Datasheet

100 1k 10k 100k 1M
FREQUENCY (Hz)
-40
-32.5
-25
-17.5
-10
-2.5
5
ERROR AMP GAIN (dB)
100 1k 10k 100k 1M
FREQUENCY (Hz)
-180
-120
-60
0
60
120
180
ERROR AMP PHASE (°)
G
EA-ACTUAL
=
G
EA
x OPG
1 + G
EA
x OPG
OPG =
2S x GBW
2S x GBW
A
DC
s +
C1 =
C2
2S x C2 x R1 x f
P1
-1
:
C2 =
1
2S x R1 x f
Z1
:
LM5022
www.ti.com SNVS480G –JANUARY 2007–REVISED DECEMBER 2013
For this example, C2 = 125 nF
8. Select a frequency for the compensation pole, f
P1
: The suggested placement for this pole is at one-fifth of
the switching frequency. For this example, f
P1
= 100 kHz
9. Set
For this example, C1 = 530 pF
10. Plug the closest 1% tolerance values for R
FB2
and R1, then the closest 10% values for C1 and C2 into
G
EA
and model the error amp: The open-loop gain and bandwidth of the LM5022’s internal error amplifier
are 75 dB and 4 MHz, respectively. Their effect on G
EA
can be modeled using the following expression:
A
DC
is a linear gain, the linear equivalent of 75 dB is approximately 5600V/V. C1 = 560 pF 10%, C2 = 120 nF
10%, R1 = 3.01 kΩ 1%
11. Plot or evaluate the actual error amplifier transfer function:
Figure 21. Error Amplifier Gain and Phase Figure 22. Error Amplifier Gain and Phase
12. Plot or evaluate the complete control loop transfer function: The complete control loop transfer function
is obtained by multiplying the power stage and error amplifier functions together. The bandwidth and phase
margin can then be read graphically or evaluated numerically.
Copyright © 2007–2013, Texas Instruments Incorporated Submit Documentation Feedback 23
Product Folder Links: LM5022