR ATIO
WU
UU
Figure 14. Back Termination Using Differential Postive Feedback
Figure 13. Back Termination Using Postive Feedback
+
–
6301 F13
RF
RBT
RP2
RP1
RG
VI
VA
VP
VO
RL
RF
RG
1 +
RL
n
=
VO
VI
= 1 –
–
1
n
FOR RBT =
()
RF
RG
1 +
()
RP1
RP1 + RP2
RP1
RP2 + RP1
RP2/(RP2 + RP1)
()
1 + 1/n
–
+
RBT
RF
RG
RP
RG
RL
–VI
VA
–VA
VI
–VO
VO
–
+
RBT
6301 F14
RF
RL
n
=
VO
VI
n =
1 –
2
FOR RBT =
RF
RP
RF
RP
+
RF
RG
1 +
1 –
RF
RP
1
()