19
LT1506
APPLICATIONS INFORMATION
WUU
U
Figure 10. Response from V
C
Pin to Output
FREQUENCY (Hz)
GAIN (µMho)
PHASE (DEG)
3000
2500
2000
1500
1000
500
200
150
100
50
0
–50
100 10k 100k 10M
1506 F11
1k 1M
GAIN
PHASE
R
OUT
200k
C
OUT
12pF
V
C
ERROR AMPLIFIER EQUIVALENT CIRCUIT
R
LOAD
= 50
V
FB
2 × 10
–3
)(
Figure 11. Error Amplifier Gain and Phase
Figure 12. Overall Loop Characteristics
What About a Resistor in the Compensation Network?
It is common practice in switching regulator design to add
a “zero” to the error amplifier compensation to increase
loop phase margin. This zero is created in the external
network in the form of a resistor (R
C
) in series with the
compensation capacitor. Increasing the size of this resis-
tor generally creates better and better loop stability, but
there are two limitations on its value. First, the combina-
tion of output capacitor ESR and a large value for R
C
may
cause loop gain to stop rolling off altogether, creating a
gain margin problem. An approximate formula for R
C
where gain margin falls to zero is:
R Loop
V
G G ESR
C
OUT
MP MA
Gain =1
()
=
()()()()
242.
FREQUENCY (Hz)
GAIN: V
C
PIN TO OUTPUT (dB)
PHASE: V
C
PIN TO OUTPUT (DEG)
40
20
0
–20
–40
40
0
–40
–80
120
10 1k 10k 1M
1505 F10
100 100k
GAIN
PHASE
V
IN
= 10V
V
OUT
= 5V
I
OUT
= 2A
FREQUENCY (Hz)
LOOP GAIN (dB)
LOOP PHASE (DEG)
80
60
40
20
0
–20
200
150
100
50
0
–50
10 1k 10k 1M
1505 F12
100 100k
GAIN
PHASE
V
IN
= 10V
V
OUT
= 5V, I
OUT
= 2A
C
OUT
= 100µF, 10V, AVX TPS
C
C
= 1.5nF, R
C
= 0, L = 10µH
Error amplifier transconductance phase and gain are shown
in Figure 11. The error amplifier can be modeled as a
transconductance of 2000µMho, with an output imped-
ance of 200k in parallel with 12pF. In all practical
applications, the compensation network from V
C
pin to
ground has a much lower impedance than the output
impedance of the amplifier at frequencies above 500Hz.
This means that the error amplifier characteristics them-
selves do not contribute excess phase shift to the loop, and
the phase/gain characteristics of the error amplifier sec-
tion are completely controlled by the external compensa-
tion network.
In Figure 12, full loop phase/gain characteristics are
shown with a compensation capacitor of 1.5nF, giving the
error amplifier a pole at 530Hz, with phase rolling off to 90°
and staying there. The overall loop has a gain of 74dB at
low frequency, rolling off to unity-gain at 100kHz. Phase
shows a two-pole characteristic until the ESR of the output
capacitor brings it back above 10kHz. Phase margin is
about 60° at unity-gain.
Analog experts will note that around 4.4kHz, phase dips
very close to the zero phase margin line. This is typical of
switching regulators, especially those that operate over a
wide range of loads. This region of low phase is not a
problem as long as it does not occur near unity-gain. In
practice, the variability of output capacitor ESR tends to
dominate all other effects with respect to loop response.
Variations in ESR
will
cause unity-gain to move around,
but at the same time phase moves with it so that adequate
phase margin is maintained over a very wide range of ESR
( ±3:1).
20
LT1506
APPLICATIONS INFORMATION
WUU
U
cases, the resistor may have to be larger to get acceptable
phase response, and some means must be used to control
ripple voltage at the V
C
pin. The suggested way to do this
is to add a capacitor (C
F
) in parallel with the R
C
/C
C
network
on the V
C
pin. Pole frequency for this capacitor is typically
set at one-fifth of switching frequency so that it provides
significant attenuation of switching ripple, but does not
add unacceptable phase shift at loop unity-gain frequency.
With R
C
= 3k,
C
fR
k
pF
F
C
=
()()()
=
()
=
5
2
5
2 500 10 3
531
3
π
π
How Do I Test Loop Stability?
The “standard” compensation for LT1506 is a 1.5nF
capacitor for C
C
, with R
C
= 0. While this compensation will
work for most applications, the “optimum” value for loop
compensation components depends, to various extent, on
parameters which are not well controlled. These include
inductor value
(±30% due to production tolerance, load
current and ripple current variations),
output capacitance
(±20% to ±50% due to production tolerance, tempera-
ture, aging and changes at the load),
output capacitor ESR
(±200% due to production tolerance, temperature and
aging), and finally,
DC input voltage and output load
current
. This makes it important for the designer to check
out the final design to ensure that it is “robust” and tolerant
of all these variations.
I check switching regulator loop stability by pulse loading
the regulator output while observing transient response at
the output, using the circuit shown in Figure 13. The
regulator loop is “hit” with a small transient AC load
current at a relatively low frequency, 50Hz to 1kHz. This
causes the output to jump a few millivolts, then settle back
to the original value, as shown in Figure 14. A well behaved
loop will settle back cleanly, whereas a loop with poor
phase or gain margin will “ring” as it settles. The
number
of rings indicates the degree of stability, and the
frequency
of the ringing shows the approximate unity-gain fre-
quency of the loop.
Amplitude
of the signal is not particu-
larly important, as long as the amplitude is not so high that
the loop behaves nonlinearly.
G
MP
= Transconductance of power stage = 5.3A/V
G
MA
= Error amplifier transconductance = 2(10
–3
)
ESR = Output capacitor ESR
2.42 = Reference voltage
With V
OUT
= 5V and ESR = 0.03, a value of 6.5k for R
C
would yield zero gain margin, so this represents an upper
limit. There is a second limitation however which has
nothing to do with theoretical small signal dynamics. This
resistor sets high frequency gain of the error amplifier,
including the gain at the switching frequency. If switching
frequency gain is high enough, output ripple voltage will
appear at the V
C
pin with enough amplitude to muck up
proper operation of the regulator. In the marginal case,
subharmonic
switching occurs, as evidenced by alternat-
ing pulse widths seen at the switch node. In more severe
cases, the regulator squeals or hisses audibly even though
the output voltage is still roughly correct. None of this will
show on a theoretical Bode plot because Bode is an
amplitude insensitive analysis.
Tests have shown that if
ripple voltage on the V
C
is held to less than 100mV
P-P
, the
LT1506 will be well behaved
. The formula below will give
an estimate of V
C
ripple voltage when R
C
is added to the
loop, assuming that R
C
is large compared to the reactance
of C
C
at 500kHz.
V
R G V V ESR
VLf
C RIPPLE
C MA IN OUT
IN
(
)
=
()( )
()()()
()()()
24.
G
MA
= Error amplifier transconductance (2000µMho)
If a computer simulation of the LT1506 showed that a
series compensation resistor of 3k gave best overall loop
response, with adequate gain margin, the resulting V
C
pin
ripple voltage with V
IN
= 10V, V
OUT
= 5V, ESR = 0.1,
L = 10µH, would be:
V
k
V
C RIPPLE
(
)
=
()
()()()
()
=
3 2 10 10 5 0 1 2 4
10 10 10 500 10
0 144
3
63
•..
••
.
This ripple voltage is high enough to possibly create
subharmonic switching. In most situations a compromise
value (<2k in this case) for the resistor gives acceptable
phase margin and no subharmonic problems. In other
21
LT1506
APPLICATIONS INFORMATION
WUU
U
Figure 13. Loop Stability Test Circuit
TO
OSCILLOSCOPE
SYNC
ADJUSTABLE
DC LOAD
ADJUSTABLE
INPUT SUPPLY
100Hz TO 1kHz
100mV TO 1V
P-P
100µF TO
1000µF
RIPPLE FILTER
1506 F13
TO X1
OSCILLOSCOPE
PROBE
3300pF 330pF
50
470
4.7k
SWITCHING
REGULATOR
+
0.2ms/DIV 1375/76 F14
10mV/DIV
V
OUT
AT I
OUT
=
500mA
BEFORE FILTER
V
OUT
AT I
OUT
=
500mA
AFTER FILTER
V
OUT
AT I
OUT
= 50mA
AFTER FILTER
LOAD PULSE
THROUGH 50
f 780Hz
5A/DIV
Figure 14. Loop Stability Check
The output of the regulator contains both the desired low
frequency transient information and a reasonable amount
of high frequency (500kHz) ripple. The ripple makes it
difficult to observe the small transient, so a two-pole,
100kHz filter has been added. This filter is not particularly
critical; even if it attenuated the transient signal slightly,
this wouldn’t matter because amplitude is not critical.
After verifying that the setup is working correctly, I start
varying load current and input voltage to see if I can find
any combination that makes the transient response look
suspiciously “ringy.” This procedure may lead to an
adjustment for best loop stability or faster loop transient
response. Nearly always you will find that loop response
looks better if you add in several k for R
C
. Do this only
if necessary, because as explained before, R
C
above 1k
may require the addition of C
F
to control V
C
pin ripple. If
everything looks OK, I use a heat gun and cold spray on the
circuit (especially the output capacitor) to bring out any
temperature-dependent characteristics.
Keep in mind that this procedure does not take initial
component tolerance into account. You should see fairly
clean response under all load and line conditions to ensure
that component variations will not cause problems. One
note here: according to Murphy, the component most
likely to be changed in production is the output capacitor,
because that is the component most likely to have manu-
facturer variations (in ESR) large enough to cause prob-
lems. It would be a wise move to lock down the sources of
the output capacitor in production.
A possible exception to the “clean response” rule is at very
light loads, as evidenced in Figure 14 with I
LOAD
= 50mA.
Switching regulators tend to have dramatic shifts in loop
response at very light loads, mostly because the inductor
current becomes discontinuous. One common result is very
slow but stable characteristics. A second possibility is low
phase margin, as evidenced by ringing at the output with
transients. The good news is that the low phase margin at
light loads is not particularly sensitive to component varia-
tion, so if it looks reasonable under a transient test, it will
probably not be a problem in production. Note that
fre-
quency
of the light load ringing may vary with component
tolerance but phase margin generally hangs in there.
CURRENT SHARING MULTIPHASE SUPPLY
The circuit in Figure 15 uses multiple LT1506s to produce
a 5V, 12A power supply. There are several advantages to
using a multiple switcher approach compared to a single
larger switcher. The inductor size is considerably reduced.
Three 4A inductors store less energy (LI
2
/2) than one 12A
coil so are far smaller. In addition, synchronizing three

LT1506CR-3.3#TRPBF

Mfr. #:
Manufacturer:
Analog Devices / Linear Technology
Description:
Switching Voltage Regulators 4.5A, 500kHz Buck Sw Reg
Lifecycle:
New from this manufacturer.
Delivery:
DHL FedEx Ups TNT EMS
Payment:
T/T Paypal Visa MoneyGram Western Union