LTC4266
1
4266fa
n High Power PSE Switches/Routers
n High Power PSE Midspans
ApplicAtions
FeAtures Description
Quad IEEE 802.3at Power
Over Ethernet Controller
The LTC
®
4266 is a quad PSE controller designed for use
in IEEE 802.3 Type 1 and Type 2 (high power) compliant
Power over Ethernet systems. External power MOSFETs
enhance system reliability and minimize channel resis-
tance, cutting power dissipation and eliminating the need
for heatsinks even at Type 2 power levels. External power
components also allow use at very high power levels while
remaining otherwise compatible with the IEEE standard.
80V-rated port pins provide robust protection against
external faults.
The LTC4266 includes advanced power management
features, including current and voltage readback and
programmable ICUT and ILIM thresholds. Available C librar-
ies simplify power-management software development;
an optional AUTO mode provides fully IEEE-compliant
standalone operation with no software required. Proprietary
4-point PD detection circuitry minimizes false PD detec-
tion while supporting legacy phone operation. Midspan
operation is supported with built-in 2-event classification
and backoff timing. Host communication is via a 1MHz
I2C serial interface.
The LTC4266 is available in a 5mm × 7mm QFN package
that significantly reduces board space compared with
competing solutions. A legacy-compatible 36-pin SSOP
package is also available.
n Four Independent PSE Channels
n Compliant with IEEE 802.3at Type 1 and 2
n 0.34Ω Total Channel Resistance
130mW/Port at 600mA
n Advanced Power Management
8-Bit Programmable Current Limit (ILIM)
7-Bit Programmable Overload Currents (ICUT)
Fast Shutdown of Preselected Ports
14.5-Bit Port Current/Voltage Monitoring
2-Event Classification
n Very High Reliability 4-Point PD Detection:
2-Point Forced Voltage
2-Point Forced Current
n High Capacitance Legacy Device Detection
n LTC4259A-1 and LTC4258 Pin and SW Compatible
n 1MHz I2C Compatible Serial Control Interface
n Midspan Backoff Timer
n Supports Proprietary Power Levels Above 25W
n Available in 38-Pin 5mm × 7mm QFN and 36-Pin
SSOP Packages
typicAl ApplicAtion
Complete 4-Port Ethernet High Power Source
L, LT, LTC, LTM, Linear Technology and the Linear logo are registered trademarks and
ThinSOT is a trademark of Linear Technology Corporation. All other trademarks are the property
of their respective owners.
4266 TA01
SCL
SDAIN
SDAOUT
AD0
AD1
AD2
AD3 AGND VEE SENSE1
–50V
1µFSMAJ58A
GATE1 OUT1 OUT2 OUT3SENSE2 GATE2 SENSE3 GATE3 SENSE4 GATE4 OUT4
PORT1
–50V
0.22µF 100V
s4
S1B
s4
S1B
s4
PORT2
PORT3
PORT4
DGND
VDD
INT SHDN1 SHDN2 AUTO MSD RESETSHDN3 SHDN4 MID
0.1µF
3.3V
LTC4266
LTC4266
2
4266fa
Absolute MAxiMuM rAtings
pin conFigurAtion
orDer inForMAtion
LEAD FREE FINISH TAPE AND REEL PART MARKING* PACKAGE DESCRIPTION TEMPERATURE RANGE
LTC4266CGW#PBF LTC4266CGW#TRPBF LTC4266 36-Lead Plastic Wide SSOP 0°C to 70°C
LTC4266IGW#PBF LTC4266IGW#TRPBF LTC4266 36-Lead Plastic Wide SSOP –40°C to 85°C
LTC4266CUHF#PBF LTC4266CUHF#TRPBF 4266 38-Lead (5mm × 7mm) Plastic QFN 0°C to 70°C
LTC4266IUHF#PBF LTC4266IUHF#TRPBF 4266 38-Lead (5mm × 7mm) Plastic QFN –40°C to 85°C
Consult LTC Marketing for parts specified with wider operating temperature ranges. *The temperature grade is identified by a label on the shipping container.
Consult LTC Marketing for information on non-standard lead based finish parts.
For more information on lead free part marking, go to: http://www.linear.com/leadfree/
For more information on tape and reel specifications, go to: http://www.linear.com/tapeandreel/
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
TOP VIEW
GW36 PACKAGE
36-LEAD PLASTIC WIDE SSOP
TJMAX = 125°C, QJA = 80°C/W
36
35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
20
19
RESET
MID
INT
SCL
SDAOUT
SDAIN
AD3
AD2
AD1
AD0
NC
NC
NC
NC
DGND
VDD
SHDN1
SHDN2
MSD
AUTO
OUT1
GATE1
SENSE1
OUT2
GATE2
SENSE2
VEE
OUT3
GATE3
SENSE3
OUT4
GATE4
SENSE4
AGND
SHDN4
SHDN3
13 14 15 16
TOP VIEW
39
UHF PACKAGE
38-LEAD (5mm s 7mm) PLASTIC QFN
EXPOSED PAD IS VEE (PIN 39) MUST BE SOLDERED TO PCB
TJMAX = 125°C, QJA = 34°C/W
17 18 19
38 37 36 35 34 33 32
24
25
26
27
28
29
30
31
8
7
6
5
4
3
2
1SDAOUT
NC
SDAIN
AD3
AD2
AD1
AD0
DNC
NC
DGND
NC
NC
GATE1
SENSE1
OUT2
GATE2
SENSE2
VEE
VEE
OUT3
GATE3
SENSE3
OUT4
GATE4
SCL
INT
MID
RESET
MSD
AUTO
OUT1
VDD
SHDN1
SHDN2
SHDN3
SHDN4
AGND
SENSE4
23
22
21
20
9
10
11
12
Supply Voltages (Note 1)
AGND – VEE ............................................. 0.3V to 80V
DGND – VEE ................................................. 0.3V to 80V
VDD – DGND .............................................. 0.3V to 5.5V
Digital Pins
SCL, SDAIN, SDAOUT, INT, SHDNn, MSD, ADn,
RESET, AUTO, MID........... DGND –0.3V to VDD + 0.3V
Analog Pins
GATEn, SENSEn, OUTn .......... VEE0.3V to VEE + 80V
Operating Temperature Range
LTC4266C ................................................ 0°C to 70°C
LTC4266I .............................................40°C to 85°C
Junction Temperature (Note 2) ............................. 125°C
Storage Temperature Range .................. 65°C to 150°C
Lead Temperature (Soldering, 10 sec)...................300°C
LTC4266
3
4266fa
electricAl chArActeristics
SYMBOL PARAMETER CONDITIONS MIN TYP MAX UNITS
–48V Supply Voltage AGND – VEE
For IEEE Type 1 Complaint Output
For IEEE Type 2 Complaint Output
l
l
45
51
57
57
V
V
Undervoltage Lock-out Level l20 25 30 V
VDD VDD Supply Voltage VDD – DGND l3.0 3.3 4.3 V
Undervoltage Lock-out l2.2 V
Allowable Digital Ground Offset DGND – VEE l25 57 V
IEE VEE Supply Current (AGND – VEE) = 55V l–2.4 –5 mA
IDD VDD Supply Current (VDD – DGND) = 3.3V l1.1 3 mA
Detection
Detection Current – Force Current First Point, AGND – VOUTn = 9V
Second Point, AGND – VOUTn = 3.5V
l
l
220
140
240
160
260
180
µA
µA
Detection Voltage – Force Voltage AGND – VOUTn, 5µA ≤ IOUTn ≤ 500µA
First Point
Second Point
l
l
7
3
8
4
9
5
V
V
Detection Current Compliance AGND – VOUTn = 0V l0.8 0.9 mA
VOC Detection Voltage Compliance AGND – VOUTn, Open Port l10.4 12 V
Detection Voltage Slew Rate AGND – VOUTn, CPORT = 0.15µF l0.01 V/µs
Min. Valid Signature Resistance l15.5 17 18.5
Max. Valid Signature Resistance l27.5 29.7 32
Classification
VCLASS Classification Voltage AGND – VOUTn, 0mA ≤ ICLASS ≤ 50mA l16.0 20.5 V
Classification Current Compliance VOUTn = AGND l53 61 67 mA
Classification Threshold Current Class 0 – 1
Class 1 – 2
Class 2 – 3
Class 3 – 4
Class 4 – Overcurrent
l
l
l
l
l
5.5
13.5
21.5
31.5
45.2
6.5
14.5
23
33
48
7.5
15.5
24.5
34.9
50.8
mA
mA
mA
mA
mA
VMARK Classification Mark State Voltage AGND – VOUTn, 0.1mA ≤ ICLASS ≤ 10mA l7.5 9 10 V
Mark State Current Compliance VOUTn = AGND l53 61 67 mA
Gate Driver
GATE Pin Pull-Down Current Port Off, VGATEn = VEE + 5V
Port Off, VGATEn = VEE + 1V
l
l
0.4
0.08
0.12
mA
mA
GATE Pin Fast Pull-Down Current VGATEn = VEE + 5V 30 mA
GATE Pin On Voltage VGATEn – VEE, IGATEn = 1µA l8 14 V
Output Voltage Sense
VPG Power Good Threshold Voltage VOUTn – VEE l2 2.4 2.8 V
OUT Pin Pull-Up Resistance to AGND 0V ≤ (AGND – VOUTn) ≤ 5V l300 500 700
The l denotes the specifications which apply over the full operating
temperature range, otherwise specifications are at TA = 25°C. AGND – VEE = 54V, AGND = DGND, and VDD – DGND = 3.3V unless otherwise
noted. (Notes 3, 4)
LTC4266
4
4266fa
SYMBOL PARAMETER CONDITIONS MIN TYP MAX UNITS
Current Sense
VCUT Overcurrent Sense Voltage VSENSEn – VEE, icut12 = icut34 = hpen = 00h
hpen = 0Fh, cutn[5:0] ≥ 4 (Note 12)
cutrng = 0
cutrng = 1
l
l
l
180
9
4.5
188
9.38
4.69
196
9.75
4.88
mV
mV/LSB
mV/LSB
Overcurrent Sense in Auto Mode Class 0, Class 3
Class 1
Class 2
Class 4
l
l
l
l
90
26
49
152
94
28
52
159
98
30
55
166
mV
mV
mV
mV
VLIM Active Current Limit in 802.3af Compliant
Mode
VSENSEn – VEE, icut12 = icut34 = hpen = 00h
VEE = 55V (Note 12)
VEE < VOUT < AGND – 29V
AGND – VOUT = 0V
l
l
204
40
212
220
100
mV
mV
VLIM Active Current Limit in High Power Mode hpen = 0Fh, limn = C0h, VEE = 55V
VOUT – VEE = 0V to 10V
VEE + 23V < VOUT < AGND – 29V
AGND – VOUT = 0V
l
l
l
204
100
20
212
106
221
113
50
mV
mV
mV
VLIM Active Current Limit in Auto Mode VOUT – VEE = 0V to 10V, VEE = 55V
Class 0 to Class 3
Class 4
l
l
102
204
106
212
110
221
mV
mV
VMIN DC Disconnect Sense Voltage VSENSEn – VEE, rdis = 0
VSENSEn – VEE, rdis = 1
l
l
2.6
1.3
3.8
1.9
4.8
2.41
mV
mV
VSC Short-Circuit Sense VSENSEn – VEE – VLIM, rdis = 0
VSENSEn – VEE – VLIM, rdis = 1
l
l
160
75
200
100
255
135
mV
mV
Port Current ReadBack
Resolution No missing codes, fast_iv = 0 14 bits
LSB Weight VSENSEn – VEE 30.5 µV/LSB
50-60Hz Noise Rejection (Note 7) 30 dB
Port Voltage ReadBack
Resolution No missing codes, fast_iv = 0 14 bits
LSB Weight AGND – VOUTn 5.835 mV/LSB
50-60Hz noise rejection (Note 7) 30 dB
Digital Interface
VILD Digital Input Low Voltage (Note 6) l0.8 V
VIHD Digital Input High Voltage (Note 6) l2.2 V
Digital Output Low Voltage ISDAOUT = 3mA, IINT = 3mA
ISDAOUT = 5mA, IINT = 5mA
l
l
0.4
0.7
V
V
Internal Pull-Up to VDD ADn, SHDNn, RESET, MSD 50
Internal Pull-Down to DGND AUTO, MID 50
electricAl chArActeristics
The l denotes the specifications which apply over the full operating
temperature range, otherwise specifications are at TA = 25°C. AGND – VEE = 54V, AGND = DGND, and VDD – DGND = 3.3V unless otherwise
noted. (Notes 3, 4)
LTC4266
5
4266fa
SYMBOL PARAMETER CONDITIONS MIN TYP MAX UNITS
Timing Characteristics
tDET Detection Time Beginning to End of Detection (Note 7) l270 290 310 ms
tDETDLY Detection Delay From PD Connected to Port to Detection
Complete (Note 7)
l300 470 ms
tCLE1 First Class Event Duration (Note 7) l11 12 13 ms
tME1 First Mark Event Duration (Notes 7, 11) l6.8 8.6 10.3 ms
tCLE2 Second Class Event Duration (Note 7) l11 12 13 ms
tME2 Second Mark Event Duration (Note 7) l19 22 ms
tCLE3 Third Class Event Duration CPORT = 0.6µF (Note 7) l0.1 ms
tPON Power On Delay in Auto Mode From End of Valid Detect to Application of
Power to Port (Note 7)
l60 ms
Turn On Rise Time (AGND VOUT): 10% to 90% of (AGND VEE),
CPORT = 0.15µF (Note 7)
l15 24 µs
Turn On Ramp Rate CPORT = 0.15µF (Note 7) l10 V/µs
Fault Delay From ICUT Fault to Next Detect l1.0 1.1 s
Midspan Mode Detection Backoff Rport = 15.5kΩ (Note 7) l2.3 2.5 2.7 s
Power Removal Detection Delay From Power Removal After tDIS to Next
Detect (Note 7)
l1.0 1.3 2.5 s
tSTART Maximum Current Limit Duration During Port
Start-Up
tSTART1 = 0, tSTART0 = 0 (Notes 7, 12) l52 62.5 66 ms
tLIM, tICUT Maximum Current Limit Duration After Port
Start-Up
tICUT1 = 0, tICUT0 = 0 (Notes 7, 12) l52 62.5 66 ms
Maximum Current Limit Duty Cycle (Note 7) l5.8 6.3 6.7 %
tMPS Maintain Power Signature (MPS) Pulse Width
Sensitivity
Current Pulse Width to Reset Disconnect
Timer (Notes 7, 8)
l1.6 3.6 ms
tDIS Maintain Power Signature (MPS) Dropout
Time
tconf [1:0] = 00b (Notes 5, 12) l320 350 380 ms
tMSD Masked Shut Down Delay (Note 7) l6.5 µs
tSHDN Port Shut Down Delay (Note 7) l6.5 µs
I2C Watchdog Timer Duration l1.5 2 3 s
Minimum Pulse Width for Masked Shut
Down
(Note 7) l3 µs
Minimum Pulse Width for SHDN (Note 7) l3 µs
Minimum Pulse Width for RESET (Note 7) l4.5 µs
electricAl chArActeristics
The l denotes the specifications which apply over the full operating
temperature range, otherwise specifications are at TA = 25°C. AGND – VEE = 54V, AGND = DGND, and VDD – DGND = 3.3V unless otherwise
noted. (Notes 3, 4)
LTC4266
6
4266fa
SYMBOL PARAMETER CONDITIONS MIN TYP MAX UNITS
I2C Timing
Clock Frequency (Note 7) l1 MHz
t1Bus Free Time Figure 5 (Notes 7, 9) l480 ns
t2Start Hold Time Figure 5 (Notes 7, 9) l240 ns
t3SCL Low Time Figure 5 (Notes 7, 9) l480 ns
t4SCL High Time Figure 5 (Notes 7, 9) l240 ns
t5Data Hold Time Figure 5 (Notes 7, 9) Data into chip
Data out of chip
l
l
60
120
ns
ns
t6Data Set-Up Time Figure 5 (Notes 7, 9) l80 ns
t7Start Set-Up Time Figure 5 (Notes 7, 9) l240 ns
t8Stop Set-Up Time Figure 5 (Notes 7, 9) l240 ns
trSCL, SDAIN Rise Time Figure 5 (Notes 7, 9) l120 ns
tfSCL, SDAIN Fall Time Figure 5 (Notes 7, 9) l60 ns
Fault Present to INT Pin Low (Notes 7, 9, 10) l150 ns
Stop Condition to INT Pin Low (Notes 7, 9, 10) l1.5 µs
ARA to INT Pin High Time (Notes 7, 9) l1.5 µs
SCL Fall to ACK Low (Notes 7, 9) l120 ns
Note 1: Stresses beyond those listed under Absolute Maximum Ratings
may cause permanent damage to the device. Exposure to any Absolute
Maximum Rating condition for extended periods may affect device
reliability and lifetime.
Note 2: This IC includes overtemperature protection that is intended
to protect the device during momentary overload conditions. Junction
temperature will exceed 140°C when overtemperature protection is active.
Continuous operation above the specified maximum operating junction
temperature may impair device reliability.
Note 3: All currents into device pins are positive; all currents out of device
pins are negative.
Note 4: The LTC4266 operates with a negative supply voltage (with respect
to ground). To avoid confusion, voltages in this data sheet are referred to
in terms of absolute magnitude.
Note 5: tDIS is the same as tMPDO defined by IEEE 802.3at.
Note 6: The LTC4266 digital interface operates with respect to DGND. All
logic levels are measured with respect to DGND.
Note 7: Guaranteed by design, not subject to test.
Note 8: The IEEE 802.3af specification allows a PD to present its
Maintain Power Signature (MPS) on an intermittent basis without being
disconnected. In order to stay powered, the PD must present the MPS for
tMPS within any tMPDO time window.
Note 9: Values measured at VILD(MAX) and VIHD(MIN).
Note 10: If fault condition occurs during an I2C transaction, the INT pin
will not be pulled down until a stop condition is present on the I2C bus.
Note 11: Load Characteristic of the LTC4266 during Mark:
7V < (AGND – VOUTn) < 10V or IOUT < 50µA
Note 12: See the LTC4266 Software Programming documentation for
information on serial bus usage and device configuration and status
registers.
electricAl chArActeristics
The l denotes the specifications which apply over the full operating
temperature range, otherwise specifications are at TA = 25°C. AGND – VEE = 54V, AGND = DGND, and VDD – DGND = 3.3V unless otherwise
noted. (Notes 3, 4)
LTC4266
7
4266fa
typicAl perForMAnce chArActeristics
Power On Sequence in Auto Mode Powering Up into a 180µF Load
802.3af Classification in
Auto Mode
2-Event Classification in
Auto Mode
Classification Transient Response
to 40mA Load Step Classification Current Compliance
VDD Supply Current vs Voltage
100ms/DIV
–70
–60
PORT VOLTAGE (V)
10
0
–10
–20
–30
–40
–50
4266 G01
PORT 1
VDD = 3.3V
VEE = –54V
FORCED CURRENT DETECTION
FORCED VOLTAGE
DETECTION
802.3af
CLASSIFICATION
POWER ON
GND
VEE
5ms/DIV
GND
0mA
4266 G02
VEE
VEE
GATE
VOLTAGE
10V/DIV
PORT
CURRENT
200 mA/DIV
PORT
VOLTAGE
20V/DIV
FOLDBACK
FET ON
425mA
CURRENT LIMIT
LOAD
FULLY
CHARGED
VDD = 3.3V
VEE = –54V
5ms/DIV
VEE
–18.4
PORT
VOLTAGE
10V/DIV
GND
4266 G03
PORT 1
VDD = 3.3V
VEE = –55V
PD IS CLASS 1
10ms/DIV
VEE
–17.6
PORT
VOLTAGE
10V/DIV
GND
4266 G04
PORT 1
VDD = 3.3V
VEE = –55V
PD IS CLASS 4
1ST CLASS EVENT
2ND CLASS EVENT
50µs/DIV
40mA
0mA
4266 G05
–20V
PORT
VOLTAGE
1V/DIV
PORT
CURRENT
20mA/DIV
VDD = 3.3V
VEE = –54V
CLASSIFICATION CURRENT
–20
CLASSIFICATION VOLTAGE (V)
–18
–16
–14
–12
–10
–8
–6
–4
–2
0
0 10 20 30
4266 G06
40 50 60 70
VDD = 3.3V
VEE = –54V
TA = 25°C
VEE Supply Current vs Voltage
802.3at ILIM Threshold vs
Temperature
VEE SUPPLY VOLTAGE (V)
–60
2.0
IEE SUPPLY CURRENT (mA)
2.1
2.2
2.3
2.4
–55 –50 –45 –40
4266 G08
–35 –30 –25 –20
–40°C
25°C
85°C
TEMPERATURE (°C)
–40
210
VLIM (mV)
ILIM (mA)
211
213
212
214
215
840
844
852
848
856
860
0 40
4266 G09
–80 120
VDD = 3.3V
VEE = –54V
RSENSE = 0.25Ω
REG 48h = C0h
LTC4266
8
4266fa
typicAl perForMAnce chArActeristics
802.3af ILIM Threshold vs
Temperature
DC Disconnect Threshold vs
Temperature
Current Limit Foldback
ADC Noise Histogram
Current Readback in Fast Mode
ADC Integral Nonlinearity
Current Readback in Fast Mode
TEMPERATURE (°C)
–40
105.00
VLIM (mV)
ILIM (mA)
106.50
105.75
107.25
108.00
420
423
426
429
432
0 40
4266 G10
80 120
VDD = 3.3V
VEE = –54V
RSENSE = 0.25Ω
REG 48h = 80h
PORT 1
802.3at ICUT Threshold vs
Temperature
802.3af ICUT Threshold vs
Temperature
TEMPERATURE (°C)
–40
158
VCUT (mV)
ICUT (mA)
161
160
159
162
163
630
636
640
648
644
652
0 40
4266 G11
80 120
VDD = 3.3V
VEE = –54V
RSENSE = 0.25Ω
REG 47h = E2h
PORT 1
TEMPERATURE (°C)
–40
93.00
VCUT (mV)
ICUT (mA)
94.50
93.75
95.25
96.00
372
375
378
381
384
0 40
4266 G12
80 120
VDD = 3.3V
VEE = –54V
RSENSE = 0.25Ω
REG 47h = D4h
PORT 1
TEMPERATURE (°C)
–40
VMIN (mV)
IMIN (mV)
7.00
7.50
7.25
7.75
8.00
1.7500
1.8125
1.8750
1.9375
2.0000
0 40
4266 G13
80 120
VDD = 3.3V
VEE = –54V
RSENSE = 0.25Ω
REG 47h = E2h
PORT 1
VOUTn (V)
–54
0
ILIM (mA)
VLIM (mV)
900
800
700
600
500
400
300
200
100
0
225
200
175
150
125
100
75
50
25
–36–45 –27
4266 G14
–18 –9 0
VDD = 3.3V
VEE = –54V
RSENSE = 0.25Ω
REG 48h = C0h
ADC OUTPUT
191
0
BIN COUNT
400
350
300
250
200
150
100
50
193192 194
4266 G15
195 196
VSENSEn – VEE = 110.4mV
CURRENT SENSE RESISTOR INPUT VOLTAGE (mV)
0
ADC INTEGRAL NONLINEARITY (LSBs)
0
0.5
400
4266 G16
–0.5
–1.0 100 200 250 500
1.0
300
50 150 450
350
LTC4266
9
4266fa
typicAl perForMAnce chArActeristics
ADC Noise Histogram
Current Readback in Slow Mode
ADC Integral Nonlinearity
Current Readback in Slow Mode
ADC Noise Histogram Port
Voltage Readback in Fast Mode
ADC Integral Nonlinearity
Voltage Readback in Fast Mode
ADC Noise Histogram Port
Voltage Readback in Slow Mode
ADC Integral Nonlinearity
Voltage Readback in Slow Mode
INT and SDAOUT Pull Down Voltage
vs Load Current
MOSFET Gate Drive With Fast
Pull Down
ADC OUTPUT
0
BIN COUNT
300
250
200
150
100
50
6139 6141
4266 G17
6143 6145 6147
VSENSEn – VEE = 110.4mV
ADC OUTPUT
260
0
BIN COUNT
600
500
400
300
200
100
262261 263
4266 G19
264 265
AGND – VOUTn = 48.3V
ADC OUTPUT
8532
0
BIN COUNT
600
500
400
300
200
100
85348533 8535
4266 G21
8536
AGND – VOUTn = 48.3V
100µs/DIV
GND
0mA
4266 G24
VEE
VEE
PORT
CURRENT
500mA/DIV
GATE
VOLTAGE
10V/DIV
PORT
VOLTAGE
20V/DIV
CURRENT LIMIT
50Ω FAULT REMOVED
50Ω
FAULT
APPLIED
VDD = 3.3V
VEE = –54V
FAST PULL DOWN
CURRENT SENSE RESISTOR INPUT VOLTAGE (mV)
0
ADC INTEGRAL NONLINEARITY (LSBs)
0
0.5
400
4266 G18
–0.5
–1.0 100 200 250 500
1.0
300
50 150 450
350
PORT VOLTAGE (V)
0
ADC INTEGRAL NONLINEARITY (LSBs)
0
0.5
50
4266 G22
–0.5
–1.0 20 30 60
1.0
40
10
LTC4266
10
4266fa
test tiMing DiAgrAMs
Figure 1. Detect, Class and Turn-On Timing in Auto or Semiauto Modes
Figure 2. Current Limit Timing Figure 3. DC Disconnect Timing
Figure 4. Shut Down Delay Timing Figure 5. I2C Interface Timing
VLIM VCUT
0V
VSENSEn TO VEE
INT
4266 F02
tSTART, tICUT
SCL
SDA
t1
t2
t3tr
tf
t5t6t7t8
t4
4266 F05
VMIN
VSENSEn
TO VEE
INT
tDIS
tMPS 4266 F03
VPORTn
INT
tDETDLY
VOC
VEE
tDET
tME1
tME2
VMARK
VCLASS
15.5V
20.5V
tCLE1
tCLE2
tCLE3
PD
CONNECTED
0V
4266 F01
FORCED-CURRENT
CLASSIFICATION
tPON
FORCED-
VOLTAGE
VGATEn
VEE
MSD or
SHDNn
tSHDN
tMSD
4266 F04
LTC4266
11
4266fa
i2c tiMing DiAgrAMs
Figure 6. Writing to a Register
Figure 7. Reading from a Register
SCL
SDA
4266 F06
0 01 AD3 AD2 AD1 AD0 A7 A6 A5 A4 A3 A2 A1 A0
R/W ACK D7 D6 D5 D4 D3 D2 D1 D0
ACK ACK
START BY
MASTER
ACK BY
SLAVE
ACK BY
SLAVE
ACK BY
SLAVE
FRAME 1
SERIAL BUS ADDRESS BYTE
FRAME 2
REGISTER ADDRESS BYTE
FRAME 3
DATA BYTE
STOP BY
MASTER
SCL
SDA 0 01 AD3 AD2 AD1 AD0 A7 A6 A5 A4 A3 A2 A1 A0
R/W ACK ACK 0 01 AD3 AD2 AD1 AD0 D7 D6 D5 D4 D3 D2 D1 D0
R/W ACK ACK
START BY
MASTER
ACK BY
SLAVE
ACK BY
SLAVE
4266 F07
STOP BY
MASTER
REPEATED
START BY
MASTER
ACK BY
SLAVE
NO ACK BY
MASTER
FRAME 1
SERIAL BUS ADDRESS BYTE
FRAME 2
REGISTER ADDRESS BYTE
FRAME 1
SERIAL BUS ADDRESS BYTE
FRAME 2
DATA BYTE
LTC4266
12
4266fa
Figure 8. Reading the Interrupt Register (Short Form)
Figure 9. Reading from Alert Response Address
SCL
SDA
4266 F08
0 1 0AD3 AD2 AD1 AD0 D7 D6 D5 D4 D3 D2 D1 D0
R/W ACK ACK
START BY
MASTER
ACK BY
SLAVE
NO ACK BY
MASTER
FRAME 1
SERIAL BUS ADDRESS BYTE
FRAME 2
DATA BYTE
STOP BY
MASTER
SCL
SDA
4266 F09
0 0 110AD30000 1 AD2 AD1 AD0
R/W ACK ACK1
START BY
MASTER
ACK BY
SLAVE
NO ACK BY
MASTER
FRAME 1
ALERT RESPONSE ADDRESS BYTE
FRAME 2
SERIAL BUS ADDRESS BYTE
STOP BY
MASTER
i2c tiMing DiAgrAMs
LTC4266
13
4266fa
pin Functions
RESET: Chip Reset, Active Low. When the RESET pin is
low, the LTC4266 is held inactive with all ports off and all
internal registers reset to their power-up states. When
RESET is pulled high, the LTC4266 begins normal opera-
tion. RESET can be connected to an external capacitor
or RC network to provide a power turn-on delay. Internal
filtering of the RESET pin prevents glitches less than 1µs
wide from resetting the LTC4266. Internally pulled up to
VDD.
MID: Midspan Mode Input. When high, the LTC4266 acts
as a midspan device. Internally pulled down to DGND.
INT: Interrupt Output, Open Drain. INT will pull low when
any one of several events occur in the LTC4266. It will
return to a high impedance state when bits 6 or 7 are set
in the Reset PB register (1Ah). The INT signal can be used
to generate an interrupt to the host processor, eliminating
the need for continuous software polling. Individual INT
events can be disabled using the Int Mask register (01h).
See Register Functions and Applications Information for
more information. The INT pin is only updated between
I2C transactions.
SCL: Serial Clock Input. High impedance clock input for
the I2C serial interface bus. SCL must be tied high if not
used.
SDAOUT: Serial Data Output, Open Drain Data Output for
the I2C Serial Interface Bus. The LTC4266 uses two pins
to implement the bidirectional SDA function to simplify
optoisolation of the I2C bus. To implement a standard
bidirectional SDA pin, tie SDAOUT and SDAIN together.
SDAOUT should be grounded or left floating if not used.
See Applications Information for more information.
SDAIN: Serial Data Input. High impedance data input for
the I2C serial interface bus. The LTC4266 uses two pins
to implement the bidirectional SDA function to simplify
optoisolation of the I2C bus. To implement a standard
bidirectional SDA pin, tie SDAOUT and SDAIN together.
SDAIN must be tied high if not used. See Applications
Information for more information.
AD3: Address Bit 3. Tie the address pins high or low to
set the I2C serial address to which the LTC4266 responds.
This address will be 010A3A2A1A0b. Internally pulled up
to VDD.
AD2: Address Bit 2. See AD3.
AD1: Address Bit 1. See AD3.
AD0: Address Bit 0. See AD3.
NC, DNC: All pins identified with “NC” or “DNC” must be
left unconnected.
DGND: Digital Ground. DGND is the return for the VDD
supply.
VDD: Logic Power Supply. Connect to a 3.3V power supply
relative to DGND. VDD must be bypassed to DGND near
the LTC4266 with at least a 0.1µF capacitor.
SHDN1: Shutdown Port 1, Active Low. When pulled low,
SHDN1 shuts down port 1, regardless of the state of the
internal registers. Pulling SHDN1 low is equivalent to set-
ting the Reset Port 1 bit in the Reset Pushbutton register
(1Ah). Internal filtering of the SHDN1 pin prevents glitches
less than 1µs wide from reseting the port. Internally pulled
up to VDD.
SHDN2: Shutdown Port 2, Active Low. See SHDN1.
SHDN3: Shutdown Port 3, Active Low. See SHDN1.
SHDN4: Shutdown Port 4, Active Low. See SHDN1.
AGND: Analog Ground. AGND is the return for the VEE
supply.
SENSE4: Port 4 Current Sense Input. SENSE4 monitors
the external MOSFET current via a 0.5Ω or 0.25Ω sense
resistor between SENSE4 and VEE. Whenever the voltage
across the sense resistor exceeds the overcurrent detection
threshold VCUT, the current limit fault timer counts up. If
the voltage across the sense resistor reaches the current
limit threshold VLIM, the GATE4 pin voltage is lowered to
maintain constant current in the external MOSFET. See
Applications Information for further details. If the port is
unused, the SENSE4 pin must be tied to VEE.
LTC4266
14
4266fa
pin Functions
GATE4: Port 4 Gate Drive. GATE4 should be connected
to the gate of the external MOSFET for port 4. When the
MOSFET is turned on, the gate voltage is driven to 13V
(typ) above VEE. During a current limit condition, the
voltage at GATE4 will be reduced to maintain constant
current through the external MOSFET. If the fault timer
expires, GATE4 is pulled down, turning the MOSFET off
and recording a tCUT or tSTART event. If the port is unused,
float the GATE4 pin.
OUT4: Port 4 Output Voltage Monitor. OUT4 should be
connected to the output port. A current limit foldback
circuit limits the power dissipation in the external MOSFET
by reducing the current limit threshold when the drain-to-
source voltage exceeds 10V. The port 4 Power Good bit is
set when the voltage from OUT4 to VEE drops below 2.4V
(typ). A 500k resistor is connected internally from OUT4
to AGND when the port is idle. If the port is unused, OUT4
pin must be floated.
SENSE3: Port 3 Current Sense Input. See SENSE4.
GATE3: Port 3 Gate Drive. See GATE4.
OUT3: Port 3 Output Voltage Monitor. See OUT4.
VEE: Main Supply Input. Connect to a –45V to –57V
supply, relative to AGND.
SENSE2: Port 2 Current Sense Input. See SENSE4.
GATE2: Port 2 Gate Drive. See GATE4.
OUT2: Port 2 Output Voltage Monitor. See OUT4.
SENSE1: Port 1 Current Sense Input. See SENSE4.
GATE1: Port 1 Gate Drive. See GATE 4.
OUT1: Port 1 Output Voltage Monitor. See OUT4.
AUTO: Auto Mode Input. Auto mode allows the LTC4266
to detect and power up a PD even if there is no host con-
troller present on the I2C bus. The voltage of the AUTO
pin determines the state of the internal registers when
the LTC4266 is reset or comes out of VDD UVLO (see the
Register map). The states of these register bits can sub-
sequently be changed via the I2C interface. The real-time
state of the AUTO pin is read at bit 0 in the Pin Status
register (11h). Internally pulled down to DGND. Must be
tied locally to either VDD or DGND.
MSD: Maskable Shutdown Input. Active low. When pulled
low, all ports that have their corresponding mask bit set
in the mconfig register (17h) will be reset, equivalent to
pulling the SHDN pin low. Internal filtering of the MSD pin
prevents glitches less than 1µs wide from resetting ports.
Internally pulled up to VDD.
LTC4266
15
4266fa
operAtion
Figure 10. Power Over Ethernet System Diagram
4266 F10
S1B
S1B
SMAJ58A
0.22µF
100V
X7R
1µF
100V
X7R
Tx
Rx
Rx
Tx
SMAJ58A
58V
DATA PAIR
DATA PAIR
VEE SENSE GATE OUT
VDD
INT
SCL
SDAIN
SDAOUT
0.25Ω
IRFM120A
SPARE PAIR
SPARE PAIR
1/4
LTC4266
DGND AGND
I2C
3.3V
INTERRUPT
–48V
CAT 5
20Ω MAX
ROUNDTRIP
0.05µF MAX
RJ45
4
5
4
5
1
2
1
2
3
6
3
6
7
8
7
8
RJ45
1N4002
s4
1N4002
s4
PSE PD
RCLASS
–48VIN
PWRGD
–48VOUT
LTC4265
GND
DC/DC
CONVERTER
5µF ≤ CIN
≤ 300µF
+
VOUT
GND
0.1µF
Overview
Power over Ethernet, or PoE, is a standard protocol for
sending DC power over copper Ethernet data wiring.
The IEEE group that administers the 802.3 Ethernet data
standards added PoE powering capability in 2003. This
original PoE spec, known as 802.3af, allowed for 48V DC
power at up to 13W. This initial spec was widely popular,
but 13W was not adequate for some requirements. In
2009, the IEEE released a new standard, known as 802.3at
or PoE+, increasing the voltage and current requirements
to provide 25W of power.
The IEEE standard also defines PoE terminology. A device
that provides power to the network is known as a PSE, or
power sourcing equipment, while a device that draws power
from the network is known as a PD, or powered device.
PSEs come in two types: Endpoints (typically network
switches or routers), which provide data and power; and
Midspans, which provide power but pass through data.
Midspans are typically used to add PoE capability to existing
non-PoE networks. PDs are typically IP phones, wireless
access points, security cameras, and similar devices, but
could be nearly anything that runs from 25W or less and
includes an RJ45-style network connector.
The LTC4266 is a third-generation quad PSE controller
that implements four PSE ports in either an endpoint or
midspan design. Virtually all necessary circuitry is included
to implement a IEEE 802.3at compliant PSE design, requir-
ing only an external power MOSFET and sense resistor per
channel; these minimize power loss compared to alterna-
tive designs with on-board MOSFETs and increase system
reliability in the event a single channel is damaged.
PoE Basics
Common Ethernet data connections consist of two or four
twisted pairs of copper wire (commonly known as CAT-5
cable), transformer-coupled at each end to avoid ground
loops. PoE systems take advantage of this coupling ar-
rangement by applying voltage between the center-taps
of the data transformers to transmit power from the PSE
to the PD without affecting data transmission. Figure 10
shows a high-level PoE system schematic.
To avoid damaging legacy data equipment that does not
expect to see DC voltage, the PoE spec defines a protocol
that determines when the PSE may apply and remove
power. Valid PDs are required to have a specific 25kΩ
common-mode resistance at their input. When such a PD
is connected to the cable, the PSE detects this signature
resistance and turns on the power. When the PD is later
disconnected, the PSE senses the open circuit and turns
power off. The PSE also turns off power in the event of a
current fault or short circuit.
When a PD is detected, the PSE optionally looks for a
classification signature that tells the PSE the maximum
power the PD will draw. The PSE can use this information
to allocate power among several ports, police the current
consumption of the PD, or to reject a PD that will draw
LTC4266
16
4266fa
operAtion
more power that the PSE has available. The classification
step is optional; if a PSE chooses not to classify a PD, it
must assume that the PD is a 13W (full 802.3af power)
device.
New in 802.3at
The newer 802.3at standard supersedes 802.3af and brings
several new features:
A PD may draw as much as 25.5W. Such PDs (and the
PSEs that support them) are known as Type 2. Older
13W 802.3af equipment is classified as Type 1. Type 1
PDs will work with all PSEs; Type 2 PDs may require
Type 2 PSEs to work properly. The LTC4266 is designed
to work in both Type 1 and Type 2 PSE designs, and
also supports non-standard configurations at higher
power levels.
The Classification protocol is expanded to allow Type 2
PSEs to detect Type 2 PDs, and to allow Type 2 PDs to
determine if they are connected to a Type 2 PSE. Two
versions of the new Classification protocol are avail-
able: an expanded version of the 802.3af Class Pulse
protocol, and an alternate method integrated with the
existing LLDP protocol (using the Ethernet data path).
The LTC4266 fully supports the new Class Pulse protocol
and is also compatible with the LLDP protocol (which
is implemented in the data communications layer, not
in the PoE circuitry).
Fault protection current levels and timing are adjusted
to reduce peak power in the MOSFET during a fault;
this allows the new 25.5W power levels to be reached
using the same MOSFETs as older 13W designs.
BACKWARDS COMPATIBILITY
The LTC4266 is designed to be backward compatible with
earlier PSE chips in both software and pin functions. Exist-
ing systems using either the LTC4258 or LTC4259A (or
compatible) devices can be substituted with the LTC4266
without software or PCB layout changes; only minor BOM
changes are required to implement a fully compliant
802.3at design.
Because of the backwards compatibility features, some of
the internal registers are redundant or unused when the
LTC4266 is operated as recommended. For more details
on usage in compatibility mode, refer to the LTC4258/
LTC4259A device datasheets.
Special Compatibility Mode Notes
The LTC4266 can use either 0.5Ω or 0.25Ω sense
resistors, while the LTC425x chips always used 0.5Ω.
To maintain compatibility, if the AUTO pin is low when
the LTC4266 powers up it assumes the sense resistor
is 0.5Ω; if it is high at power up, the LTC4266 assumes
0.25Ω. The resistor value setting can be reconfigured
at any time after power up. In particular, systems that
use 0.25Ω sense resistors and have AUTO tied low
must reconfigure the resistor settings after power up.
The LTC4259A included both AC and DC disconnect
sensing circuitry, but the LTC4266 has only DC discon-
nect sensing. For the sake of compatibility, register
bits used to enable AC disconnect in the LTC4259A are
implemented in the LTC4266, but they simply mirror
the bits used for DC disconnect.
The LTC4258 and LTC4259A required 10k resistors
between the OUTn pins and the drains of the external
MOSFETs. These resistors must be shorted or replaced
with zero ohm jumpers when using the LTC4266.
The LTC4258 and LTC4259A included a BYP pin, de-
coupled to AGND with 0.1µF. This pin changes to the MID
pin on the LTC4266. The capacitor should be removed
for Endspan applications, or replaced with a zero ohm
jumper for Midspan applications.
LTC4266
17
4266fa
ApplicAtions inForMAtion
Operating Modes
The LTC4266 includes four independent ports, each of
which can operate in one of four modes: Manual, Semi-
auto, Auto, or Shutdown.
In manual mode, the port waits for instructions from the
host system before taking any action. It runs a single
detection or classification cycle when commanded to
by the host, and reports the result in its Port Status
register. The host system can command the port to
turn on or off the power at any time.
In semi-auto mode, the port repeatedly attempts to
detect and classify any PD attached to it. It reports the
status of these attempts back to the host, and waits for
a command from the host before turning on power to
the port. The host must enable detection (and optionally
classification) for the port before detection will start.
Auto mode operates the same as Semi-auto mode except
that it will automatically turn on the power to the port
if detection is successful.
In shutdown mode, the port is disabled and will not
detect or power a PD.
Regardless of which mode it is in, the LTC4266 will remove
power automatically from any port that generates a current
limit fault. It will also automatically remove power from
any port that generates a disconnect event if disconnect
detection is enabled. The host controller may also com-
mand the port to remove power at any time.
Power-On Reset and the AUTO/MID pins
The initial LTC4266 configuration depends on the state
of the AUTO and MID pins during reset. Reset occurs at
power-up, or whenever the RESET pin is pulled low or the
global Reset All bit is set. Note that the AUTO pin is only
sampled when a reset occurs. Changing the state of AUTO
or MID after power-up will not change the port behavior
of the LTC4266 until a reset occurs.
Although typically used with a host controller, the LTC4266
can also be used in a standalone mode with no connection
to the serial interface. If there is no host present, the AUTO
pin should be tied high so that, at reset, all ports will be
configured to operate automatically. Each port will detect
and classify repeatedly until a PD is discovered, set ICUT
and ILIM according to the classification results, apply power
after successful detection, and remove power when a PD
is disconnected. Similarly, if the standalone application
is a midspan, the MID pin should be tied high to enable
correct midspan detection timing.
Table 1 shows the ICUT and ILIM values that will be
automatically set in standalone mode, based on the dis-
covered class.
Table 1. ICUT and ILIM Values in Standalone Mode
CLASS ICUT ILIM
Class 1 112mA 425mA
Class 2 206mA 425mA
Class 3 or Class 0 375mA 425mA
Class 4 638mA 850mA
The automatic setting of the ICUT and ILIM values only oc-
curs if the LTC4266 is reset with the AUTO pin high.
DETECTION
Detection Overview
To avoid damaging network devices that were not designed
to tolerate DC voltage, a PSE must determine whether the
connected device is a real PD before applying power. The
IEEE specification requires that a valid PD have a common-
mode resistance of 25kΩ ±5% at any port voltage below
10V. The PSE must accept resistances that fall between
19kΩ and 26.5kΩ, and it must reject resistances above
33kΩ or below 15kΩ (shaded regions in Figure 11). The
PSE may choose to accept or reject resistances in the
undefined areas between the must-accept and must-reject
ranges. In particular, the PSE must reject standard computer
network ports, many of which have 150Ω common-mode
termination resistors that will be damaged if power is ap-
plied to them (the black region at the left of Figure 11).
RESISTANCE
PD
PSE
10k
15k
4266 F11
19k 26.5k
26.25k23.75k
150Ω (NIC)
20k 30k
33k
Figure 11. IEEE 802.3af Signature Resistance Ranges
LTC4266
18
4266fa
ApplicAtions inForMAtion
4-Point Detection
The LTC4266 uses a 4-point detection method to discover
PDs. False-positive detections are minimized by check-
ing for signature resistance with both forced-current and
forced-voltage measurements. Initially, two test currents
are forced onto the port (via the OUTn pin) and the resulting
voltages are measured. The detection circuitry subtracts
the two V-I points to determine the resistive slope while
removing offset caused by series diodes or leakage at
the port (see Figure 12). If the forced-current detection
yields a valid signature resistance, two test voltages are
then forced onto the port and the resulting currents are
measured and subtracted. Both methods must report
valid resistances for the port to report a valid detection.
PD signature resistances between 17k and 29k (typically)
are detected as valid and reported as Detect Good in the
corresponding Port Status register. Values outside this
range, including open and short circuits, are also reported.
If the port measures less than 1V at the first forced-current
test, the detection cycle will abort and Short Circuit will be
reported. Table 2 shows the possible detection results.
Table 2. Detection Status
MEASURED PD SIGNATURE DETECTION RESULT
Incomplete or Not Yet Tested Detect Status Unknown
<2.4k Short Circuit
Capacitance > 2.7µF CPD too High
2.4k < RPD < 17k RSIG too Low
17k < RPD < 29k Detect Good
>29k RSIG too High
>50k Open Circuit
Voltage > 10V Port Voltage Outside Detect Range
Operating Modes
The port’s operating mode determines when the LTC4266
runs a detection cycle. In manual mode, the port will
idle until the host orders a detect cycle. It will then run
detection, report the results, and return to idle to wait for
another command.
In semi-auto mode, the LTC4266 autonomously polls a
port for PDs, but it will not apply power until commanded
to do so by the host. The port status register is updated
at the end of each detection cycle. If a valid signature
resistance is detected and classification is enabled, the
port will classify the PD and report that result as well.
The port will then wait for at least 100ms (or 2 seconds if
midspan mode is enabled), and will repeat the detection
cycle to ensure that the data in the port status register is
up-to-date.
If the port is in semi-auto mode and high power opera-
tion is enabled, the port will not turn on in response to
a power-on command unless the current detect result is
detect good. Any other detect result will generate a tSTART
fault if a power-on command is received. If the port is not
in high power mode, it will ignore the detection result and
apply power when commanded, maintaining backwards
compatibility with the LTC4259A.
Behavior in auto mode is similar to semi-auto; however,
after detect good is reported and the port is classified (if
classification is enabled), it is automatically powered on
without further intervention. In standalone mode, the ICUT
and ILIM thresholds are automatically set in auto mode;
see the power-on Reset and the AUTO Pin section for
more information.
The signature detection circuitry is disabled when the port
is initially powered up with the AUTO pin low, in shutdown
mode, or when the corresponding detect enable bit is
cleared.
Detection of Legacy PDs
Proprietary PDs that predate the original IEEE 802.3af
standard are commonly referred to today as legacy de-
vices. One type of legacy PD uses a large common mode
Figure 12. PD Detection
FIRST
DETECTION
POINT
SECOND
DETECTION
POINT
VALID PD
25kΩ SLOPE
275
165
CURRENT (µA)
0V-2V
OFFSET VOLTAGE
4266 F12
LTC4266
19
4266fa
ApplicAtions inForMAtion
capacitance (>10μF) as the detection signature. Note that
PDs in this range of capacitance are defined as invalid, so
a PSE that detects legacy PDs is technically noncompliant
with the IEEE spec.
The LTC4266 can be configured to detect this type of legacy
PD. Legacy detection is disabled by default, but can be
manually enabled on a per-port basis. When enabled, the
port will report detect good when it sees either a valid IEEE
PD or a high-capacitance legacy PD. With legacy mode
disabled, only valid IEEE PDs will be recognized.
CLASSIFICATION
802.3af Classification
A PD can optionally present a classification signature to
the PSE to indicate the maximum power it will draw while
operating. The IEEE specification defines this signature as
a constant current draw when the PSE port voltage is in the
VCLASS range (between 15.5V and 20.5V), with the current
level indicating one of 5 possible PD classes. Figure 14
shows a typical PD load line, starting with the slope of
the 25kΩ signature resistor below 10V, then transitioning
to the classification signature current (in this case, Class
3) in the VCLASS range. Table 3 shows the possible clas-
sification values.
Table 3. Classification Values
CLASS RESULT
Class 0 No Class Signature Present; Treat Like Class 3
Class 1 3W
Class 2 7W
Class 3 13W
Class 4 25.5W (Type 2)
If classification is enabled, the port will classify the PD
immediately after a successful detection cycle in semi-auto
or auto modes, or when commanded to in manual mode. It
measures the PD classification signature by applying 18V
for 12ms (both values typical) to the port via the OUTn pin
and measuring the resulting current; it then reports the
discovered class in the port status register. If the LTC4266
was reset with the AUTO pin high and the port is in auto
mode, it will additionally use the classification result to set
the ICUT and ILIM thresholds. See the Power-On Reset and
the AUTO/MID Pin section for more information.
The classification circuitry is disabled when the port is
initially powered up with the AUTO pin low, in shutdown
mode, or when the corresponding class enable bit is
cleared.
802.3at 2-Event Classification
The 802.3at spec defines two methods of classifying a
Type 2 PD.
One method adds extra fields to the Ethernet LLDP data
protocol; although the LTC4266 is compatible with this
classification method, it cannot perform classification
directly since it doesn’t have access to the data path.
LLDP classification requires the PSE to power the PD as
a standard 802.3af (Type 1) device. It then waits for the
host to perform LLDP communication with the PD and
update the PSE port data. The LTC4266 supports chang-
ing the ILIM and ICUT levels on the fly, allowing the host
to complete LLDP classification.
The second 802.3at classification method, known as 2-
event classification or ping-pong, is fully supported by the
LTC4266. A Type 2 PD that is requesting more than 13W
will indicate Class 4 during normal 802.3af classification.
If the LTC4266 sees Class 4, it forces the port to a speci-
fied lower voltage (called the mark voltage, typically 9V),
pauses briefly, and then re-runs classification to verify the
Class 4 reading (Figure 1). It also sets a bit in the high
VOLTAGE (VCLASS)
0
CURRENT (mA)
60
50
40
30
20
10
05 10 15 20
4266 F13
25
TYPICAL
CLASS 3
PD LOAD
LINE
48mA
33mA
PSE LOAD LINE
23mA
14.5mA
6.5mA
CLASS 4
CLASS 2
CLASS 1
CLASS 0
CLASS 3
OVER
CURRENT
Figure 13. PD Classification
LTC4266
20
4266fa
ApplicAtions inForMAtion
power status register to indicate that it ran the second
classification cycle. The second cycle alerts the PD that
it is connected to a Type 2 PSE which can supply Type 2
power levels.
2-event ping-pong classification is enabled by setting a bit
in the port’s high power mode register. Note that a ping-
pong enabled port only runs the second classification cycle
when it detects a Class 4 device; if the first cycle returns
Class 0 to 3, the port assumes it is connected to a Type 1
PD and does not run the second classification cycle.
Invalid Type 2 Class Combinations
The 802.3at spec defines a Type 2 PD class signature as
two consecutive Class 4 results; a Class 4 followed by
a Class 0-3 is not a valid signature. In auto mode, the
LTC4266 will power a detected PD regardless of the clas-
sification results, with one exception: if the PD presents
an invalid Type 2 signature (Class 4 followed by Class 0
to 3), the LTC4266 will not provide power and will restart
the detection process. To aid in diagnosis, the port status
register will always report the results of the last class pulse,
so an invalid Class 4–Class 2 combination would report
a second class pulse was run in the High Power Status
register (which implies that the first cycle found Class 4),
and Class 2 in the port status register.
POWER CONTROL
External MOSFET, Sense R Summary
The primary function of the LTC4266 is to control the
delivery of power to the PSE port. It does this by control-
ling the gate drive voltage of an external power MOSFET
while monitoring the current via an external sense resis-
tor and the output voltage at the OUT pin. This circuitry
serves to couple the raw VEE input supply to the port in
a controlled manner that satisfies the PD’s power needs
while minimizing power dissipation in the MOSFET and
disturbances on the VEE backplane.
The LTC4266 is designed to use 0.25Ω sense resistors to
minimize power dissipation. It also supports 0.5Ω sense
resistors, which are the default when LTC4258/LTC4259A
compatibility is desired.
Inrush Control
Once the command has been given to turn on a port, the
LTC4266 ramps up the GATE pin of that port’s external
MOSFET in a controlled manner. Under normal power-up
circumstances, the MOSFET gate will rise until the port
current reaches the inrush current limit level (typically
450mA), at which point the GATE pin will be servoed to
maintain the specified IINRUSH current. During this inrush
period, a timer (tSTART ) runs. When output charging is
complete, the port current will fall and the GATE pin will
be allowed to continue rising to fully enhance the MOSFET
and minimize its on-resistance. The final VGS is nominally
13V. If the tSTART timer expires before the inrush period
completes, the port will be turned back off and a tSTART
fault reported.
Current Limit
Each LTC4266 port includes two current limiting thresholds
(ICUT and ILIM), each with a corresponding timer (tCUT
and tLIM). Setting the ICUT and ILIM thresholds depends
on several factors: the class of the PD, the voltage of the
main supply (VEE), the type of PSE (1 or 2), the sense
resistor (0.5Ω or 0.25Ω), the SOA of the MOSFET, and
whether or not the system is required to implement class
enforcement.
Per the IEEE spec, the LTC4266 will allow the port cur-
rent to exceed ICUT for a limited period of time before
removing power from the port, whereas it will actively
control the MOSFET gate drive to keep the port current
below ILIM. The port does not take any action to limit the
current when only the ICUT threshold is exceeded, but
does start the tCUT timer. The tLIM timer starts when the
ILIM threshold is exceeded and current limit is active. If
the current drops below the ICUT current threshold before
its timer expires, the tCUT timer counts back down, but
at 1/16 the rate that it counts up. This allows the current
limit circuitry to tolerate intermittent overload signals with
duty cycles below about 6%; longer duty cycle overloads
will turn the port off.
ICUT is typically set to a lower value than ILIM to allow the
port to tolerate minor faults without current limiting.
LTC4266
21
4266fa
ApplicAtions inForMAtion
Per the IEEE specification, the LTC4266 will automatically
set ILIM to 425mA (shown in bold in Table 4) during inrush
at port turn-on, and then switch to the programmed ILIM
setting once inrush has completed. To maintain IEEE
compliance, ILIM should kept at 425mA for all Type 1 PDs,
and 850mA if a Type 2 PD is detected. ILIM is automatically
reset to 425mA when a port turns off.
Table 4. Example Current Limit Settings
ILIM (mA)
INTERNAL REGISTER SETTING (hex)
RSENSE = 0.5Ω RSENSE = 0.25Ω
53 88
106 08 88
159 89
213 80 08
266 8A
319 09 89
372 8B
425 00 80
478 8E
531 92 8A
584 CB
638 10 90
744 D2 9A
850 40 C0
956 4A CA
1063 50 D0
1169 5A DA
1275 60 E0
1488 52 49
1700 40
1913 4A
2125 50
2338 5A
2550 60
2975 52
ILIM Foldback
The LTC4266 features a two-stage foldback circuit that
reduces the port current if the port voltage falls below
the normal operating voltage. This keeps MOSFET power
dissipation at safe levels for typical 802.3af MOSFETs,
even at extended 802.3at power levels. Current limit and
foldback behavior are programmable on a per-port basis.
Figure 14 shows MOSFET power dissipation with 802.3af-
style foldback compared with a typical MOSFET SOA curve;
Figure 15 demonstrates how two-stage foldback keeps the
FET within its SOA under the same conditions. Table 4 gives
examples of recommended ILIM register settings.
The LTC4266 will support current levels well beyond the
maximum values in the 802.3at specification. The shaded
areas in Table 4 indicate settings that may require a larger
external MOSFET, additional heat sinking, or a reduced
tLIM setting.
PD Voltage (V) at VPSE = 58V
0
0.0
PSE Current (A)
0.2
0.4
0.6
10 20 30 40
4266 F14
50
0.8
1.0
0.1
0.3
0.5
0.7
0.9
60
802.3af FOLDBACK
2 x 802.3af FOLDBACK
SOA 75ms AT 25°C
SOA DC AT 90°C
PD Voltage (V) at VPSE = 58V
0
0.0
PSE Current (A)
0.2
0.4
0.6
10 20 30 40
4266 F15
50
0.8
1.0
0.1
0.3
0.5
0.7
0.9
60
LTC4266 FOLDBACK
802.3af FOLDBACK
SOA DC AT 90°C
SOA 75ms AT 90°C
SOA 75ms AT 25°C
Figure 14. Turn On Currents vs FET Safe Operating
Area at 90°C Ambient
Figure 15. LTC4266 Foldback vs FET Safe Operating
Area at 90°C Ambient
LTC4266
22
4266fa
ApplicAtions inForMAtion
MOSFET Fault Detection
LTC4266 PSE ports are designed to tolerate significant
levels of abuse, but in extreme cases it is possible for
the external MOSFET to be damaged. A failed MOSFET
may short source to drain, which will make the port ap-
pear to be on when it should be off; this condition may
also cause the sense resistor to fuse open, turning off
the port but causing the LTC4266 SENSE pin to rise to
an abnormally high voltage. A failed MOSFET may also
short from gate to drain, causing the LTC4266 GATE pin
to rise to an abnormally high voltage. The LTC4266 SENSE
and GATE pins are designed to tolerate up to 80V faults
without damage.
If the LTC4266 sees any of these conditions for more than
180μs, it disables all port functionality, reduces the gate
drive pull-down current for the port and reports a FET Bad
fault. This is typically a permanent fault, but the host can
attempt to recover by resetting the port, or by resetting
the entire chip if a port reset fails to clear the fault. If the
MOSFET is in fact bad, the fault will quickly return, and
the port will disable itself again. The remaining ports of
the LTC4266 are unaffected.
An open or missing MOSFET will not trigger a FET Bad
fault, but will cause a tSTART fault if the LTC4266 attempts
to turn on the port.
Voltage and Current Readback
The LTC4266 measures the output voltage and current
at each port with an internal A/D converter. Port data is
only valid when the port power is on. The converter has
two modes:
Slow mode: 14 samples per second, 14.5 bits resolution
Fast mode: 440 samples per second, 9.5 bits resolution
In fast mode, the least significant 5 bits of the lower byte are
zeroes so that bit scaling is the same in both modes.
Disconnect
The L
TC4266 monitors the port to make sure that the PD
continues to draw the minimum specified current. A dis-
connect timer counts up whenever port current is below
7.5mA (typ), indicating that the PD has been disconnected.
If the tDIS timer expires, the port will be turned off and
the disconnect bit in the fault event register will be set.
If the current returns before the tDIS timer runs out, the
timer resets and will start counting from the beginning
if the undercurrent condition returns. As long as the PD
exceeds the minimum current level more often than tDIS,
it will stay powered.
Although not recommended, the DC disconnect feature
can be disabled by clearing the corresponding enable
bits. Note that this defeats the protection mechanisms
built into the IEEE spec, since a powered port will stay
powered after the PD is removed. If the still-powered port
is subsequently connected to a non-PoE data device, the
device may be damaged.
The LTC4266 does not include AC disconnect circuitry, but
includes AC disconnect enable bits to maintain compat-
ibility with the LTC4259A. If the AC disconnect enable bits
are set, DC disconnect will be used.
Shutdown Pins
The LTC4266 includes a hardware SHDN pin for each port.
When a SHDN pin is pulled to DGND, the corresponding
port will be shut off immediately. The port remains shut
down until re-enabled via I2C or a device reset in auto
mode.
Masked Shutdown
The LTC4266 provides a low latency port shedding fea-
ture to quickly reduce the system load when required. By
allowing a pre-determined set of ports to be turned off,
the current on an overloaded main power supply can be
reduced rapidly while keeping high priority devices pow-
ered. Each port can be configured to high or low priority;
all low-priority ports will shut down within 6.5μs after
the MSD pin is pulled low. If multiple ports in a LTC4266
device are shut down via MSD, they are staggered by at
least 0.55μs to help reduce voltage transients on the main
supply. If a port is turned off via MSD, the corresponding
detection and classification enable bits are cleared, so
the port will remain off until the host explicitly re-enables
detection.
LTC4266
23
4266fa
SERIAL DIGITAL INTERFACE
Overview
The LTC4266 communicates with the host using a standard
SMBus/I2C 2-wire interface. The LTC4266 is a slave-only
device, and communicates with the host master using
the standard SMBus protocols. Interrupts are signaled to
the host via the INT pin. The Timing Diagrams (Figures 6
through 10) show typical communication waveforms and
their timing relationships. More information about the
SMBus data protocols can be found at www.smbus.org.
The LTC4266 requires both the VDD and VEE supply rails
to be present for the serial interface to function.
Bus Addressing
The LTC4266’s primary serial bus address is 010xxxxb, with
the lower four bits set by the AD3-AD0 pins; this allows up
to 16 LTC4266s on a single bus. All LTC4266s also respond
to the address 0110000b, allowing the host to write the
same command (typically configuration commands) to
multiple LTC4266s in a single transaction. If the LTC4266
is asserting the INT pin, it will also respond to the alert
response address (0001100b) per the SMBus spec.
Interrupts and SMBAlert
Most LTC4266 port events can be configured to trigger
an interrupt, asserting the INT pin and alerting the host
to the event. This removes the need for the host to poll
the LTC4266, minimizing serial bus traffic and conserving
host CPU cycles. Multiple LTC4266s can share a common
INT line, with the host using the SMBAlert protocol (ARA)
to determine which LTC4266 caused an interrupt.
Register Description
For information on serial bus usage and device configura-
tion and status, refer to the LTC4266 Software Program-
ming documentation.
EXTERNAL COMPONENT SELECTION
Power Supplies and Bypassing
The LTC4266 requires two supply voltages to operate. VDD
requires 3.3V (nominally) relative to DGND. VEE requires a
negative voltage of between –44V and –57V for Type 1 PSEs,
or –50V to –57V for Type 2 PSEs, relative to AGND. The
relationship between the two grounds is not fixed; AGND
can be referenced to any level from VDD to DGND, although
it should typically be tied to either VDD or DGND.
VDD provides power for most of the internal LTC4266 cir-
cuitry, and draws a maximum of 3mA. A ceramic decoupling
cap of at least 0.1μF should be placed from VDD to DGND,
as close as practical to each LTC4266 chip.
Figure 16 shows a three component low dropout regulator
for a negative supply to DGND generated from the negative
VEE supply. VDD is tied to AGND and DGND is negative
referenced to AGND. This regulator drives a single LTC4266
device. In Figure 17, DGND is tied to AGND in this boost
converter circuit for a positive VDD supply of 3.3V above
AGND. This circuit can drive multiple LTC4266 devices
and opto couplers.
ApplicAtions inForMAtion
4266 F16
R5
750k
D1
CMHZ4687-4.3V C1
0.1µF
Q2
CMPTA92
VEE
VDD
LTC4266
AGND
VEE
AGND
DGND
Figure 16. Negative LDO to DGND
VEE is the main supply that provides power to the PDs.
Because it supplies a relatively large amount of power and
is subject to significant current transients, it requires more
design care than a simple logic supply. For minimum IR
loss and best system efficiency, set VEE near maximum
amplitude (57V), leaving enough margin to account for
transient over- or undershoot, temperature drift, and the line
regulation specs of the particular power supply used.
Bypass capacitance between AGND and VEE is very im-
portant for reliable operation. If a short circuit occurs at
one of the output ports it can take as long as 1μs for the
LTC4266 to begin regulating the current. During this time
the current is limited only by the small impedances in the
circuit and a high current spike typically occurs, causing a
LTC4266
24
4266fa
ApplicAtions inForMAtion
voltage transient on the VEE supply and possibly causing
the LTC4266 to reset due to a UVLO fault. A 1μF, 100V
X7R capacitor placed near the VEE pin is recommended
to minimize spurious resets.
Isolating the Serial Bus
The LTC4266 includes a split SDA pin (SDAIN and SDAOUT)
to ease opto-isolation of the bidirectional SDA line.
IEEE 802.3 Ethernet specifications require that network
segments (including PoE circuitry) be electrically isolated
from the chassis ground of each network interface device.
However, network segments are not required to be isolated
from each other, provided that the segments are connected
to devices residing within a single building on a single
power distribution system.
For simple devices such as small PoE switches, the isola-
tion requirement can be met by using an isolated main
power supply for the entire device. This strategy can be
used if the device has no electrically conducting ports
other than twisted-pair Ethernet. In this case, the SDAIN
and SDAOUT pins can be tied together and will act as a
standard I2C/SMBus SDA pin.
If the device is part of a larger system, contains additional
external non-Ethernet ports, or must be referenced to
protective ground for some other reason, the Power over
Ethernet subsystem (including all LTC4266s) must be
electrically isolated from the rest of the system. Figure 18
shows a typical isolated serial interface. The SDAOUT pin
of the LTC4266 is designed to drive the inputs of an opto-
coupler directly. Standard I2C/SMBus devices typically
cannot drive opto-couplers, so U1 is used to buffer the
signals from the host controller side.
External MOSFET
Careful selection of the power MOSFET is critical to system
reliability. LTC recommends either Fairchild IRFM120A,
FDT3612, FDMC3612 or Philips PHT6NQ10T for their
proven reliability in Type 1 and Type 2 PSE applications.
Non-standard applications that provide more current than
the 850mA IEEE maximum may require heat sinking and
other MOSFET design considerations. Contact LTC Ap-
plications before using a MOSFET other than one of these
recommended parts.
Sense R
The LTC4266 is designed to use either 0.5Ω or 0.25Ω
current sense resistors. For new designs 0.25Ω is recom-
mended to reduce power dissipation; the 0.5Ω option is
intended for existing systems where the LTC4266 is used
as a drop-in replacement for the LTC4258 or LTC4259A.
The lower sense resistor values reduce heat dissipation.
Four commonly available resistors (0402 or larger
package size) can be used in parallel in place of a single
0.25Ω resistor. In order to meet the ICUT and ILIM accuracy
required by the IEEE specification, the sense resistors
should have ±1% tolerance or better, and no more than
±200ppm/°C temperature coefficient.
Figure 17. Positive VDD Boost Converter
4266 F17
R54
56k
C79
2200pF
GND
ITH/RUN
LTC3803
VCC
2
5
VFB
1
3
NGATE Q15
FDC2512
Q13
FMMT723
Q14
FMMT723
SENSE
6
4
VEE
C74
100µF
6.3V
C75
10µF
16V
L3
100µH
SUMIDA CDRH5D28-101NC
R51
4.7k
1%
R53
4.7k
1%
R52
3.32k
1%
3.3V AT 400mA
R55
806Ω
1%
R59
0.100Ω
1%, 1W
R56
47.5k
1%
R57
1k
D28
B1100
R58
10Ω
R60
10Ω
C73
10µF
6.3V
L4
10µH
SUMIDA CDRH4D28-100NC
+
C77
0.22µF
100V
C78
0.22µF
100V
C76
10µF
63V
LTC4266
25
4266fa
ApplicAtions inForMAtion
4266 F18
VDD
INT
SCL
SDAIN
SDAOUT
AD0
AD1
AD2
AD3
DGND
AGND
LTC4266
VDD
INT
SCL
SDAIN
SDAOUT
AD0
AD1
AD2
AD3
DGND
AGND
LTC4266
VDD
INT
SCL
SDAIN
SDAOUT
AD0
AD1
AD2
AD3
DGND
AGND
LTC4266
VDD
INT
SCL
SDAIN
SDAOUT
AD0
AD1
AD2
AD3
DGND
AGND
LTC4266
VDD
INT
SCL
SDAIN
SDAOUT
AD0
AD1
AD2
AD3
DGND
AGND
LTC4266
0.1µF
0.1µF
0.1µF
0.1µF
0.1µF
+
10µF
2k
2k
0.1µF
0.1µF
200Ω
200Ω
200Ω
200Ω
U2
U3
U1
HCPL-063L
HCPL-063L
VDD CPU
SCL
SDA
SMBALERT
GND CPU
U1: FAIRCHILD NC7WZ17
U2, U3: AGILENT HCPL-063L
TO
CONTROLLER
ISOLATED
3.3V
ISOLATED
GND
0100000
0100001
0100010
0101110
0101111
I2C ADDRESS
Figure 18. Opto-Isolating the I2C Bus
LTC4266
26
4266fa
Output Cap
Each port requires a 0.22μF cap across its outputs to keep
the LTC4266 stable while in current limit during startup
or overload. Common ceramic capacitors often have sig-
nificant voltage coefficients; this means the capacitance
is reduced as the applied voltage increases. To minimize
this problem, X7R ceramic capacitors rated for at least
100V are recommended.
ESD/Cable Discharge Protection
Ethernet ports can be subject to significant ESD events
when long data cables, each potentially charged to thou-
sands of volts, are plugged into the low impedance of the
RJ45 jack. To protect against damage, each port requires a
pair of clamp diodes; one to AGND and one to VEE (Figure
10). An additional surge suppressor is required for each
LTC4266 chip from VEE to AGND. The diodes at the ports
steer harmful surges into the supply rails, where they are
absorbed by the surge suppressor and the VEE bypass
capacitance. The surge suppressor has the additional
benefit of protecting the LTC4266 from transients on the
VEE supply.
S1B diodes work well as port clamp diodes, and an
SMAJ58A or equivalent is recommended for the VEE surge
suppressor.
LAYOUT GUIDELINES
Standard power layout guidelines apply to the LTC4266:
place the decoupling caps for the VDD and VEE supplies
near their respective supply pins, use ground planes, and
use wide traces wherever there are significant currents.
The main layout challenge involves the arrangement of
the current sense resistors, and their connections to
the LTC4266. Because the sense resistor values are very
low, layout parasitics can cause significant errors. Care is
required to achieve specified accuracy, particularly with
disconnect currents.
Figure 19 illustrates the problem. In the example on the
left, two ports have load currents I1 and I2 that return to
the VEE power supply through a mutual resistance RM.
RM represents the combined resistances of any traces,
planes, and vias in the PCB that I1 and I2 share as they
return to the VEE supply. The LTC4266 measures the volt-
age difference between its SENSE and VEE pins to sense
the voltage drop across RS1, but as the example shows,
RM introduces errors.
The example on the right shows how errors can be
minimized with a good layout. The circuit is rearranged
so that RM no longer affects VS, and the VEE connection
to the LTC4266 is used as a Kelvin sense trace. VEE is not
a perfect Kelvin connection because all four ports con-
trolled by the LTC4266 share the same sense trace, and
because the current through the trace (IEE) is not zero.
However, as the equation shows, the remaining error is a
small offset term.
Figure 20 shows two LTC4266 chips controlling eight ports
(A though H). The ports are separated into two groups
of four; each has its own trace on the top PCB layer that
ApplicAtions inForMAtion
RM
+
VS
+
VS
RS1
MUTUAL RESISTANCE
RS2
4266 F19
IEE
I1I2
I1 + I2 + IEE
VS = I1RS1 + I1RM + I2RM
LTC4266
GATE
SENSE
SIGNAL
SCALE ERROR
CROSSTALK ERROR
VEE
RKRM
RS1
KELVIN SENSE LINE
RS2
IEE
I1I2
VS = I1RS1 – IEERK
I1 + I2 + IEE
LTC4266
GATE
SENSE
SIGNAL
SMALL OFFSET ERROR
VEE
Figure 19. Layout Affects Current Readback Accuracy
LTC4266
27
4266fa
U1
LTC4266
PORTS A THROUGH D
SENSE1
SENSE2
SENSE3
SENSE4
VEE
VIA VIA
4266 F20
BY KEEPING THESE COPPER FILLS SEPARATE ON
THE SURFACE, MUTUAL RESISTANCE BETWEEN
PORTS A-D AND E-H IS ELIMINATED
THIS TRACE PROVIDES VEE TO U1
BUT ALSO ACTS AS A KELVIN
SENSE LINE FOR PORTS A-D
VEE COPPER FILL ON SURFACE LAYER
VEE PLANE ON INNER LAYER
U2
LTC4266
PORTS E THROUGH H
SENSE1
SENSE2
SENSE3
SENSE4
VEE RSENSE
RETURN TO
VEE POWER SUPPLY
Figure 20. Layout Strategy to Reduce Mutual Resistance
ApplicAtions inForMAtion
Figure 21. Good PCB Layout Example
4266 F21
PORT A RSENSE PORT B RSENSE
FOUR LARGE VIAS
TO VEE PLANE
PIN 1
U1 THE PADDLE IS
CONNECTED TO
VEE PINS
KELVIN SENSE TRACE CONNECTS U1
TO VEE THROUGH THE VIAS ON THE RIGHT
EDGE OF VEE PLANE (ON SOME INNER LAYER)
VIAS TO SOURCE PIN OF
THE PORT D MOSFET
LOCATED ON THE OPPOSITE
SIDE OF THE BOARD
HOLE IN VEE PLANE
PORT C RSENSE PORT D RSENSE
connects to the VEE plane with a via. Currents from the U1
sub-circuit are effectively isolated from the U2 sub-circuit,
reducing the layout problem down to 4-port chunks; this
arrangement can be expanded for any number of ports.
Figure 21 shows an example of good 4-port layout. Each
0.25Ω sense resistor consists of four resistors in paral-
lel. The four groups of resistors are arranged to minimize
the overlap in their current flows, which minimizes mutual
resistance. The horizontal slits cut in the copper help to
keep the currents separate. Wide copper paths connect
each group of resistors to the vias at the center, so the
resistance is very low.
Proper connection of the sense line is also important. In
Figure 21, U1 is not connected directly to the VEE plane
but is connected instead to a Kelvin sense trace that
leads to the sense resistor array. Similarly, the via at the
center of the sense resistor array has a matching hole
in the VEE plane. This arrangement prevents the mutual
resistance of the four large vias from influencing the
current measurements.
LTC4266
28
4266fa
pAckAge Description
GW36 SSOP 0204
0° – 8° TYP
0.355
REF
0.231 – 0.3175
(.0091 – .0125)
0.40 – 1.27
(.015 – .050)
7.417 – 7.595**
(.292 – .299)
s 45°
0.254 – 0.406
(.010 – .016)
2.286 – 2.388
(.090 – .094)
0.1 – 0.3
(.004 – .0118)
2.44 – 2.64
(.096 – .104)
0.800
(.0315)
BSC
0.28 – 0.51
(.011 – .02)
TYP
15.291 – 15.545*
(.602 – .612)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
10.11 – 10.55
(.398 – .415)
36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19
10.804 MIN
RECOMMENDED SOLDER PAD LAYOUT
7.75 – 8.258
1936
181
0.800 BSC0.520 ±0.0635
1.40 ±0.127
DIMENSION DOES NOT INCLUDE MOLD FLASH. MOLD FLASH
SHALL NOT EXCEED 0.152mm (0.006") PER SIDE
*
DIMENSION DOES NOT INCLUDE INTERLEAD FLASH. INTERLEAD
FLASH SHALL NOT EXCEED 0.254mm (0.010") PER SIDE
**
MILLIMETERS
(INCHES)
NOTE:
1. CONTROLLING DIMENSION: MILLIMETERS
2. DIMENSIONS ARE IN
GW Package
36-Lead Plastic SSOP (Wide .300 Inch)
(Reference LTC DWG # 05-08-1642)
LTC4266
29
4266fa
pAckAge Description
5.00 p 0.10
NOTE:
1. DRAWING CONFORMS TO JEDEC PACKAGE
OUTLINE M0-220 VARIATION WHKD
2. DRAWING NOT TO SCALE
3. ALL DIMENSIONS ARE IN MILLIMETERS
PIN 1
TOP MARK
(SEE NOTE 6)
37
1
2
38
BOTTOM VIEW—EXPOSED PAD
5.50 REF
5.15 ± 0.10
7.00 p 0.10
0.75 p 0.05
R = 0.125
TYP
R = 0.10
TYP
0.25 p 0.05
(UH) QFN REF C 1107
0.50 BSC
0.200 REF
0.00 – 0.05
RECOMMENDED SOLDER PAD LAYOUT
APPLY SOLDER MASK TO AREAS THAT ARE NOT SOLDERED
3.00 REF
3.15 ± 0.10
0.40 p0.10
0.70 p 0.05
0.50 BSC
5.5 REF
3.00 REF 3.15 ± 0.05
4.10 p 0.05
5.50 p 0.05 5.15 ± 0.05
6.10 p 0.05
7.50 p 0.05
0.25 p 0.05
PACKAGE
OUTLINE
4. DIMENSIONS OF EXPOSED PAD ON BOTTOM OF PACKAGE DO NOT INCLUDE
MOLD FLASH. MOLD FLASH, IF PRESENT, SHALL NOT EXCEED 0.20mm ON ANY SIDE
5. EXPOSED PAD SHALL BE SOLDER PLATED
6. SHADED AREA IS ONLY A REFERENCE FOR PIN 1 LOCATION
ON THE TOP AND BOTTOM OF PACKAGE
PIN 1 NOTCH
R = 0.30 TYP OR
0.35 s 45o CHAMFER
Information furnished by Linear Technology Corporation is believed to be accurate and reliable.
However, no responsibility is assumed for its use. Linear Technology Corporation makes no representa-
tion that the interconnection of its circuits as described herein will not infringe on existing patent rights.
UHF Package
38-Lead Plastic QFN (5mm × 7mm)
(Reference LTC DWG # 05-08-1701 Rev C)
LTC4266
30
4266fa
relAteD pArts
PART NUMBER DESCRIPTION COMMENTS
LT3803 Constant Frequency Current Mode Flyback DC/DC
Controller in ThinSOT™
200kHz Operation, Adjustable Slope Compensation
LTC4258 Quad IEEE 802.3af PoE PSE Controller DC Disconnect Sensing Only
LTC4263 Single IEEE 802.3af PSE Controller Internal FET Switch
LTC4263-1 High Power Single PoE PSE Controller With Internal FET Switch
LTC4265 IEEE 802.3at PD Interface Controller 100V, 1A Internal Switch, 2-Event Classification Recognition
LTC4267 IEEE 802.3af PD Interface With Integrated Switching
Regulator
Internal 100V, 400mA Switch, Dual Inrush Current, Programmable Class
LTC4268-1 High Power PD With Synchronous Flyback Controller No Opto-coupler Required
LTC4269-1 IEEE 802.3at PD Interface Integrated Switching Regulator 2-Event Classification, Programmable Classification, Synchronous No-Opto
Flyback Controller, 50kHz to 250kHz
LTC4269-2 IEEE 802.3at PD Interface Integrated Switching Regulator 2-Event Classification, Programmable Classification, Synchronous Forward
Controller, 100kHz to 500kHz
LINEAR TECHNOLOGY CORPORATION 2009
LT 0810 REV A • PRINTED IN USA
Linear Technology Corporation
1630 McCarthy Blvd., Milpitas, CA 95035-7417
(408) 432-1900 FAX: (408) 434-0507 www.linear.com
4266 F22
1
2
3
4
5
6
7
8
2k
2k
1µF
100V
X7R
0.1µF
0.1µF
200Ω
200Ω
200Ω
200Ω
U2
U3
U1
HCPL-063L
HCPL-063L
VDD CPU
SCL
SDA
GND CPU
INTERRUPT
DTSS: DIODES INC SMAJ58A
Q1: FAIRCHILD IRFM120A OR PHILIPS PHT6NQ10T
U1: FAIRCHILD NC7WZ17
U2, U3: AGILENT HCPL-063L
FB1, FB2:TDK MPZ2012S601A
T1: PULSE H6096NL OR COILCRAFT ETH1-230LD
TO
CONTROLLER
PHY
(NETWORK
PHYSICAL
LAYER
CHIP)
0.22µF
100V
X7R
VEE SENSE GATE OUT
VDD
SCL
SDAIN
SDAOUT
INT
RS
0.25Ω
Q1
T1
1/4
LTC4266
DGND AGND
–48V
ISOLATED
SMAJ58A S1B
ISOLATED
3.3V
ISOLATED
GND
0.01µF
200V
0.01µF
200V
0.01µF
200V
0.01µF
200V
75Ω75Ω
75Ω75Ω
RJ45
CONNECTOR
1000pF
2000V
FB1
FB2
S1B
Figure 22. One Complete Isolated Powered Etherent Port
typicAl ApplicAtion