Datasheet

R
C1
=
x
C
C1
C
OUT
I
OUT
V
OUT
+
15 x D
V
IN
-1
+
1-D
f
SW
x L
COMP
C
C1
R
C1
C
C2
LM20134
(optional)
A
M
f
SW
/2
0 dB
FREQUENCY (Hz)
GAIN (dB)
Error Amp Zero, f
Z(EA)
Complex Double Pole, f
P(MOD)
Optional Error Amp
Pole, f
P2(EA)
0 dB
0 dB
A
EA
+ A
M
Error Amplifier
Transfer Function
Modulator and Output Filter
Transfer Function
Compensated Closed
Loop Transfer Function
A
EA
Error Amp Pole, f
P1(EA)
Complex Double Pole, f
P(MOD)
Output Filter Zero, f
Z(FIL)
Output Filter Pole, f
P(FIL)
f
C
Error Amp Pole, f
P(EA)
LM20134, LM20134Q
SNVS527G –OCTOBER 2007–REVISED MARCH 2013 www.ti.com
Figure 30. LM20134 Loop Compensation
The power stage transfer function is dictated by the modulator, output LC filter, and load; while the feedback
transfer function is set by the feedback resistor ratio, error amp gain, and external compensation network.
To achieve a -20dB/decade slope, the error amplifier zero, located at f
Z(EA)
, should positioned to cancel the
output filter pole (f
P(FIL)
). An additional error amp pole, located at f
P2(EA)
, can be added to cancel the output filter
zero at (f
Z(FIL)
). Cancellation of the output filter zero is recommended if larger value, non-ceramic output
capacitors are used.
Compensation of the LM20134 is achieved by adding an RC network as shown in Figure 31 below.
Figure 31. Compensation Network for LM20134
A good starting value for C
C1
for most applications is 4.7 nF. Once the value of C
C1
is chosen the value of RC
should be calculated using the equation below to cancel the output filter pole (f
P(FIL)
) as shown in Figure 30.
(8)
A higher crossover frequency can be obtained, usually at the expense of phase margin, by lowering the value of
C
C1
and recalculating the value of R
C1
. Likewise, increaseing C
C1
and recalculating R
C1
will provide additional
phase margin at a lower crossover frequency. As with any attempt to compensate the LM20134 the stability of
the system should be verified for desired transient droop and settling time.
If the output filter zero, (f
Z(FIL)
) approaches the crossover frequency (F
C
), an additional capacitor (C
C2
) should be
placed at the COMP pin to ground. This capacitor adds a pole to cancel the output filter zero assuring the
crossover frequency will occur before the double pole at f
SW
/2 degrades the phase margin. The output filter zero
is set by the output capacitor value and ESR as shown in the equation below.
18 Submit Documentation Feedback Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM20134 LM20134Q