For more information, visit our website at: www.marvell.com
Disclaimer
No part of this document may be repr od uced or transmitted in any form or by any means, electronic or m ech ani ca l, in cl ud in g p hotocopying and recording, for any purpose,
without the express written permission of Marvell. Marvell retains the right to m ake cha ng es to th is document at any time, without notice. Marvell makes no warranty of any
kind, expressed or implied, with r egard to any info rmation conta ined in this do cument, in cluding, but no t limited to, the implie d warra nties of merchant abil ity or fi tness for a ny
particular purpose. Further, Marvell does not warrant the accuracy or completeness of the information, text, graphics, or other items contained within this document.
Marvell products are not designed for use in life-support equipment or applications that would cause a life-threatening situation if any such products failed. Do not use
Marvell products in these types of equipment or applications.
With respect to the products described herein, the user or recipient, in the absence of appropriate U.S. government authorization, agrees:
1) Not to re-export or release any such information consisting of technology, software or source code controlled for national security reasons by the U.S. Export Control
Regulations ("EAR"), to a national of EAR Count ry Gr oups D:1 or E:2;
2) Not to export the direct product of such technology or such software, to EAR Country Groups D:1 or E:2, if such technology or software and direct products thereof are
controlled for national security reasons by the EAR; and,
3) In the case of technology controlled for national security reasons under the EAR where the direct product of the technology is a complete plant or component of a plant,
not to export to EAR Country Groups D:1 or E:2 the direct product of the plant or major component thereof, if such direct product is controlled for national security reasons
by the EAR, or is subject to controls under the U.S. Munitions List ("USML").
At all times hereunder , the recipi ent o f any such info rmati on agre es t hat the y shal l be deem ed to h ave manu all y sign ed this d ocument in conne ction with their re ceipt of any
such information.
The Marvell® 88F6281 is a high-performance, highly integrated controller. The 88F6281 is based on the Marvell
proprietary, ARMv5TE-compliant, high-speed Sheeva
™
CPU core. The CPU core integrates a 256 KB L2 cache.
About this Document.......................................................................................................................................15
Related Documentation...................................................................................................................................15
1.3Internal Pull-up and Pull-down Pins................................................................................................................48
6.6Serial ROM Initialization..................................................................................................................................70
8.1Absolute Maximum Ratings ............................................................................................................................75
8.3Thermal Power Dissipation .............................................................................................................................79
Figure 5:SDRAM DDR2 Interface Test Circuit................................................................................................91
Figure 6:SDRAM DDR2 Interface Write AC Timing Diagram .........................................................................91
Figure 7:SDRAM DDR2 Interface Address and Control AC Timing Diagram.................................................92
Figure 8:SDRAM DDR2 Interface Read AC Timing Diagram.........................................................................92
Figure 9:RGMII Test Circuit ................. ... ... .....................................................................................................94
Figure 10:RGMII AC Timing Diagram...............................................................................................................94
Figure 11: GMII Test Circuit...............................................................................................................................95
Figure 12: GMII Output AC Timing Diagram............................................. ... ......................................................96
Figure 13: GMII Input AC Timing Diagram.........................................................................................................96
Figure 14: MII/MMII MAC Mode Test Circuit......................................................................................................97
Figure 15: MII/MMII MAC Mode Output Delay AC Timing Diagram...................................................................97
Figure 16: MII/MMII MAC Mode Input AC Timing Diagram................................................................................98
Figure 17: MDIO Master Mode Test Circuit.......................................................................................................99
Figure 18: MDC Master Mode Test Circuit ......................................................................................................100
Figure 19: SMI Master Mode Output AC Timing Diagram...............................................................................100
Figure 20: SMI Master Mode Input AC Timing Diagram..................................................................................100
Figure 21: JTAG Interface Test Circuit ............................................................................................................101
Figure 22: JTAG Interface Output Delay AC Timing Diagram .........................................................................102
Figure 23: JTAG Interface Input AC Timing Diagram ......................................................................................102
Figure 24: TWSI Test Circuit............................................................................................................................104
Figure 25: TWSI Output Delay AC Timing Diagram.........................................................................................104
Figure 26: TWSI Input AC Timing Diagram .....................................................................................................104
Figure 27: S/PDIF Test Circuit............................................................. ... .. ................................ .......................106
This datasheet provides the hardware specifications for the 88F6281 integrated controller. The
hardware specifications include detailed pin information, configuration settings, electrical
characteristics and physical specifications.
This datasheet is intended to be the basic source of information for designers of new systems.
In this document, the “88F6281” is often referred to as the “device”.
Related Documentation
The following documents contain additional information related to the 88 F6281:
88F6180, 88F6190, 88F6192, and 88F62 81 Functional Specifications,
Doc No. MV-S104860-U0
Sheeva
Doc No. MV-S104950-U0
Unified Layer 2 (L2) Cache for Sheeva
88F6180, 88F6190, 88F6192, and 88F6281 Functional Errata, Interface Guidelines, and
Restrictions, Doc No. MV-S501157-U0
88F6180, 88F6190, 88F6192, and 88F6281 Design Guide, Doc No. MV-S301398-00
AN-63: Thermal Management for Marvell Technology Products Doc No. MV-S300281-00
AN-179: TWSI Software Guidelines for Discovery™, Horizon™, and Feroceon
Doc No. MV-S300754-00
AN-183: 88F5181 and 88F5281 Big Endian and Little Endian Support,
Doc No. MV-S300767-00
AN-249: Configuring the Marvell
Doc No. MV-S301342-00
AN-260 System Power-Saving Methods for 88F6180, 88F6190, 88F6192, and 88F6281,
Doc No. MV-S301454-00
TB-227: Differences Between the 88F6190, 88F6192, and 88F6281 Stepping Z0 and A0,
Doc No. MV-S105223-00
White Paper, ThetaJC, ThetaJA, and Temperature Calculations, Doc No. MV-S700019-00
ARM Architecture Reference Manual, Second Edition
PCI Express Base Specification, Revision 1.1
Universal Serial Bus Specification, Revision 2.0, April 2000, Compaq, Hewlett-Packard, Intel,
Lucent, Microsoft, NEC, Philips
Enhanced Host Controller Interface Specification for Universal Serial Bus, Revision 0.95,
November 2000, Intel Corporation
ARC USB-HS OTG High-Speed Controller Core reference V 4.0.1
Federal Information Processing Standards (FIPS) 46-2 (Data Encryption Standard)
FIPS 81 (DES Modes of Operation)
FIPS 180-1 (Secure Hash Standard)
FIPS draft - Advanced Encryption Standard (Rijndeal)
1. This document is a Marvell proprietary, confidential document, requiring an NDA and can be downloaded from the
™
88SV131 ARM v5TE Processor Core with MMU and L1/L2 Cache Datasheet,
RFC 1321 (The MD5 Message-Digest Algorithm)
RFC 1851 – The ESP Triple DES Transform
RFC 2104 (HMAC: Keyed-Hashing for Message Authentication).
RFC 2405 – The ESP DES-CBC Cipher Algorithm With Explicit IV
IEEE standard, 802.3-2000 Clause 14
ANSI standard X3.263-1995
See the Marvell Extranet website for the latest product documentation.
Document Conventions
The following conventions are used in this document:
Signal RangeA signal name followed by a range enclosed i n bracket s represent s a range of l ogical ly relat ed
signals. The first number in the range indicates the most significant bit (MSb) and the last
number indicates the least significant bit (LSb).
Example: DB_Addr[12:0]
Active Low Signals #An n letter at the end of a signal name indicates that the signal’s active state occurs when
voltage is low.
Example: INTn
State NamesState names are indicated in italic font.
Example: linkfail
Register Naming
Conventions
Reset ValuesReset values have the following meanings:
AbbreviationsKb: kil obit
Numbering ConventionsUnless otherwise indicated, all numbers in this document are decimal (base 10).
Register field names are indicated by angle brackets.
Example: <RegInit>
Register field bits are enclosed in brackets.
Example: Field [1:0]
Register addresses are represented in hexadecimal format.
Example: 0x0
Reserved: The contents of the register are reserved for interna l use only or for future use.
A lowercase <n> in angle brackets in a register indicates that there are multiple registers with
this name.
Example: Multicast Configuration Register< n>
This section details all the pins for the different interfaces providing a functional description of each
pin and pin attributes.
Table 1<Default ¬¹ Font> defines the abbreviations and acronyms used in the pin description tables.
Table 1:Pin Functions and Assignments Table Key
TermDefinition
[n]n - Represents the SERDES pair number
<n>Represents port number when there are more than one ports
AnalogAnalog Driver/Receiver or Power Supply
CalibCalibration pad type
CMLCommon Mode Logic
CMOSComplementary Metal-Oxide-Semiconductor
DDRDouble Data Rate
GNDGround Supply
HCSLHigh-speed Current Steering Logic
IInput
I/OInput/Output
OOutput
o/dOpen Drain pin
The pin allows multiple drivers simultaneously (wire-OR conn ection).
A pull-up is required to sustain the inactive value.
PowerVDD Power Supply
SSTLStub Series Terminated Logic for 1.8V
t/sTri-State pin
XXXnn - Suffix represents an Active Low Signal
Table 3 provides the voltage levels for the various interface pins. These do not include the analog
power supplies for the PLLs or PHYs which are explicitly mentioned in the other pin description
tables.
Table 3:Power Pin Assignments
Pin and Signal Descriptions
Pin Descriptions
Pin NameI/OPin
Description
Type
VDDIPower1.0V Digital core voltage
VDD_CPUIPower1.1V Digital CPU voltage
VDDOIPower3.3V I/O power for MPP[49:36],MPP[19:0] and JTAG pins
VDD_GE_AIPower1.8V or 3.3V I/O supply voltage for RGMII and SMI interfaces
3.3V I/O supply voltage for GMII, MII/MMII, and SMI interfaces
VDD_GE_BIPowerI/O power for MPP[35:20]
1.8V or 3.3V I/O supply voltage for RGMII interfaces
3.3V I/O supply voltage for GMII and MII/MMII interfaces
VDD_M IPower1.8V I/O supply voltage for the DDR2 SDRAM interface
VSSIGNDVSS
CPU_PLL_AVDDIPower1.8V analog quiet power to CPU PLL
NOTE: See the 88F6180, 88F6190, 88F6192, and 88F6281 Design
Guide
for power supply filtering recommendations.
CPU_PLL_AVSSIGNDCPU PLL ground
CORE_PLL_AVDDIPower1.8V analog quiet power to Core PLL
NOTE: See the 88F6180, 88F6190, 88F6192, and 88F6281 Design
Guide for power supply filtering recommendations.
CORE_PLL_AVSSIGNDCore PLL ground
SSCG_AVDDIPower1.8V quiet power supply to the internal Spread Spectrum Clock
Generator
SSCG_AVSSIGNDGround for the internal Spread Spectrum Clock Generator
XTAL_ AVDDIPower1.8V analog quiet power to on-chip clock inverter for supporting external
crystal, and on-chip current reference for SATA and USB PHYs
NOTE: See the 88F6180, 88F6190, 88F6192, and 88F6281 Design
Guide for power supply filtering recommendations.
XTAL_ A VSSIGNDGround for supporting external crystal, and on-chip current reference for
SATA and USB PHYs
VHVIPowerI/O supply voltage for eFuse:
This pin is multiplexed on the MPP pins (see Section 4, Pin
Multiplexing, on page 51).
TPOAnalogAnalog Test Point for SATA, USB, and PCI Express
interfaces
For internal use. Leave this pin unconnected.
ISETIAnalogCurrent reference for both the USB and SATA PHYs.
Terminate this pin with a 6.04 k
Ω resistor, pulled down.
MRnICMOSVDD_GE_AActive-Low, Manual Reset Input
SYSRST_OUTn is asserted low as long as the MRn input
signal is asserted low, and for additional 20 ms after MRn
(manual reset) de-assertion
This pin is internally pulled up.
88F6281 indication of starting a burst read transaction.
Asserted with the first M_CASn cycle of SDRAM access.
NOTE: Must be routed on board to the SDRAM, and back to
the 88F6281 as M_STARTBURST_IN. For the
recommended length calculation for this routing and
termination requirements, see the 8 8F6180, 88F619 0, 88F6192, and 88F6281 Design Guide.
M_START
ISSTLVDD_MStart Burst Input
BURST_IN
M_PCALICalibSDRAM interface P channel output driver calibration. Connect
to VSS through a resistor. The resistor value can vary
between 30–70 ohm.
NOTE: See the 88F6180, 88F6190, 88F6192, and 88F6281
Design Guide for the recommended values of the
calibration resistors.
M_NCALICalibSDRAM interface N channel output driver calibration. Connect
to M_VDD through a resistor. The resistor value can vary
between 30–70 ohm.
NOTE: See the 88F6180, 88F6190, 88F6192, and 88F6281
100 MHz, differential
This clock can be configured as input or output according to the
reset strap (see Table 32, Reset Configuration, on page 67).NOTE: For Output mode, 50-ohm, pull-down resistors are
required.
PEX_TX_P/NOCMLPEX_AVDDTransmit Lane
Differential pair of PCI Express transmit data
PEX_RX_P/NICMLPEX_AVDDReceive Lane
Differential pair of PCI Express receive data
PEX_ISETIAnalogCurrent reference. Pull down to VSS through a 5 k
See the 88F6180, 88F6190, 88F6192, and 88F6281 Design Guide for the recommended resistor value.
RGMII transmit reference output clock for GE_TXD[3:0] and
GE_TXCTL.
Provides 125 MHz, 25 MHz or 2.5 MHz clock.
Not used in MII/MMII mode.
IMII/MMII Transmit Clock
MII/MMII transmit reference clock fro m PHY.
Provides the timing reference for the transmission of the MII
transmit clock, transmit enable, and GE_TXD[3:0] signals. This
clock operates at 2.5 MHz or 25 MHz.
t/s
O
GMII Transmit Clock
Provides the timing reference for the transfer of the transmit
enable, transmit error and transmit data signals. This clock
operates at 125 MHz.
GE_TXD[3:0]t/s OCMOSVDD_GE_ARGMII Transmit Data
Contains the transmit data nibble outp ut s that run at doubl e dat a
rate with bits [3:0] driven on the rising edge of GE_TXCLKOUT
and bits [7:4] driven on the falling edge.
MII/MMII Transmit Data
Contains the transmit data nibble outputs that are synchronous
to the transmit clock input.
GMII Transmit Data
Contains the transmit data nibble outputs.
GE_TXCTL t/s OCMOSVDD_GE_ARGMII Transmit Control
Transmit control synchronous to the GE_TXCLKOUT output
rising/falling edge.
GE_TXEN is driven on the rising edge of GE_TXCLKOUT.
A logical derivative of transmit enable and transmit error is driven
on the falling edge of GE_TXCLKOUT.
MII/MMII Transmit Enable
Indicates that the packet is being transmitted to the PH Y. It Is
synchronous to transmit clock.
GMII Transmit Enable
Indicates that the packet is being transmitted to the PH Y.
It Is synchronous to GE_TXCLKOUT.
Contains the receive data nibble inputs that are synchronous to
GE_RXCLK input rising/falling edge.
MII/MMII Receive Data
Contains the receive data nibble inputs that are synchronous to
GE_RXCLK input.
GMII Receive Data
Contains the receive data nibble inputs.
GE_RXCTLICMOSVDD_GE_ARGMII Receive Control
GE_RXCTL is presented on the rising edge of GE_RXCLK.
A logical derivative of receive data vali d and receive d ata error is
presented on the falling edge of RXCLK.
MII/MMII Receive Data Valid
GMII Receive Data Valid.
GE_RXCLKICMOSVDD_GE_ARGMII Receive Clock
The receive clock provides a 125 MHz, 25 MHz, or 2.5 MHz
reference clock derived from the received data stream.
MII/MMII Receive Clock
Provides the timing reference for the reception of the receive
data valid, receive error, and GE_RXD[3:0] signals. This clock
operates at 2.5 MHz or 25 MHz.
GMII Receive Clock
Provides the timing reference for the reception of the GE_RXDV,
receive error and receive data signals. This clock operates at
125 MHz
Port1—Multiplexed GbE Pins
MPP[23:20]/
GE1[3:0]
t/s OCMOSVDD_GE_BRGMII Transmit Data
Contains the transmit data nibble outp ut s that run at doubl e dat a
rate with bits [3:0] presented on the rising edge of
GE_TXCLKOUT and bits [7:4] presented on the falling edge.
MII/MMII Transmit Data
Contains the transmit data nibble outputs that are synchronous
to the transmit clock input.
GMII Transmit Data
Contains the transmit data nibble outputs.
MII/MMII transmit reference clock fro m PHY.
Provides the timing reference for the transmission of the MII
transmit clock, transmit enable, and GE_TXD[3:0] signals. This
clock operates at 2.5 MHz or 25 MHz.
t/s
O
GMII Transmit Clock
Provides the timing reference for the transfer of the transmit
enable, transmit error and transmit data signals. This clock
operates at 125 MHz.
MPP[30]/GE1[10]ICMOSVDD_GE_BRGMII Receive Control
GE_RXCTL is presented on the rising edge of GE_RXCLK.
A logical derivative of receive data vali d and receive d ata error is
presented on the falling edge of RXCLK.
MII/MMII Receive Data Valid
GMII Receive Error
MPP[31]/GE1[11]ICMOSVDD_GE_BRGMII Receive Clock
The receive clock provides a 125 MHz, 25 MHz, or 2.5 MHz
reference clock derived from the received data stream.
MII/MMII Receive Clock
Provides the timing reference for the reception of the receive
data valid, receive error, and GE_RXD[3:0] signals. This clock
operates at 2.5 MHz or 25 MHz.
RGMII transmit reference output clock for GE_TXD[3:0] and
GE_TXCTL Provides 125 MHz, 25 MHz or 2.5 MHz clock.
Not used in MII/MMII mode.
MII/MMII Carrier Sense
Indicates that the receive medium is non-idle. In half-duplex
mode, GE_CRS is also asserted during transmission. Carrier
sense is not synchronous to any clock.
GMII Carrier Sense
MPP[33]/GE1[13]t/s OCMOSVDD_GE_BRGMII Transmit Control
Transmit control synchronous to the GE_TXCLKOUT output
rising/falling edge.
GE_TXEN is presented on the rising edge of GE_TXCLKOUT.
A logical derivative of transmit enabl e transmit error i s presented
on the falling edge of GE_TXCLKOUT.
MII/MMII Transmit Error
It is synchronous to transmit clock.
NOTE: Multiplexed on MPP.
GMII Transmit Error
It Is synchronous to GE_TXCLKOUT.
NOTE: Multiplexed on MPP.
Indicates that an error symbol, a false carrier, or a carrier
extension symbol is detected on the cable. It is synchronous to
GE_RXCLK input.
NOTE: Multiplexed on MPP.
Clock input for the JTAG controller.
NOTE: This pin is internally pulled down to 0.
JT_RSTnICMOSVDDOJTAG Reset
When asserted, resets the JTAG controller.
NOTE: This pin is internally pulled down to 0.
1
JT_TMS_CPUICMOSVDDOCPU JTAG Mode Select
Controls CPU JTAG controller state.
Sampled with the rising edge of JT_CLK.
NOTE: This pin is internally pulled up to 1.
JT_TMS_COREICMOSVDDOCore JTAG Mode Select
Controls the Core JTAG controller state.
Sampled with the rising edge of JT_CLK.
NOTE: This pin is internally pulled up to 1.
JT_TDOOCMOSVDDOJTAG Data Out
Driven on the falling edge of JT_CLK.
JT_TDIICMOSVDDOJTAG Data In
JTAG serial data input. Sampled with the JT_CLK rising edge.
NOTE: This pin is internally pulled up to 1.
1. If this pull-down conflicts with other devices, the JTAG tool must not use this signal. This signal is not mandatory for the
JTAG interface, since the TAP (Test Access Port) can be reset by driving the JT_TMS signal HIGH for 5 JT_CLK cycles.
Used to transfer a command serially from the SDIO host to the
SDIO device. Used to transfer a command response serially
from the SDIO device to the SDIO host.
NOTE: This pin requires a pull up on board.
Used to transfer data from the SDIO host to the SDIO device or
vice versa.
NOTE: These pins require a pull up on board.
All of the TDM signals are multiplexed on the MPP pins (see Section 4, Pin
Multiplexing, on page 51).
The TDM signals are powered on VDDO or on VDD_GE_B, based on the pin
multiplexing option (see Section 4, Pin Multiplexing, on page 51).
Table 20: Time Division Multiplexing (TDM) Interface Signal Assignment
Pin Name
I/O
Pin TypePower Rail
TDM_CH0_TX_QLOCMOSVDDO/
VDD_GE_B
TDM_CH2_TX_QLOCMOSVDDO/
VDD_GE_B
TDM_CH0_RX_QLOCMOSVDDO/
VDD_GE_B
TDM_CH2_RX_QLOCMOSVDDO/
VDD_GE_B
TDM_CODEC_
INTn
TDM_CODEC_
RSTn
ICMOSVDDO/
VDD_GE_B
OCMOSVDDO/
VDD_GE_B
TDM_PCLKI/OCMOSVDDO/
VDD_GE_B
TDM_FSI/OCMOSV DDO/
VDD_GE_B
TDM_DRXICMOSVDDO/
VDD_GE_B
TDM_DTXOCMOSVDDO/
VDD_GE_B
TDM_SPI_CS[1:0]
OCMOSVDDO/
VDD_GE_B
TDM_SPI_SCKOCMOSVDDO/
VDD_GE_B
Description
TDM Channel0 Transmit Qualifier
TDM Channel2 Transmit Qualifier
TDM Channel0 Receive Qualifier
TDM Channel2 Receive Qualifier
Interrupt Signal FROM the SLIC/codec
SLIC/codec
Reset Signal
PCM Audio Bit Clock
TDM Frame Sync Signal
PCM Audio Input Data (for recording)
PCM Audio Output Data (for playback)
Active low SPI chip selects driven by the host to the codec for
register access. Always asserted for eight SCLK cycles at a time.
Only Byte-by-Byte mode codec register read/write is supported.
Serial SPI clock from the host to the codec for register acce ss.
This is an RTO (return to one) clock. It toggles for eight cycles at
a time (for 1 byte transfer) during codec register access, then it
returns to high.
The host drives write data on TDM_SPI_MOSI on the negative
edge of TDM_SPI_SCK, and captures read data from the codec
on the positive edge of TDM_SPI_SCK.
Table 20: Time Division Multiplexing (TDM) Interface Signal Assignment (Continued)
Pin Name
I/O
Pin TypePower Rail
TDM_SPI_MOSIOCMOSVDDO/
VDD_GE_B
TDM_SPI_MISOICMOSVDDO/
VDD_GE_B
Description
Serial SPI data from the host to the codec for register access.
When TDM_SPI_CS is asserted low, the data is driven from the
host on the negative edge of TDM_SPI_SCK. It is always driven
for eight TDM_SPI_SCK cycles at a time.
In a byte, the data can be driven MSB or LSB first.
Serial SPI read data from the CODEC to the host for register
access.
When TDM_SPI_CS is asserted low, this data is driven from
CODEC on negative edge of TDM_SPI_SCK. It is always driven
for eight TDM_SPI_SCK cycles at a time. The CODEC drives
data on this line only for a read o peration, when it ge ts command
and address in previous bytes from the host on TDM_SPI_MOSI
In a byte, the data can be driven MSB or LSB first.
All of the TS signals are multiplexed on the MPP pins (see Section 4, Pin
Multiplexing, on page 51).
The TS signals are powered on VDDO or on VDD_GE_B based on the pin
multiplexing option (see Section 4, Pin Multiplexing ).
Table 21: Transport Stream (TS) Interface Signal Assignment
Pin Name
I/O
Pin TypePower Rail
TSMP[0]ICMOSVDDO/
VDD_GE_B
TSMP[1]I/OCMOSVDDO/
VDD_GE_B
TSMP[2]I/OCMOSVDDO/
VDD_GE_B
TSMP[3]I/OCMOSVDDO/
VDD_GE_B
TSMP[4]I/OCMOSVDDO/
VDD_GE_B
TSMP[5]I/OCMOSVDDO/
VDD_GE_B
TSMP[6]I/OCMOSVDDO/
VDD_GE_B
Description
EXT_CLK
External clock that can be used to drive the TS0_CLK and
TS1_CLK
TS0_CLK
Port0 TS clock.
• If TS0_VA L is used, the clock may be continuous.
• If TS0_VAL is not used, the clock may toggle onl y when valid
data is available on TS0_DATA.
TS0_SYNC
Port0 Sync/Frame Start Indicator or Packet Clock.
The TS0_SYNC in parallel mode is a pulse that is active during
the first (Sync) byte of the TS packet. In serial mode, the
TS0_SYNC pulse may be active for the entire byte or only for the
first bit. The polarity is programmable to be eit her active high or
active low.
TS0_VAL
Port0 V alid Data Indicator
When this signal is used and is valid, it indicates that valid dat a is
present on TS0_DATA. TS0_VAL is active during the TS frame
packet data and inactive when there is no TS synchronization.
In output mode, the polarity of TS0_VAL is programmable to be
either active high or active low.
TS0_ERR
Port0 Uncorrectable Packet Error
When this signal is used, an error indicates that the packet
contains an uncorrectable error, and therefore should not be
used.
In output mode, the TS0_ERR is active during the entire TS
frame.
TS0_DATA[0]
Port0 TS Data bit 0 in both parallel and serial modes.
In Serial mode TS0_DATA[0] is used as data input or output.
• Parallel Mode:
TS0_DATA[1]: Port0 TS Data bit 1
• Serial Mode:
TS1_CLK: Port1 TS clock.
- If TS1_VAL is used, the clock may be continuous.
- If TS1_VAL is not used, the clock may toggle only when
valid data is available on TS1_DATA
Table 21: Transport Stream (TS) Interface Signal Assignment (Continued)
Pin Name
I/O
Pin TypePower Rail
TSMP[7]I/OCMOSVDDO/
VDD_GE_B
TSMP[8]I/OCMOSVDDO/
VDD_GE_B
TSMP[9]I/OCMOSVDDO/
VDD_GE_B
TSMP[10]I/OCMOSVDDO/
VDD_GE_B
TSMP[11]I/OCMOSVDDO/
VDD_GE_B
TSMP[12]I/OCMOSVDDO/
VDD_GE_B
Description
• Parallel Mode:
TS0_DATA[2]: Port0 TS Data bit 2
• Serial Mode:
TS1_SYNC: Port1 Sync/Frame Start Indicator or Packet
Clock.
The TS1_SYNC pulse may be active for the entire byte or
only for the first bit. The p olarity i s programmable to be e ither
active high or active low
• Parallel Mode:
TS0_DATA[3]: Port0 TS Data bit 3
• Serial Mode:
TS1_VAL: Port1Valid Data Indicator
When this signal is used and is valid, it indicates that valid
data is present on TS1_DATA[0].
TS1_VAL is active during the TS frame packet data and
inactive when there is no TS synchronization.
In output mode, the polarity of TS1_VAL is programmable to
be either active high or active low.
• Parallel Mode:
TS0_DATA[4]: Port0 TS Data bit 4
• Serial Mode:
TS1_ERR: Port1 Uncorrectable Packet Error
When this signal is used, an error indicates that the packet
contains an uncorrectable error, and, therefore, should not
be used.
In output mode the TS1_ERR is active during the entire TS
frame.
• Parallel Mode:
TS0_DATA[5]: Port0 TS Data bit 5
• Serial Mode:
TS1_DATA[0]: Port1 TS Data bit 0, used as data input or
output.
TS0_DATA[6]
Port0 TS Data bit 6
This pin is only valid in Parallel mode.
TS0_DATA[7]
Port0 TS Data bit 7
This pin is only valid in Parallel mode.
Some pins of the device package are connected to internal pull-up and pull-down resistors. When
these pins are Not Connected (NC) on the system board, these resistors set the default value for
input and sample at reset configuration pins.
The internal pull-up and pull-down resistor value is 50 kΩ. An external resistor with a lower value can
override this internal resistor.
Table 23: Internal Pull-up and Pull-down Pins
Pin NamePin NumberPull up/Pull down
GE_TXD[0]H02Pull down
GE_TXD[1]H01Pull down
GE_TXD[2]H03Pull up
GE_TXD[3]H04Pull up
GE_TXCTLJ04Pull down
GE_MDCL03Pull up
JT_TMS_CORET14Pull up
JT_RSTnT15Pull down
JT_TDIR14Pull up
JT_TMS_CPUV15Pull up
NF_ALER10Pull up
NF_REnU11Pull down
NF_CLER11Pull down
NF_CEnV11Pull up
NF_WEnV12Pull up
MRnF04Pull up
MPP[1]V08Pull down
MPP[2]V07Pull down
MPP[3]V09Pull down
MPP[4]T09Pull up
MPP[5]T10Pull up
MPP[7]R06Pull up
MPP[10]R07Pull down
MPP[11]T07Pull up
MPP[12]U12Pull down
MPP[14]V13Pull up
MPP[18]V10Pull up
MPP[19]U10Pull up
MPP[33]N03Pull down
Table 24 lists the signal strapping to be used for systems in which some of the device interfaces are unused (not
connected).
Table 24: Unused Interface Strapping
Unused InterfaceStrapp ing
Ethernet SMIPull up GE_MDIO.
MPP Configure any unused MPP pin to GPIO output.
Leave the power supply connected.
• If the related power supply is VDDO, leave it connected to 3.3V.
• If the related power supply is VDD_GE_B, leave it connected to either 3.3V or 1.8V.
USBDiscard the power filter.
Leave USB_AVDD connected to 3.3V.
All other signals can be left unconnected.
PCI ExpressDiscard the analog power filters.
Leave PEX_AVDD connected to 1.8V.
Pull down the PEX_CLK_N signal through a 50 k
Pull up the PEX_CLK_P signal through a 16 k
All other signals can be left unconnected.
Configure the PEX_CLK_P and PEX_CLK_N signals as inputs, as indicated in Table 32,
Reset Configuration, on page 67.
SATADiscard the analog power filters.
SATA0_AVDD/SATA1_AVDD can be left unconnected.
RTCConnect RTC_AVDD, RTC_AVSS, RTC_XIN, and RTC_XOUT to GND.
SSCGDiscard the power filter.
Leave SSCG_AVDD connected to 1.8V.
eFuseConnect VHV to VDD
SYSRST_OUTn: Reset request from the device to the board reset logic. This pin is an output.
SYSRST_OUTn is the default setting for MPP[6].
PEX_RST_OUTn: Optional PCI Express Endpoint card reset output.
MII/MMII/GMII/RGMII interface signals
SATA0/1_ACTn/SATA0/1_PRESENTn (port 0 and port 1): SATA active and SATA present
indications—see the SATA section in the 88F6180, 88F6190, 88F6192, and 88F6281
Functional Specifications.
NF_IO[7:0] (NAND Flash data [7:0])
SPI interface: SPI_MOSI, SPI_MISO, SPI_SCK, SPI_CSn
UART interface (port 0 and port 1): Transmit and receive functions: UA0_TXD, UA0_RXD,
UA1_TXD, UA1_RXD, and Modem control functions: UA0_RTSn, UA0_CTSn, UA1_RTSn,
UA1_CTSn
For MPPs assigned as NAND flash and SPI flash, wake-up mode after reset
depends on Boot mode (see the Boot Device field in Table 32, Reset
Configuration, on page 67):
• When Boot mode is NAND Flash, MPP[5:0] and MPP[19:18] wake up after reset
in NAND Flash mode.
• When Boot mode is SPI Flash, either MPP[3:0] or {MPP[3:1] and MPP[7]} wake
up after reset in SPI mode, (according to boot mode configured by reset strap
pins).
Pin MPP[6] wakes up after reset in 0x1 mode (SYSRST_OUTn)
Pin MPP[7] wakes up after reset:
• As SPI_CSn, if the boot device—selected according to boot device reset
strapping—is 0x2 (boot from SPI flash, SPI_CSn on MPP[7]).
• As PEX_RST_OUTn, if the boot device—selected according to boot device
reset strapping—is any option other than 0x2.
When TWSI serial ROM initialization is enabled (see TWSI Serial ROM
Initialization in Table 32, Reset Configuration, on page 67), MPP[8] and MPP[9]
wake up as TWSI data and clock pins, respectively.
All other MPP interface pins wake up after reset in 0x0 mode (GPIO/GPO) and are
default set to Data Output disabled (Tri-State). Therefore, those MPPs that are
GPIO are in fact inputs, and those that are GPO are Tri-State.
The SPI interface can be configured using one of the following sets of MPP pins:
• MPP[3:0]
• MPP[11], MPP[10], MPP[7], and MPP[6]
• MPP[3:1] and MPP[7]
Do not configure both MPP[3] and MPP[11] as SPI_MISO.
UART0 and UART1 signals are duplicated on a few MPPs. The UART0 or UART1
signals must not be configured to more than one MPP.
When selecting the MII/MMII interface (MPP[35:20]) and the TDM interface
(MPP[49:35]), the TDM signal TDM_CH0_TX_QL and the MII/MMII signal
MII1_RXERR are both multiplexed on MPP[35]. However, MPP[35] can only be
configured to one of these functions at a time.
Some of the MPP pins are sampled during SYSRSTn de-assertion to set the
device configuration. These pins must be set to the correct value during reset (see
Section 6.5, Pins Sample Configuration, on page 66).
Pins that are left as GPIO and are not connected should be set to output after
When using Gigabit Ethernet signals on MPPs, all relevant Gigabit Ethernet signals
(except those marked as NA) must be implemented. For example, if using MII, and the
chosen PHY does not have an MII_RXERR out signal, the MII_RX_ERR (in) (MPP[35])
must still be configured accordingly and must have a pull-down resistor.
The TS interface can be configured to one of five modes:
One or two serial in interfaces
One or two serial out interfaces
Serial in and serial out interface
Parallel in interface
Parallel out interface
In parallel in or serial in mode, all TS signals are inputs.
In parallel out or serial out mode, all TS signals are outputs.
Table 28 summarizes the TS port pins multiplexing.
NOTE: See Table 32, Reset Configuration, on page 67 for CPU, L2 cache and
DDR frequency configuration.
L2 cache clock frequency must be equal or higher then DDR clock
frequency.
If the SSCG enable bit in the Sampled at Reset register is set, then the
SSCG circuit is applied for the CPU PLL reference clock (refer to the
Sampled at Reset register in the 88F6180, 88F6190, 88F6192, and 88F6281 Functional Specifications).
• Reference clock:
REF_CLK_XIN (25 MHz)
• Derivative clocks:
- TCLK (core clock, 200 MHz)
- SDIO Clock (100 MHz)
- Gigabit Ethernet Clock (125 MHz)
- TS unit Clock(100/91/83/77MHz)
- SPI clock (TCLK/30–TCLK/4 MHz)
- SMI clock (TCLK/128 MHz)
- TWSI clock (up to TCLK/1600)
NOTE: See Table 32, Reset Configuration, on page 67 for TCLK frequency
configuration.
NOTE: See the TS Interface Configuration re gister in the 88F6180, 88F6190,
88F6192, and 88F6281 Functional Specifications for TS clock frequency
There are two options for the ref erence c lock configur a tion, depending on the PCI
Express clock 100 MHz differential clock:
• The device uses an external source for PCI Express clock. The PEX_CLK_P
pin is an input.
• The device uses an internal generated clock for PCI Express clock. The
PEX_CLK_P pin is an output, driving out the PCI Express differential clock.
• Reference clock:
REF_CLK_XIN (25 MHz)
Page 61
Table 29: 88F6281Clocks (Continued)
Clock TypeDescription
Clocking
SATA PHY PLL
RTC
PTP
• Reference clock:
REF_CLK_XIN (25 MHz)
• Derivative clock:
SATA Clock (150 MHz)
• Reference clock:
RTC_XIN (32.768 kHz)
Used for real time clock functionality, see the Real Time Clock section in the
88F6180, 88F6190, 88F6192, and 88F6281 Functional Specifications.
• Reference clock:
PTP_CLK (125 MHz)
The PTP_CLK can be used for the following functions:
• PTP time stamp clock
Two options for reference clock:
- PTP_CLK
- Gigabit Ethernet Clock (125 MHz)
• TS unit clock
Two options for reference clock:
- PTP_CLK/2
- Core PLL
• Audio unit clock
Two options for reference clock:
- PTP_CLK
- REF_CLK_XIN (25 MHz)
For clocking configuration registers, see the 88F6180, 88F6190, 88F6192, and 88F6281 Functional Specifications.
The following table lists the supported combinations of the CPU_CLK Frequency select, CPU_CLK
to DDR CLK ratio, and to CPU_CLK to CPU L2 clock ratio (see Section 6.5, Pins Sample
The SSCG (Spread S pectrum Clock Generator) may be used to generate the spread spectrum clock
for the PLL input. See SSCG Disable in Table 32, Reset Configuration, on page 67, for SSCG
enable/bypass configuration settings.
The SSCG block can be configured to perform up spread, down spread and center spread.
The modulation frequency is configurable. Typical frequency is 30 kHz.
The spread percentage can also be configured up to 1%.
For additional details, see the SSCG Configuration Register description in the 88F6180, 88F6190,
This section provides information about the device power-up/down sequence and configuration at
reset.
6.1Power-Up/Down Sequence Requirements
6.1.1Power-Up Sequence Requirements
These guidelines must be applied to meet the 88F6281 device power-up requirements:
The non-core voltages (I/O and Analog) as listed in Table 31 must reach 70% of their voltage
level before the core voltages reach 70% of their voltage level.
The order of the power-up sequence between the non-core voltages is unimportant so long as
the non-core voltages power up before the core voltages reach 70% of their voltage level
(shown in Figure 2).
The order of the power-up sequence between the core voltages (VDD and VDD_CPU) is
unimportant.
The reset signal(s) must be asserted before the core voltages reach 70% of their voltage level
(shown in Figure 2).
The reference clock(s) inputs must toggle with their respective voltage levels before the core
voltages reach 70% of their voltage level (shown in Figure 2).
If VHV is set to burning mode (2.5V), which is a higher voltage than the VDD voltage, VDD must
be powered before VHV, to prevent the fuse from being accidentally burned.
It is the designer's responsibility to verify that the power sequencing requirements
of other components are also met.
Although the non-core voltages can be powered up any time before the core
voltages, allow a reasonable time limitation (for example, 100 ms) between the
first non-core voltage power-up and the last core voltage power-up.
6.1.2Power-Down Sequence Requirements
There are no special requirements for the core supply to go down before non-core power, or for
reset assertion when powering down (except for VHV, as described below). However, allow a
reasonable time limitation (no more than 100 ms) between the first and last voltage power-down.
When using the eFuse in Burning mode, VHV must be po wered down before VDD.
6.2Hardware Reset
The device has one reset input pin—SYSRSTn. When asserted, the entire chip is placed in its initial
state. Most outputs are placed in high-z, except for the following output pins, that are still active
during SYSRSTn assertion:
Reset (SYSRSTn signal) must be active for a minimum length of 5 ms. core power, I/O
power, and analog power must be stable (VDD +/- 5%) during that time and onward.
6.2.1Reset Out Signal
The device has an optional SYSRST_OUTn output signal, multiplexed on an MPP pin, that is used
as a reset request from the device to the board reset logic. SYSRST_OUTn is the default option for
that MPP pin.
This signal is asserted low for 20 ms, when one of the following maskable events occurs:
Received hot reset indication from the PCI Express link (only relevant when used as a PCI
Express endpoint), and bit <PexRstOutEn> is set to 1 in the RSTOUTn Mask Register(see the
Reset register section of the 88F6180, 88F6190, 88F6192, and 88F6281 Functional Specifications).
PCI Express link failure (only relevant when used as a PCI Express endpoint), and bit
<PexRstOutEn> is set to 1 in the RSTOUTn Mask Register.
Watchdog timer expiration and bit <WDRstOutEn> is set to 1 in the RSTOUTn Mask Register.
Bit <SystemSoftRst> is set to 1 in System Soft Reset Register and bit <SoftRstOutEn> is set
to 1 in RSTOUTn Mask Register.
This signal is asserted low for 20 ms, when one of the following non-maskable events occurs:
Power on reset (The device includes a power-on-reset (POR) circuit for VDD power.)
SYSRST_OUTn is asserted low as long as the MRn input signal is asserted low and for an
additional 20 ms after MRn de-assertion. (This is useful for implementations that include a
manual reset button.)
System Power Up/Down and Reset Settings
Hardware Reset
6.2.2Power On Reset (POR)
The SYSRST_OUTn output signal is asserted low for 20 ms, when the power-on-reset (POR) circuit
is triggered.
POR is triggered when VDD power up (digital core voltage) reaches a VDD threshold (threshold
maximum value 0.8V).
Hysteresis: Another trigger will only occur after the power first drops to 50 mV, and then a power up
occurs.
6.2.3SYSRSTn Duration Counter
When SYSRSTn is asserted low, a SYSRSTn duration counter is running.
The counter clock is the 25 MHz reference clock.
It is a 29-bit counter, yielding a maximum counting duration of 2^29/25 MHz (21.4 seconds).
The host software can read the counter value and reset the counter.
When the counter reach its maximum value, it remains at this value until counter reset is
triggered by software.
The SYSRSTn duration counter is useful for implementing manufacturer/factory reset.
Upon a long reset assertion, greater than a pre-configured threshold, the host software
may reset all settings to the factory default values.
As a Root Complex, the device may generate a Hot Reset to the PCI Express port. Upon CPU
setting the PCI Express Control register’s <conf_mstr_hot_reset> bit, the PCI Express unit sends a
Hot Reset indication to the Endpoint, see the PCI Express Interface section in the 88F6180, 88F6190, 88F6192, and 88F6281 Functional Specifications.
6.3.2PCI Express Endpoint Reset
When a Hot Reset packet is received:
A maskable interrupt is asserted.
If the <conf_dis_hot_rst_reg_rst> field in the PCI Express Debug Control register is cleared, the
device also resets the PCI Express regist er fi l e to its default values.
The device triggers an internal reset, if not masked by the <conf_msk_hot_reset> field in the
PCI Express Debug Control register.
Link failure is detected if the PCI Express link was up (LTTSSM L0 state) and dropped back to an
inactive state (LTSSM Detect state). When Link failure is detected:
A maskable interrupt is asserted.
If the <conf_dis_link_fail_reg_rst> field in the PCI Express Debug Control register is cleared,
the device also resets the PCI Express register file to its default values.
The device triggers an internal reset, if the <conf_msk_link_fail> field is not masked by PCI
Express Debug Control register.
Both link fail and hot reset conditions trigger a chip internal reset (if not masked in the PCI Express
interface). All the chip logic is reset to the default values, except for sticky registers and the sample
on reset logic. In addition, these events can trigger reset to the board, using one of the following:
PEX_RST_OUTn signal (multiplexed on MPP).
SYSRST_OUTn output (multiplexed on MPP)—if not masked by the <PexRstOutEn> bit.
The external reset logic (on the board) may assert the SYSRSTn input pin and reset the entire chip.
6.4Sheeva™ CPU TAP Controller Reset
The Sheeva™ CPU Test Access Port (TAP) controller is reset when JT_RSTn is set and
JT_TMS_CPU is active.
6.5Pins Sample Configuration
The following pins are sampled during SYSRSTn de-assertion:
Internal pull up/down resistors set the default mode (see Section 1.3, Internal Pull-up and
Pull-down Pins, on page 48).
Higher value, external pull up/down resistors are required to change the default mode of
operation.
These signals must remain pulled up or down until SYSRSTn de-assertion (zero hold time in respect
to SYSRSTn de-assertion).
If external logic is used instead of pull-up and pull-down resistors, the logic must
drive all of these signals to the desired values during SYSRSTn assertion. To
prevent bus contention on these pins, the external logic must float the bus no later
than the third TCLK cycle after SYSRSTn de-assertion.
All reset sampled values are registered in the Sample at Reset register (see the
MPP Registers in the 88F6180, 88F6190, 88F6192, and 88F6281 Functional
Specifications). This is useful for board debug purposes and identification of board
and system settings for the host software.
If a signal is pulled up on the board, it must be pulled to the proper voltage level.
Certain reset configuration pins are powered by VDD_GE_A and VDD_GE_B.
Those pins have multiple voltage options (see Table 36, Recommended Operating
Conditions, on page 77).
In each row of Table 32, the order of the pins is from MSb to LSb (e.g., for in the row CPU_CLK
Frequency Select, MPP[2] is the MSB and MPP[10] is the LSB).
Table 32: Reset Configuration
PinConfiguration Function
System Power Up/Down and Reset Settings
Pins Sample Configuration
MPP[1]TWSI Serial ROM Initialization
0 = Disabled
1 = Enabled
NOTE: Internally pulled down to 0x0.
When this pin is set to 0x1, MPP[8] and MPP[9] wake up as TWSI data and clock pins,
respectively (see Section 4.1, Multi-Purpose Pins Functional Summary, on page 51).
MPP[2],MPP[5],
CPU_CLK Frequency Select
MPP[19],
MPP[10]
0x0–0x6 = Reserved
0x7 = 1000 MHz
0x8 = Reserved
0x9 = 1200 MHz
0xA–0xB = Reserved
0xC = 1500 MHz
0xD–0xF = Reserved
NOTE: Internally pulled to 0x6.
The supported combination for CPU_CLK Frequency select, CPU_CLK to DDR CLK ratio,
and CPU_CLK to CPU L2 clock ratio are listed in Table 30, Supported Clock Combinations,
The supported combination for CPU_CLK Frequency select, CPU_CLK to DDR CLK ratio,
and CPU_CLK to CPU L2 clock ratio are listed in Table 30, Supported Clock Combinations,
0x0 = Reserved
0x1 = Reserved
0x2 = Boot from SPI flash (SPI_CSn on MPP[7])
0x3 = Reserved
0x4 = Boot from SPI flash (SPI_CSn on MPP[0])
0x5 = Boot from NAND flash
0x6 = Boot from SATA
0x7 = Boot from the PCI Express port
NOTE:
• Internally pulled to 0x4.
• Only SPI signals configured on pins MPP[3:0] or on pins MPP[7] and MPP[3:1] can be used for
booting from SPI.
SPI signals that are multiplexed on other MPPs can only be used after booting (see Section 4.1,
Multi-Purpose Pins Functional Summary, on page 51).
• When GE_TXD[2:0] is set to 0x4, MPP[3:0] wake up as SPI signals.
• When GE_TXD[2:0] is set to 0x2, MPP[7] and MPP[3:1] wake up as SPI signals.
• When GE_TXD[2:0] is set to 0x5, MPP[5:0] and MPP[19:18] wake up as NAND Flash signals.
• For a more detailed descripti on of the bootROM, see the BootROM section in the 88F6180, 88F6190, 88F6192, and 88F6281 Functional Specifications.
• For a more detailed description of the boot from SPI flash or NAND flash, see the SPI Interface
and NAND Flash Interface sections in the 88F6180, 88F6190, 88F6192, a nd 88F6281 Functional Specifications.
• There is an option to boot from UART when GE_ TXD[2:0] = 0x2–0x7. For a more detailed
description of the boot from UART, see the BootROM section in the 88F6180, 88F6190, 88F6192, and 88F6281 Functional Specifications.
NOTE: MUST be externally pulled down to 0x0 during reset.
6.6Serial ROM Initialization
The device supports initialization of ALL of its internal and configuration registers through the TWSI
master interface. If serial ROM initializa ti o n is e nabled, the device TWSI master starts reading
initialization data from serial ROM and writes it to the appropriate registers, upon de-assertion of
SYSRSTn.
When using Serial ROM Initialization, the MPP[9:8] pins must be configured to as TW_SCK
(MPP[9]) and TW_SDA (MPP[8]).
6.6.1Serial ROM Data Structure
Serial ROM data structure consists of a sequence of 32-bit address and 32-bit data pairs, as shown
in Figure 3.
The serial ROM initialization logic reads eight bytes at a time. It performs address decoding on the
32-bit address being read, and based on address decoding result, writes the next four bytes to the
required target.
The Serial Initialization Last Data Register contains the expected value of last serial data item
(default value is 0xFFFFFFFF). When the device reaches last data, it stops the initialization
sequence.
6.6.2Serial ROM Initialization Operation
On SYSRSTn de-assertion, the device starts the initialization process. It first performs a dummy
write access to the serial ROM, with data byte(s) of 0x0, to set the ROM byte offset to 0x0. Then, it
performs the sequence of reads, until it reaches last data item, as shown in Figure 4.
Figure 4: Serial ROM Read Example
Boot Sequence
For a detailed description of TWSI implementation, see the Two-Wire Serial Interface section in the
88F6180, 88F6190, 88F6192, and 88F6281 Functional Specifications.
Initialization data must be programmed in the serial ROM starting at offset 0x0.
The device assumes 7-bit serial ROM address of ‘b1010000.
After receiving the last data identifier (default value is 0xFFFFFFFF), the device receives an
additional byte of dummy data. It responds with no-ack and then asserts the stop bit.
The serial EEPROM must contain two address offset bytes (It must not be less than a 256 byte
ROM.).
6.7Boot Sequence
The device requires that SYSRSTn stay asserted for at least 300 μs after power and clocks are
stable. The following procedure describes the boot sequence starting with the reset assertion:
1.While SYSRSTn is asserted, the CPU PLL and the core PLL are locked.
2.Upon SYSRSTn de-assertion, the pad drive auto-calibration process starts. It takes 512 TCLK
cycles.
3.If Serial ROM initialization is enabled, an initialization sequence is started.
4.If configured to boot from NAND flash (and BootROM is disabled), the device also performs a
NAND Flash boot sequence to prepare page 0 in the NAND fl ash device for read.
Page 72
88F6281
Hardware Specifications
Upon completing the above sequence, the internal CPU reset is de-asserted, and the CPU starts
executing boot code from the boot device (SPI flash, NAND flash, or internal Boot ROM), according
to sample at reset setting, see Table 32, Reset Configuration, on page 67.
For bootROM details, see the BootROM section in the 88F6180, 88F6190, 88F6192, and 88F6281Functional Specifications.
As part of the CPU boot code, the CPU typically performs the following:
Configures the PCI Express address map.
Configures the proper SDRAM controller parameters, and then triggers SDRAM initialization
(sets <InitEn> bit [0] to 1 in the SDRAM Initialization Control register).
Sets the <PEXEn> bits in the CPU Control and Status register to wake up the PCI Express link.
To enable board testing, the device supports a test mode operation through its JTAG boundary scan
interface.
The JTAG interface is IEEE 1149.1 standard compliant. It supports mandatory and optional
boundary scan instructions.
7.1TAP Controller
The Test Access Port (T AP) is constructed with a 5-pin interface and a 16-state Finite State Machine
(FSM), as defined by IEEE JTAG standard 1149.1.
To place the device in a functional mode, reset the JTAG state machine to disable the JTAG
interface.
According to the IEEE 1149.1 standard, the JTAG state machine is not reset when the 88F6281
SYSRSTn is asserted. The JTAG state machine can only be reset by one of the following methods:
Asserting JT_RSTn.
Setting JT_TMS_CORE for at least five JT_CLK cycles.
To place the device in one of the boundary scan test mode, the JT AG state machine must be moved
to its control states. JT_TMS_CORE and JT_TDI inputs control the state transitions of the JTAG
state machine, as specified in the IEEE 1149.1 standard. The JTAG state machine will shift
instructions into the Instruction register while in SHIFT-IR state and shift data into and from the
various data registers when in SHIFT-DR state.
JTAG Interface
TA P Co ntroller
7.2Instruction Register
The Instruction register (IR) is a 4-bit, two-stage register. It contains the command that is shifted in
when the TAP FSM is in the Shift-IR state. When the TAP FSM is in the Capture-IR state, the IR
outputs all four bits in parallel.
Table 33 lists the instructions supported by the device.
Table 33: Supported JTAG Instructions
InstructionCodeDescription
HIGHZ0011Select the single bit Bypass register between TDI and TDO.
Sets the device output pins to high-impedance state.
IDCODE0010 Selects the Identification register between TDI and TDO. This 32-bit
register is used to identify the device.
EXTEST0000Selects the Boundary Scan register between TDI and TDO. Outputs the
boundary scan register cells to drive the output pins of the devi ce. Inputs
the boundary scan register cell to sample the input pin of the device.
SAMPLE/PRE
LOAD
BYPA SS1111Selects the single bit Bypass register between TDI and TDO. This allows
0001Selects the Boundary Scan register between TDI and TDO. Samples
input pins of the device to input boundary scan register cells.
Preloads the output boundary scan register ce lls with the Boundary Scan
register value.
for rapid data movement through an untested device.
The Bypass register (BR) is a single bit serial shift register that connects TDI to TDO, when the IR
holds the Bypass command, and the TAP FSM is in Shift-DR state. Data that is driven on the TDI
input pin is shifted out one cycle later on the TDO output pin. The Bypass register is loaded with 0
when the TAP FSM is in the Capture-DR state.
7.4JTAG Scan Chain
The JTAG Scan Chain is a serial shift register used to sample and drive all of the device pins during
the JTAG tests. It is a 2-bit per pin shift register in th e de vi ce , th ere b y allo w i ng th e sh ift register to
sequentially access all of the data pins both for driving and strobing data. For further details, refer to
the BSDL Description file for the device.
7.5ID Register
The ID register is a 32-bit deep serial shift register. The ID register is loaded with vendor and device
information when the TAP FSM is in the Capture-DR state. The Identification code format of the ID
register is shown in Table 34, which describes the various ID Code fields.
Table 34: IDCODE Register Map
BitsVa l ueDescription
31:280x0Version (4'b0010 for version A0, 4'b0011 for A1, etc.)
27:12 0x6281Part number
Exposure to conditions at or beyond the maximum rating may damage the device.
Operation beyond the recommended operating conditions (Table 36) is neither
Before designing a system, Marvell recommends reading application note AN-63:
Thermal Management for Marvell Technology Products. This application note presents
basic concepts of thermal management for integrated circuits (ICs) and includes
guidelines to ensure optimal operating conditions for Marvell Technology's products.
The purpose of the Thermal Power Dissipation table is to support system engineering in thermal
design.
General comment: See the Pin Description section for internal pullup/pul ldown.
1. While I/O is in High-Z.
2. This current does not include the current flowing through the pullup/pulldown resistor.
DC Electrical Specifications
8.5DC Electrical Specifications
See Section 1.3, Internal Pull-up and Pull-down Pins, on page 48 for internal
pullup/pulldown information.
8.5.1General 3.3V (CMOS) DC Electrical Specifications
The DC electrical specifications in Table 39 are applicable for the following interfaces and signals:
JTAG
RGMII (10/100 Mbps)/GMII/MII/MMII
Secure Digital Input/Output (SDIO)
S/PDIF / I
Transport Stream (TS)
NAND flash
UART
MPP
PTP
SYSRSTn
In the following table, for the JTAG, SDIO, S/PDIF / I
interfaces, VDDIO means the VDDO power rail. For the RGMII/GMII/MII/MMII interface, VDDIO
means the VDD_GE_A and VDD_GE_B power rails.
Table 39: General 3.3V Interface (CMOS) DC Electrical Specifications
General comment: See the Pin D escription section for internal pullup/pulldown.
1. While I/O is in High-Z.
2. This current does not include the current flowing through the pullup/pulldown resistor.
88F6281
Hardware Specifications
8.5.2RGMII, SMI and REF_CLK_XIN 1.8V (CMOS) DC Electrical
Specifications
In the following table, for the RGMII interface, VDDIO means the VDD_GE_A power rail.
In the following table, for the REF_CLK_XIN pin, VDDIO means the XTAL_AVDD power rail.
Table 40: RGMII 1.8V Interface (CMOS) DC Electrical Specifications
Pa rameterSy mbolTest ConditionMinTypMaxUni ts Notes
Input low levelVI L--0.3VREF - 0.125VInput high levelVIH-VREF + 0.125VDDIO + 0.3VOutput low levelVOLIO L = 13.4 mA0.28VOutput high levelVOH IOH = -13.4 mA1.42V-
120150180ohm1 , 2
607590ohm1 , 2
405060ohm1 , 2
Deviation of VM with respect to VDDQ/2dVm See note 3-66%3
Input l eakage currentII L0 < VIN < VDD IO-1010uA4, 5
Pin capacitanceCpin-5pF-
Notes:
General com ment: See the Pin D escription section for internal pullup/pul ldown.
1. See SDR AM functional description section for ODT configuration.
2. M easureme nt definition for RTT: Apply VREF +/- 0.25 to input pin separately,
then measure current I
(VREF + 0.25)
and I
(VREF - 0.25)
respectively.
3. M easureme nt definition for VM: Measured voltage (VM) at input pin (midp oi nt) with no load.
4. While I/O i s in High-Z.
5. This current does not include the current flowing through the pullup/pull down resistor.
Rtt effective impedance valueRTTSee note 2
)25.0()25.0(
5.0
−+
−
=
VRE FVREF
II
RTT
%10 012×
⎟
⎠
⎞
⎜
⎝
⎛
−
×
=
VDDIO
Vm
dVM
DC Electrical Specifications
8.5.3SDRAM DDR2 Interface DC Electrical Specifications
In the following table, VREF is VDD_M/2 and VDDIO means the VDD_M power rail.
Table 41: SDRAM DDR2 Interface DC Electrical Specifications
1.Slew rate is defined from 20% to 80% of the reference clock signal.
2.For additional information regarding configuring this clock, see the Serial Memory Interface
Control Register in the
88F6180, 88F6190, 88F6192, and 88F6281 Functional Specifications.
3. Fs is the audio sample rate, which can be configured to 44.1 kHz, 48 kHz, or 96 kHz (see the
Audio (I
2
S / S/PDIF) Interface section in the 88F6180, 88F6190, 88F6192, and 88F6281
Functional Specifications).
4.The RTC design was optimized for a standard CL = 12.5 pF crystal. No passive components
are provided internally. Connect the crystal and the passive network as recommended by the
crystal manufacturer.
5.The frequency can be set using the TS Interface Configuration register (see the 88F6180,
88F6190, 88F6192, and 88F6281
6. For the minimum value refer to the Baud Rate Register section of the 88F6180, 88F6190,
88F6192, and 88F6281
Functional Specifications.
7.The Ethernet Reference Clock parameters refer both to the reference clock for an Ethernet port
configured using the dedicated port pins and for an Ethernet port configured using the
multiplexed port pins.
Clock frequencyfCKMHzDQ and DM valid output time before DQS transitiontDOVB0.40-nsDQ and DM valid output time after DQS transitiontDOVA0.40-nsDQ and DM output pulse w idthtDIPW0.35-tCK(avg)DQS output high pulse widthtDQSH0.35-tCK(avg)DQS output low pulse widthtDQSL0.35-tCK(avg)DQS falling edge to CLK-CLKn rising edgetDSS0.34-tCK(avg)1
DQS falling edge from CLK-CLKn rising edgetDSH0.34-tC K(avg)1
DQS latching rising transitions to associated clock edgestDQSS-0.110.11tCK (avg)DQS write preambletWPRE0.35-tCK(avg)DQS write postambletWPST0.40-tC K(avg)Average CLK-CLK n high-level w idthtCH(avg)0.480.52tCK(avg) 1, 2, 3
Average CLK-CLK n low-level w idthtCL(avg)0.480.52tC K(avg) 1, 2, 4
DQ input setup time relative to DQS in transitiontDSI-0.42-nsDQ input hold time relative to DQS in transitiontDH I0.70-nsAddress and control output pulse widthtIPW0.60-tCK(avg)-
Note s:
General comment: All timing values are defined from Vref to Vref, unless otherw ise specified.
General comment: All input timing values assum e minimum slew rate of 1 V/ns (slew rate defined from Vref +/-125 mV).
General comment: tCK(avg) is calculated as the average clock period across any consecutive 200 cycle w indow.
General comment: All timing parameters with DQS signal are defined on DQS -DQSn crossing point.
General comment: For Address and Control output tim ing parameters, refer to the Address Ti ming table.
General comment: For all signals, the load is CL = 14 pF.
1. This timing value is defined on CLK / CLKn crossing point.
2. Refer to SDRAM DD R2 clock specifications table for more information.
3. tCH(avg) is defined as the average HIGH pulse width, as calculated across any consecutive 200 HIGH pulses.
4. tCL(avg) is defined as the average LOW pulse width, as calculated across any consecutive 200 LOW pulses.
Address and Control valid output time before CLK-CLkn rising edgetAOVB0.65-ns1, 2
Address and Control valid output time after CLK-CLKn rising edgetAOVA0.65-ns1, 2
Address and Control valid output time before CLK-CLkn rising edgetAOVB2.95-ns1, 3
Address and Control valid output time after CLK-CLKn rising edgetAOVA0.65-ns1, 3
Notes :
General comment: All timing values were measured from vref to vref, unless otherw ise specified.
General comment: For all signals, the load is CL = 14 pF.
1. This timing value is defined on CLK / CLKn crossing point.
2. This timing value is defined when Address and Control signals are output on CLK-CLKn falling edge.
For more inform ation, see register settings.
3. This timing value is defined when Address and Control signals are output on CLK-CLKn falling edge.
and 2T mode is enabled. For more information, see register settings.
Except for ODT, CKE and CS signals.
Clock period jittertJIT(per)-100100ps1
Clock perior jitter during DLL locking periodtJIT(per,lck)-8080ps2
Cycle to cycle clock period ji ttertJIT(cc)-200200ps3
Cycle to cycle clock period ji tter during DLL locking periodtJIT(cc,lck)-160160ps4
Cumulative error acros s 2 cyclestERR( 2per)-150150ps5
Cumulative error acros s 3 cyclestERR( 3per)-175175ps5
Cumulative error acros s 4 cyclestERR( 4per)-200200ps5
Cumulative error acros s 5 cyclestERR( 5per)-200200ps5
Cumulative error acros s n cycles, n=6...10, i nclusivetERR(6-10per)-300300ps5
Cumulative error acros s n cycles, n=11…50, inclusivetERR(11- 50per)-450450ps5
Duty cycle jittertJIT(duty)-100100ps6
Absolute clock periodtCK(abs)ps7
Absolute clock high pulse w idthtCH(abs)ps8
Absolute clock low pulse w idthtCL(abs)ps9
Notes:
General comment: All timing values are defined on CLK / CLK n crossing point, unless otherwise specified.
1. tJIT(per) is defined as the largest deviation of any single tCK from tCK(avg).
tJIT(per) = Mi n/max of {tCKi- tCK(avg) where i=1 to 200}.
tJIT(per) defines the single period jitter when the DLL is already locked.
2. tJIT(per,lck) uses the same definition for single period jitter, during the DLL locking period only.
3. tJIT(cc) is defined as the difference in clock period between two consecutive clock cycles: tJIT(cc) = Max of |tCKi+1 – tCKi|.
tJIT(cc) defines the cycle to cycle jitter when the DLL is already locked.
4. tJIT(cc,lck) uses the same definition for cycle to cycle jitter, during the DLL locking period only.
5. tERR is defined as the cumulative error across multiple consecutive cycles from tCK(avg).
Please refer to JEDEC Standard No. 79-2C (DDR2 SDRAM Specification), Chapter 5 (page 100) for more information.
6. tJIT(duty) is defined as the cumulative set of tCH jitter and tCL jitter. tCH jitter is the largest deviation of
any single tCH from tCH(avg). tCL jitter is the largest deviation of any single tCL from tCL(avg).
tJIT(duty) = Min/max of {tJIT(CH), tJIT(CL)} where,
tJIT(CH) = {tCHi- tCH (avg) w here i=1 to 200}; tJIT(CL) = {tCLi- tC L(avg) w here i=1 to 200}.
Clock frequencyfCKMHzData to Clock output skewTskewT-0.500.50ns2
Data to Clock inpu t skewTskewR 1.002.60nsClock cycle durationTcyc 7.208.80n s1 , 2
Duty cycle fo r Gigabi tDuty_G 0.450.55tCK2
Duty cycle fo r 10/100 MegabitDuty_T 0.400.60tCK2
Notes:
General comment: All values w ere measured from vddio/2 to vddio/2, unless otherw ise specified.
General comment: tCK = 1/fCK.
General comment: If the PHY does not support internal-delay mode, the PC board design requires
routing clocks so that an additional trace delay of greater than 1.5 ns and less
than 2.0 ns is added to the associated clock signal.
For 10/100 Mbps RG MII, the Max value is unspecified.
1. For RGMII at 10 Mbps and 100 Mbps, Tcyc will scale to 400 ns +/-40 ns and 40 ns +/-4 ns, respectively.
2. For all signals, the load is CL = 5 pF.
125.0
DescriptionSym bolMinMax Units Notes
Clock frequencyfCKMHzData to Clock output skewTskew T-0.500.50ns2
Data to Clock inpu t skewTskewR 1.002.60nsClock cycle durationTcyc 7.208.80ns1 , 2
Duty cycle fo r G igabitDuty_G 0.450.55tCK2
Duty cycle fo r 10/100 MegabitDuty_T 0.400.60tCK2
Notes :
General comment: All values w ere measured from vddio/2 to vddio/2, unless otherwise specified.
General comment: tCK = 1/fCK.
General comment: If the PHY does not support internal-delay mode, the PC board design requires
routing clocks so that an additional trace delay of greater than 1.5 ns
is added to the associated clock signal.
For 10/100 Mbps RGMII, the Max value is unspecified.
1. For RGMII at 10 Mbps and 100 Mbps, Tcyc will scale to 400 ns +/-40 ns and 40 ns +/-4 ns, respectively.
2. For all signals, the load is C L = 5 pF.
25.0
AC Electrical Specifications
8.6.3Reduced Gigabit Media Independent Interface (RGMII)
AC Timing
8.6.3.1RGMII AC Timing Table
Table 49: RGMII 10/100/1000 AC Timing Table at 1.8V
GTX_CLK cycle timetCK7.58.5nsRX_CLK cycle ti metCKrx7.5-nsGTX_CLK and RX_CLK high level widthtHIGH2.5-ns1
GTX_CLK and RX_CLK low level widthtLOW2.5-ns1
GTX_C LK and RX_CLK rise timetR-1.0ns1, 2
GTX_C LK and RX_CLK fall timetF-1.0ns1, 2
Data input setup time relative to RX_CLK rising edgetSETUP2.0-nsData input hold time relative to RX_CLK rising edgetHOLD0.0-nsData output valid before GTX_CLK rising edgetOVB2.5-ns1
Data output valid after GTX_CLK rising edgetOVA0.5-ns1
Notes :
General comment: All values w ere measured from VIL(max) to VIH(min), unless otherw ise specified.
1. For all signals, the load is C L = 5 pF.
2. Rise time measured from VIL(max) to VIH(min), fall time measured from VIH(m in) to VIL(max).
Note s
125 MHz
DescriptionSym bolUnits
CL
Test Point
AC Electrical Specifications
8.6.4Gigabit Media Independent Interface (GMII) AC Timing
Data input setup relative to RX _CLK rising edgetSU3.5-nsData input hold relative to RX_CLK rising edgetHD2.0-nsData output delay relative to M II_TX_CLK rising edgetOV0.010.0ns1
Notes :
General comment: All values were measured from VIL(max) to VIH(min), unless otherw ise specified.
1. For all signals, the load is CL = 5 pF.
CL
Test Point
MII_TX_CLK
TXD, TX_EN, TX_ER
Vih(min)
Vil(max)
Vih(min)
Vil(max)
TOV
AC Electrical Specifications
8.6.5Media Independent Interface/Marvell Media Independent
Interface (MII/MMII) AC Timing
8.6.5.1MII/MMII MAC Mode AC Timing Table
Table 52: MII/MMII MAC Mode AC Timing Table
8.6.5.2MII/MMII MAC Mode Test Circuit
Figure 14: MII/MMII MAC Mode Test Circuit
8.6.5.3MII/MMII MAC Mode AC Timing Diagrams
Figure 15: MII/MMII MAC Mode Output Delay AC Timing Diagram