Datasheet
C
C2
=
C
C1
Sf
SW
R
C1
C
C1
-1
= 71 pF
1
C
C1
=
Sf
LC
R
C1
= 1.99 nF
R
C1
=
f
CROSSOVER
f
LC
'
V
RAMP
V
IN
R
FB1
=
100 kHz
17.4 kHz
10 k:
0.8 V
5.0 V
= 9.2 k:
f
Z2
= f
LC
=
f
P1
= f
ESR
=
2
f
P2
=
f
sw
f
Z1
=
f
LC
2
1
2SR
C1
C
C1
=
1
2S(R
C1
+ R
FB1
)C
C3
C
C1
+
C
C2
2SR
C1
C
C1
C
C2
1
2SR
C2
C
C3
=
LM21215A
www.ti.com
SNOSB87B –MARCH 2011–REVISED MARCH 2013
The dependancy of the pole and zero locations on the compensation components is described below.
An example of the step-by-step procedure to generate compensation component values using the typical
application setup (see Figure 40) is given. The parameters needed for the compensation values are given in the
table below.
Parameter Value
V
IN
5.0V
V
OUT
1.2V
I
OUT
15A
f
CROSSOVER
100 kHz
L 0.56 µH
R
DCR
1.8 mΩ
C
O
150 µF
R
ESR
1.0 mΩ
ΔV
RAMP
0.8V
f
SW
500 kHz
where ΔV
RAMP
is the oscillator peak-to-peak ramp voltage (nominally 0.8V), and f
CROSSOVER
is the frequency at
which the open-loop gain is a magnitude of 1. It is recommended that the f
crossover
not exceed one-fifth of the
switching frequency. The output capacitance, C
O
, depends on capacitor chemistry and bias voltage. For Multi-
Layer Ceramic Capacitors (MLCC), the total capacitance will degrade as the DC bias voltage is increased.
Measuring the actual capacitance value for the output capacitors at the output voltage is recommended to
accurately calculate the compensation network. The example given here is the total output capacitance using the
three MLCC output capacitors biased at 1.2V, as seen in the typical application schematic, Figure 40. Note that it
is more conservative, from a stability standpoint, to err on the side of a smaller output capacitance value in the
compensation calculations rather than a larger, as this will result in a lower bandwidth but increased phase
margin.
First, a the value of R
FB1
should be chosen. A typical value is 10kΩ. From this, the value of R
C1
can be
calculated to set the mid-band gain so that the desired crossover frequency is achieved:
(13)
Next, the value of C
C1
can be calculated by placing a zero at half of the LC double pole frequency (f
LC
):
(14)
Now the value of C
C2
can be calculated to place a pole at half of the switching frequency (f
SW
):
(15)
Copyright © 2011–2013, Texas Instruments Incorporated Submit Documentation Feedback 21
Product Folder Links: LM21215A