ETC 84221

84221
84221
Quad 100BaseTX/10BaseT
Physical Layer Device
PRELIMINARY
99191
Features
Note: Check for latest Data Sheet revision before
starting any designs.
■ Single Chip 100BaseTX/10BaseT
Physical Layer Solution
SEEQ Data Sheets are now on the Web, at
www.lsilogic.com
■ Four Independent Channels in One IC
This document is an LSI Logic document. Any
reference to SEEQ Technology should be considered LSI Logic.
■ 3.3V Power Supply with 5V Tolerant I/O
■ Dual Speed - 10/100 Mbps
■ Half and Full Duplex
■ Reduced Pin Count MII (RMII) Interface to
Ethernet Controller
Description
The 84221 is a highly integrated Ethernet Transceiver for
twisted pair and fiber Ethernet applications. The 84221
can be configured for either 100BaseTX or 10BaseT
Ethernet operation.
■ MI Interface for Configuration and Status
■ Optional Repeater Interface
■ AutoNegotiation for 10/100, Full/Half Duplex
Hardware Controlled Advertisement
The 84221 consists of four (4) separate and independent
channels. Each channel consists of: 4B5B/Manchester
encoder, scrambler, transmitter with wave shaping and onchip filters, transmit output driver, receiver with adaptive
equalizer, filters, baseline wander correction, clock and
data recovery, descrambler, 4B5B/Manchester decoder,
and controller interface (MII or RMII).
■ Meets all Applicable IEEE 802.3, 10BaseT,
100BaseTX Standards
■ On Chip Wave Shaping - No External Filters
Required
■ Adaptive Equalizer for 100BaseTX
■ Baseline Wander Correction
The addition of internal output waveshaping circuitry and
on-chip filters eliminates the need for external filters
normally required in 100BaseTX and 10BaseT
applications.
■ LED Outputs
Link
Activity
Collision
Full Duplex
10/100
The 84221 can automatically configure itself for 100 or 10
Mbps and Full or Half Duplex operation, for each channel
independently, using the on-chip AutoNegotiation
algorithm.
■128L PQFP
The 84221 can access eleven 16-bit registers for each
channel through the Management Interface (MI) serial
port. These registers comply to Clause 22 of IEEE 802.3u
and contain configuration inputs, status outputs, and
device capabilities.
The 84221 is ideal as a media interface for 100BaseTX/
10BaseT switching hubs, repeaters, routers, bridges, and
other multi port applications.
The 84221 is implemented in a low power CMOS
technology and operates with a 3.3V power supply.
1
MD400184/A
84221
103 RXER_3
105 TXD0_3
104 TXEN_3
107 CRS_DV3
106 TXD1_3
108 GND
110 MDIO
109 VDD
112 MDC
111 REGDEF
114 DPLX_3
113 VDD
115 DPLX_2
117 DPLX_0
116 DPLX_1
119 GND
118 VDD
120 RESET
SPEED_1
121 SPEED_0
122
123 LED3_0
124 LED3_1
126 LED3_3
125 LED3_2
127 CLKIN
128 SPEED_2
1.0 Pin Configuration
SPEED_3
1
102 NC
GND
2
101 RXD0_3
GND
3
100 RXD1_3
TPIP_3
4
99
VDD
TPIN_3
5
98
GND
VDD
6
97
CRS_DV2
TPOP_3
7
96
NC
GND
8
95
ANEG
VDD
9
94
LEDDEF
TPON_3
10
93
TXD1_2
GND
11
92
TXD0_2
VDD
12
91
TXEN_2
TPON_2
13
90
NC
GND
14
89
REPEATER
VDD
15
88
RXER_2
TPOP_2
16
87
NC
TPIN_2
17
86
RXD0_2
TPIP_2
18
85
RXD1_2
REXT
19
84
VDD
GND
20
83
GND
TPIP_1
21
82
CRS_DV1
TPIN_1
22
81
NC
VDD
23
80
GND
TPOP_1
24
79
PHYAD4
GND
25
78
PHYAD3
VDD
26
77
TXD1_1
TPON_1
27
76
TXD0_1
GND
28
75
TXEN_1
VDD
29
74
NC
TPON_0
30
73
PHYAD2
GND
31
72
RXER_1
VDD
32
71
NC
TPOP_0
33
70
RXD0_1
TPIN_0
34
69
RXD1_1
TPIP_0
35
68
CRS_DV0
NC
36
67
NC
GND
37
66
VDD
LED0_3
38
65
VDD
52
53
54
55
56
57
58
59
60
61
62
63
64
VDD
AD_REV
GND
VDD
RXD1_0
RXD0_0
RXER_0
NC
TXEN_0
TXD0_0
TXD1_0
GND
51
VDD
50
GND
LED1_1
LED2_0
46
LED0_1
49
45
GND
LED1_0
44
LED2_2
47
43
LED1_2
48
42
LED0_2
LED2_1
41
LED2_3
LED0_0
39
40
LED1_3
84221
128 Pin PQFP
Top View
2
MD400184/A
84221
1.0 PIN DESCRIPTION
Power Supplies
Pin #
Pin Name
I/O
Description
6
9
12
15
23
26
29
32
52
53
56
65
66
84
99
109
113
118
VDD
—
Positive Supply. +3.3 +/-5% Volts.
2
3
8
11
14
20
25
28
31
37
44
51
55
64
80
83
98
108
119
GND
—
Ground. 0 Volts.
36
60
67
71
74
81
87
90
96
102
NC
—
No Connect. Reserved for future use, must be left floating for proper
operation.
3
MD400184/A
84221
1.0 PIN DESCRIPTION (cont’d)
Media Interface
Pin #
Pin Name
I/O
Description
7
16
24
33
TPOP_[3:0]
O
Twisted Pair Transmit Output, Positive.
10
13
27
30
TPON_[3:0]
O
Twisted Pair Transmit Output, Negative.
4
18
21
35
TPIP_[3:0]
I
Twisted Pair Receive Input, Positive.
5
17
22
34
TPIN_[3:0]
I
Twisted Pair Receive Input, Negative.
19
REXT
—
Transmit Current Set.
An external resistor connected between this pin and GND will set the level
for the transmit outputs.
4
MD400184/A
84221
1.0 PIN DESCRIPTION (cont’d)
Controller Interface (RMII)
Pin #
Pin Name
I/O
Description
127
CLKIN
I
Clock Input.
This controller interface input latches controller interface data in and out of
the device on rising edges for all channels. There must be a 50 Mhz clock
tied to this pin.
104
91
75
61
TXEN_[3:0]
I
Transmit Enable Input. These interface inputs must be be asserted active
high to allow data on TXD and TXER to be clocked in on the rising edges of
CLKIN.
TXD[1:0]_3
TXD[1:0]_2
TXD[1:0]_1
TXD[1:0]_0
I
Transmit Data Input. These interface inputs contain input di-bit data to be
transmitted on the TP outputs and are clocked in on rising edges of CLKIN.
CRS_DV[3:0]
O
Carrier Sense Output. These interface outputs are asserted active high
when valid data is detected on the receive TP inputs and is clocked out on
the rising edge of CLKIN.
RXD[1:0]_3
RXD[1:0]_2
RXD[1:0]_1
RXD[1:0]_0
O
Receive Data Output. These interface outputs contain recovered di-bit data
from the TP inputs and are clocked out on the rising edges of CLKIN.
RXER_[3:0]
O
Receive Error Output. These interface outputs are asserted active high
when coding or other specified errors are detected on the TP inputs and are
clocked out on rising edges of CLKIN.
[106:105]
[93:92]
[77:76]
[63:62]
107
97
82
68
[100:101]
[85:86]
[69:70]
[57:58]
103
88
72
59
5
MD400184/A
84221
1.0 PIN DESCRIPTION (cont’d)
Management Interface (MI)
Pin #
Pin Name
I/O
Description
112
MDC
I
Management Interface (MI) Clock Input. This MI clock shifts serial data
into and out of MDIO on rising edges.
110
MDIO
I/O
Management Interface (MI) Data Input/Output. This bidirectional pin
contains serial data that is clocked in and out on rising edges of the MDC
clock.
111
REGDEF
I
Pullup
Invalid Register Read Select
This active low input controls the default values that are read from invalid
register locations.
If REGDEF = ‘1’ All invalid register locations will return a value of ‘0000’
when read.
If REGDEF = ‘0’ All invalid register locations will return a value of ‘ffff’ when
read.
79
78
73
PHYAD[4:2]
I
MI Physical Device Address Input. These pins set the three most
significant bits of the PHY address.
The two least significant bits of the PHY address are set internally to match
the channel number, as shown below:
Channel 3
Channel 2
Channel 1
Channel 0
6
MD400184/A
PHYAD1
PHYAD0
1
1
0
0
1
0
1
0
84221
1.0 PIN DESCRIPTION (cont’d)
LED Drivers
Pin #
Pin Name
I/O
Description
126
125
124
123
LED3_[3:0]
O
LED Output. The default functions of these pins are 100 Mbps Link Detect
outputs assuming LEDDEF = 0. These pins can drive an LED from both
VDD and GND.
40
43
47
50
LED2_[3:0]
39
42
46
49
LED1_[3:0]
38
41
45
48
LED0_[3:0]
94
LEDDEF
Please refer to Table 1a. for LED description
O
LED Output. The default functions of these pins are Activity Detect outputs
assuming LEDDEF = 0. These pins can drive an LED from both VDD and
GND.
Please refer to Table 1a. for LED description
O
LED Output. The default functions of these pins are Full Duplex Detect
outputs assuming LEDDEF = 0. These pins can drive an LED from both
VDD and GND.
Please refer to Table 1a. for LED description
O
LED Output. The default functions of these pins are 10 Mbps Link Detect
outputs assuming LEDDEF = 0. These pins can drive an LED from both
VDD and GND.
Please refer to Table 1a. for LED description
LED Default Select Input. This pin changes the default selection for the
I
Pulldown LED’s in the MI Serial Port Global Configuration Register.
1 = LINK + ACT, COL, FDX, 10/100
0 = LINK100, ACT, FDX, LNK10
7
MD400184/A
84221
1.0 PIN DESCRIPTION (cont’d)
Miscellaneous
Pin #
95
Pin Name
ANEG
I/O
I
Description
AutoNegotiation Enable Input. This digital input, ANDed with register bit
0.12, enables AutoNegotiation for all channels.
1 = AutoNegotiation On & Combined with Speed and Duplex pins,
control advertisement. Please refer to Table 1 for the
different combinations.
0 = Off
1
128
122
121
SPEED_[3:0]
114
115
116
117
DPLX_[3:0]
I
Speed Selection Input. These digital inputs, ANDed with register bit 0.13,
select speed in each corresponding channel. Please refer to Table 1 for the
different combinations.
1 = 100 Mbps Mode
0 = 10 Mbps Mode
I
Duplex Selection Input. These digital inputs, ORed with register bit 0.8,
select the duplex mode in EACH corresponding channel. They control
advertisement when ANEG is enabled. Please refer to Table 1 for the
different combinations.
1 = Full Duplex Mode
0 = Half Duplex Mode
89
REPEATER
I
Repeater Mode Enable Input. This digital input, ORed with register bit
17.14, enables repeater mode for ALL channels.
1 = Repeater Mode Enabled
0 = Normal Operation
54
120
AD_REV
RESET
I
Pullup
Address Reverse Input.
I
Pullup
Hardware Reset Input.
1 = Normal
In this mode, physical ports 0-3 are mapped to MI
addresses 0-3 in the same order.
0 = Reverse Address Mode Select
In this mode, physical ports 0-3 are mapped to MI addresses 3-0
respectively. This is the reverse to the normal order.
1 = Normal
0 = Device In Reset State
(Reset is complete 50 ms after RESET goes high).
8
MD400184/A
84221
Table 1. AutoNegotiation, Speed & Duplex Mode Combinations
Input Pins
Aneg
Speed[3:0]
Mode Selected
Duplx [3:0]
Autoneg
Forced
Modes
[Note 2]
Advertised
Capabilities
Speed
DPLX
Speed
DPLX
0
0000
0000
Off
10
Half
na
na
0
0000
1111
Off
10
Full
na
na
0
1111
0000
Off
100
Half
na
na
0
1111
1111
Off
100
Full
na
na
1
0000
0000
On
na
na
10
Half
1
0000
1111
On
na
na
10
Half/Full
1
1111
0000
On
na
na
10/100
Half/Full[3]
1
1111
1111
On
na
na
10/100
Half
Note 1: The single ANEG pin applies to all four channels. The four Speed and DPLX apply to each individual channel.
The pins above are shown for all four channels, but each channel can be individually configured for SPEED and DPLX by
appropriately setting the pin for that channel.
Note 2: Forced Modes assume that registers 0 and 4 are at default values.
Note 3: the 84221 can be either controlled through the software or through the hardware. If the device needs to be controlled through the software (Registers 0 to 4), then these seven pins (ANEG, Speed[3:0] , Duplx[3:0]) have to be tied to
their default values of 1, 1111, and 0000 respectively.
9
MD400184/A
84221
Table 1a. LED Definitions as per the LEDDEF pin
Name
Output State Description
LED tied
to GND
LED tied
to Vdd
LEDDEF
LED0
0 = 100 Mbps Link
Detected
Off
On
LEDDEF = 1
(10/100)
1 = 10 Mbps Link
Detected
On
Off
Trisate = No Link
Off
Off
LED1
0 = Full Duplex Mode
Detect with Link Pass
Off
On
(Fdx/Hdx)
1 = Half Duplex Mode
Detect with Link Pass
On
Off
Tristate = No Link
Off
Off
LED2
(Col)
0 = Collision Detect
1 = No Collision
Off
On
On
Off
LED3
0 = Link Detect
Off
On
(Link + Act)
Blink = Link Detect + Activity
Blink
Blink
1 = No Link Detect or
Activity
On
Off
LED0[1]
0 = 10 Mbps Link
Detected
Off
On
(Link10)
1 = No 10 Mbps Link
Detected
On
Off
LED1
0 = Full Duplex Mode
Detect with Link Pass
Off
On
(Fdx/Hdx)
1 = Half Duplex Mode
Detect with Link Pass
On
Off
Tristate = No Link
Off
Off
LED2
Blink = Activity Occurred
(Stretch pulse to
100 ms)
Off
On
(Act)
1 = No Activity
On
Off
LED3
0 = Link100 Detected
Off
On
1 = No Link100
On
Off
(Link100)
Detected
10
MD400184/A
LEDDEF = 0
MD400184/A
LEDDEF
LED0_3
LED1_3
LED2_3
LED3_3
RXER_3
CRS_DV[3:0]
RXD[1:0]_3
TXEN_3
TXD[1:0]_3
CLKIN
REGDEF
PHYAD[4:2]
MDIO
MDC
RESET
REPEATER
AD_REV
DPLX_[3:0]
SPEED_[3:0]
ANEG
LED
DRIVERS
RMII
CONTROLLER
INTERFACE
SERIAL
PORT
(MI)
4B5B
DECODER
MANCHESTER
ENCODER
4B5B
ENCODER
Per Channel Functions
CLOCK & DATA
RECOVERY
(MANCHESTER
DECODER)
AUTONEGOTIATION
& LINK
CLOCK & DATA
RECOVERY
CLOCK
GENERATOR
ROM
10BT TRANSMITTER
MLT3
ENCODER
DAC
LP
FILTER
LP
FILTER
SQUELCH
10BTX RECEIVER
MLT3
DECODER
SQUELCH
100BTX RECEIVER
CLOCK
GENERATOR
CURRENT
SOURCE
Figure 1. 84221 Block Diagram
Channel 0
Channel 1
Channel 2
Channel 3
DESCRAMBLER
SCRAMBLER
100BASETX TRANSMITTER
–
+
+
11
+/– Vth
–
+
Global
–
+
–
+
LP
FILTER
ADAPTIVE
EQUALIZER
TPIN_3
TPIP_3
TPON_3
TPOP_3
Global
REXT
84221
84221
2.0 Functional Description
2.1 GENERAL
• The 10 Mbps mode uses two pairs of category 3, or
better, UTP or STP twisted pair cable with Manchester
encoded 10 MHz binary data to achieve a 10 Mbps
thruput.
The 84221 is a complete 10/100 Mbps Ethernet Media
Interface IC. The 84221 has four separate and
independent channels. Each channel has the following
main sections: controller interface, encoder, decoder,
scrambler, descrambler, clock and data recovery, twisted
pair interface transmitter, twisted pair interface receiver,
and auto negotiation. A Management Interface (MI) serial
port, which provides access to eleven registers for each
channel, is common to all four channels. Figure 1 shows
the 84221 block diagram.
The data symbol format on the fiber or twisted pair cable
for the 100 and 10 Mbps modes is defined in IEEE 802.3
specifications and shown in Figure 2.
On the transmit side for 100 Mbps operation, data is
received on the controller interface from an external
Ethernet controller per the format shown in Figure 3. The
data is sent to the encoder for formatting. For TX
operation, the encoded data is sent to the scrambler. The
encoded and scrambled data is then sent to the TX
transmitter. The transmitter converts the encoded and
scrambled data into MLT3 ternary format. The transmitter
then preshapes the output and drives the twisted pair
cable.
The 84221 can operate as a 100BaseTX device (100
Mbps mode) or as a 10BaseT device (10 Mbps mode).
The 100 Mbps and 10 Mbps modes differ in data rate,
signalling protocol, and allowed wiring as follows:
• The 100 Mbps TX mode uses two pairs of category 5, or
better, UTP or STP twisted pair cable with 4B5B
encoded, scrambled, and MLT3 coded (ternary) 125
MHz data to achieve a throughput of 100 Mbps.
ETHERNET MAC FRAME
INTERFRAME
GAP
PREAMBLE
SFD
DA
SA
LN
LLC DATA
FCS
LN
LLC DATA
FCS
INTERFRAME
GAP
100 BASE-TX DATA SYMBOLS
IDLE
SSD
PREAMBLE
SFD
DA
SA
IDLE = [ 1 1 1 1 ...]
SSD = [ 1 1 0 0 0 1 0 0 0 1 ]
PREAMBLE
SFD
DA, SA, LN, LLC DATA, FCS
ESD
=
=
=
=
ESD
IDLE
SOI
IDLE
BEFORE / AFTER
4B5B ENCODING,
SCRAMBLING, AND
MLT3 CODING
[ 1 0 1 0 ...] 62 BITS LONG
[1 1 ]
[ DATA ]
[0 1 1 0 1 0 0 1 1 1]
10 BASE-T DATA SYMBOLS
IDLE
PREAMBLE
SFD
DA
SA
LN
IDLE = [ NO TRANSITIONS ]
PREAMBLE = [ 1 0 1 0 ... ] 62 BITS LONG
SFD = [ 1 1 ]
DA, SA, LN, LLC DATA, FCS = [ DATA ]
SOI = [ 1 1 ] WITH NO MID BIT
TRANSITION
Figure 2. Frame Format
12
MD400184/A
LLC DATA
FCS
BEFORE / AFTER
MANCHESTER
ENCODING
84221
On the receive side for 100BaseTX operation, the TX
receiver removes any high frequency noise from the input,
equalizes the input signal to compensate for the low pass
effects of the cable, and qualifies the data with a squelch
algorithm. The TX receiver then converts the data from
MLT3 coded twisted pair levels to internal digital levels.
The output of the receiver then goes to a clock and data
recovery block which recovers a clock from the incoming
data, uses the clock to latch in valid data into the device,
and converts the data back to NRZ data. The data is then
unscrambled and decoded by the 4B5B decoder and
descrambler, respectively, and output to an external
Ethernet controller by the controller interface.
can directly connect to any Ethernet controller that also
complies with the RMII specifications.
2.2.2 100 Mbps
The RMII consists of eight signals: (1) two transmit data
inputs (TXD[1:0]), (2) a transmit enable input (TXEN), (3)
two receive data outputs (RXD[1:0]), (4) a carrier sense/
data valid input (CRS), (5) a receive data error input
(RXER), and (6) global clock input (CLKIN). The global
clock, CLKIN, is a common signal for all eight channels.
All other signals are replicated for each channel. CLKIN
operates at a frequency of 50 MHz.
Data in RMII is transmitted/received in 2-bit wide word,
called di-bits, on TXD[1:0]/RXD[1:0], respectively. The
RMII frame format and bit order is defined in the RMII
Consortium specifications and is also shown in Figure 3.
10 Mbps operation is similar to the 100 Mbps operation,
except:
• There is no scrambler/descrambler.
• The encoder/decoder is Manchester instead of 4B5B.
On the transmit side, the CLKIN input runs continuously at
50 MHz. When no data is to be transmitted, TXEN must
be deasserted. While TXEN is deasserted, TXD[1:0] are
ignored and no data is clocked into the device. When
TXEN is asserted, it is clocked in on the rising edge of
CLKIN along with the data on TXD[1:0]. TXD[1:0] input
data is two bit wide packet data whose format needs to be
the same as specified in the RMII Consortium specification and also shown in Figure 3. When all packets have
been latched into the device, TXEN must be deasserted.
• The data rate is 10 Mbps instead of 100 Mbps.
• The twisted pair symbol data is two level Manchester
instead of ternary MLT3.
The AutoNegotiation block automatically configures each
channel for either 100BaseTX or 10BaseT, and either Full
or Half Duplex operation. This configuration is based on
the capabilities selected for the channel and capabilities
detected from the remote device connected to the
channel.
On the receive side, when no packets are being received,
CRS is deasserted and RXD[1:0] is held low. When the
start of packet is detected, CRS is asserted active high on
the rising edge of CLKIN and RXD[1:0]=00 until the
receive packet data is available to be output onto
RXD[1:0]. When the packet data is available to be output,
it is clocked out onto RXD[1:0] on rising edges of CLKIN
while CRS still remains asserted high. When the end of
packet is detected on the receive input, CRS toggles high
and low at a 25 MHz rate until all of the packet data has
been output onto the RXD[1:0] pins. Once all of the
packet data has been output on RXD[1:0], then CRS is
deasserted on rising edges of CLKIN. If dribble bits occur
on the receive input, a non-integer number of data octets
will be output on RXD[1:0]. If the device is in the Link Fail
state, CRS always stays deasserted. The packet data on
RXD[1:0] has the same frame structure and bit order as
the TXD[1:0] data and is specified in the RMII Consortium
specification and also shown in Figure 3.
The Management Interface (the MI serial port) is a two pin
bidirectional link through which configuration inputs can be
set and channel status outputs read.
Each block plus the operating modes are described in
more detail in the following sections. Since the 84221 can
operate as a 100BaseTX or a 10BaseT device, each of
the following sections describes the performance in both
100 and 10 Mbps modes.
2.2 CONTROLLER INTERFACE (RMII)
2.2.1 General
The controller interface is the interface between the device
and an external Ethernet Media Access Controller (MAC).
The specific controller interface type implemented on the
device is the Reduced Pin Count Media Independent
Interface, also referred to as RMII. The RMII is a two bit
wide packet data interface between an Ethernet PHY and
MAC that was defined by an industry group, the RMII Consortium. The RMII is a reduced pin count version of the
IEEE 802.3 MII. The 84221 meets all the RMII requirements outlined in the RMII Consortium specifications and
An elastic buffer is present in the receive data path
because the input receive data is referenced to the recovered clock and the receive data output on RXD0 has to be
referenced to CLKIN. The receive buffer is 32 bits in
length. Receive input data fills the elastic buffer to a predetermined threshold level before data is passed to the
13
MD400184/A
84221
TXEN = 1
TXEN = 0
IDLE
PREAMBLE
START
OF
FRAME
PRMBLE
SFD
62 BITS
2 BITS
PREAMBLE
SFD
DATAn
IDLE
TXEN = 0
IDLE
DATA DI_BITS
DATA 1
=
=
=
=
DATA N -1 DATA N
DATA 2
[ 1 0 1 0 ... ] 62 BITS LONG]
[11]
[BETWEEN 64-1518 DATA BYTES]
TXEN = 0
a.) MAC Frame Format
MAC’s SERIAL BIT STREAM
FIRST BIT
LSB
D0
D1
1st DI_BIT
D4
D3
D2
2nd DI_BIT
D5
3rd DI_BIT
MSB
D7
D6
4th DI_BIT
TXD0 / RXD0
TXD1 / RXD1
b.) RMII DI_BIT Order
Signals
Bit Value
TXD0
X
X
11
1
.
.
.
.
.
.
.
.
.
.
12
1
1
1
D03 D2 D4 D6
TXD1
X
X
0
0
.
.
.
.
.
.
.
.
.
.
0
0
0
1
D1 D3 D5 D74
TXEN
0
0
1
1
.
.
.
.
.
.
.
.
.
.
1
1
1
1
1.
2.
3.
4.
1
1
1
1
1st preamble di-bit transmitted.
1st sfd di-bit transmitted.
1st data di-bit transmitted.
D0 through D7 are the first 8 bits of the data field.
c.) Transmit Preamble and SFD Bits
Signals
Bit Value
RXD0
0
0
0
0
.
.
.
0
11
1
.
.
.
.
12
1
1
1
D03 D3 D4 D6
RXD1
0
0
0
0
.
.
.
0
0
0
.
.
.
.
0
0
0
1
D1 D4 D5 D74
CRS_DV
0
0
1
1
.
.
.
1
1
1
.
.
.
.
1
1
1
1
1. 1st preamble di-bit received. Device may eliminate either all or some of the preamble di-bit,
up to 1st SFD.
2. 1st sfd di-bit received.
3. 1st data di-bit received.
4. D0 through D7 are the first 8 bits of the data field.
d.) Receive Preamble and SFD Bits
Figure 3. RMII Frame Format
14
MD400184/A
1
1
1
1
84221
RMII receive outputs. This predetermined threshold level
can be configured to be either 4 bits or 16 bits by appropriately setting the receive buffer threshold select bit in the
MI serial port Global Configuration register. If the elastic
buffer underflows or overflows, RXER is asserted and
RXD[1:0]=01 for all data bits from the underflow/overflow
detect point until the end of the packet.
Table 2. 4B/5B Symbol Mapping
RXER is a receive error output that is asserted when certain errors are detected on the receive data. RXER is
asserted on rising edge of CLKIN for all data bits of the
packet starting from the RXER detect point until the end of
the packet.
In addition, any packet that contains an
RXER will substitute RXD[1:0]=01 for all the data bits from
the RXER detect point until the end of packet.
There is no collision indication for RMII. Collision can be
sensed external to the device when both TXEN and CRS
are asserted at the same time. Also, there is no internal
TXEN loopback in RMII mode.
2.2.3 10 Mbps
10 Mbps RMII operation is identical to 100 Mbps RMII
operation, except: (1) Each transmit data di-bit must be
input on TXD[1:0] for ten consecutive CLKIN cycles, and
(2) Each receive data di-bit will be output on RXD[1:0] for
ten consecutive CLKIN cycles.
2.2.4 RMII Disable
The RMII inputs and outputs can be disabled by setting
the MII disable bit in the MI serial port Control register.
When the RMII is disabled, the inputs are ignored, the outputs are placed in high impedance state, and the TP output is high impedance.
Symbol
Name
Description
5B Code
4B Code
0
Data 0
11110
0000
1
Data 1
01001
0001
2
Data 2
10100
0010
3
Data 3
10101
0011
4
Data 4
01010
0100
5
Data 5
01011
0101
6
Data 6
01110
0110
7
Data 7
01111
0111
8
Data 8
10010
1000
9
Data 9
10011
1001
A
Data A
10110
1010
B
Data B
10111
1011
C
Data C
11010
1100
D
Data D
11011
1101
E
Data E
11100
1110
F
Data F
11101
1111
I
Idle
11111
0000
0101
J
SSD #1
11000
K
SSD #2
10001
0101
T
ESD #1
01101
0000
R
ESD #2
00111
0000
H
Halt
00100
Undefined
---
Invalid codes
All others*
0000*
* These 5B codes are not used. For decoder, these 5B codes
are decoded to 4B 0000. For encoder, 4B 0000 is encoded
to 5B 11110, as shown in symbol 0.
2.3 ENCODER
2.3.1 4B5B Encoder - 100 Mbps
100BaseTX requires that this data be 4B5B encoded.
4B5B coding converts the 4 bit data nibbles into 5 bit
data words. The mapping of the 4B nibbles to the 5B code
words is specified in IEEE 802.3 and shown in Table 2.
The 4B5B encoder on the 84221 takes 4B nibbles from
the controller interface, converts them into 5B words
according to, Table 2, and sends the 5B words to the
scrambler. The 4B5B encoder also substitutes the first
eight bits of the preamble with the SSD delimiters (/J/K/
symbols) and adds an ESD delimiter (/T/R/ symbols) to
the end of each packet, as defined in IEEE 802.3 and
shown in Figure 2.
The 4B5B encoder also fills the
period between packets, called the idle period, with a
continuous stream of idle symbols, as shown in Figure 2.
2.3.2 Manchester Encoder - 10 Mbps
The Manchester encoding process combines clock and
NRZ data such that the first half of the data bit contains
the complement of the data, and the second half of the
data bit contains the true data, as specified in IEEE 802.3.
This guarantees that a transition always occurs in the
middle of the bit cell. The 84221 Manchester encoder
converts the 10 Mbps NRZ data from the controller
interface into a single data stream for the TP transmitter
and adds a start of idle pulse (SOI) at the end of the
packet as specified in IEEE 802.3 and shown in Figure 2.
15
MD400184/A
84221
equalizer has settled, the PLL input is switched to the
incoming data stream. The PLL then recovers a clock by
locking onto the transitions of the incoming signal. The
recovered clock is output to the Controller Interface block.
The Manchester encoding process is only done on actual
packet data, and the idle period between packets is not
Manchester encoded, but filled with link pulses.
2.5.2 Data Recovery - 100 Mbps
2.4 DECODER
Data recovery is performed by latching in data from the
receive inputs with the recovered clock extracted by the
PLL.
2.4.1 4B5B Decoder - 100 Mbps
Since the TX input data is 4B5B encoded on the transmit
side, it must also be decoded by the 4B5B decoder on the
receive side. The mapping of the 5B nibbles to the 4B
code words is specified in IEEE 802.3 and shown in Table
2. The 84221 4B5B decoder takes the 5B code words
from the descrambler, converts them into 4B nibbles per
Table 2, and sends the 4B nibbles to the RMII controller
interface. The 4B5B decoder also strips off the SSD
delimiter (/J/K/ symbols) and replaces them with two 4B
Data 5 nibbles (/5/ symbol), and strips off the ESD
delimiter (/T/R/ symbols) and replaces it with two 4B Data
0 nibbles (/I/ symbol), per IEEE 802.3 specifications and
shown in Figure 2.
2.5.3 Clock Recovery - 10 Mbps
The clock recovery process for 10 Mbps mode is identical
to the 100 Mbps mode, except:
• The PLL is switched from CLKIN to the TP input when
the squelch indicates valid data.
• The PLL locks onto the preamble signal in less than 12
transitions (bit times).
• Some of the preamble data symbols are lost while the
PLL is locking onto the preamble, however, the data
receiver block recovers enough preamble symbols to
pass at least 3 bytes of preamble to the controller
interface as shown in Figure 3.
The 4B5B decoder detects SSD, ESD and, codeword
errors in the incoming data stream as specified in IEEE
802.3. These errors are indicated by asserting RXER
output while the errors are being transmitted across
RXD[1:0].
2.5.4 Data Recovery
The data recovery process for 10 Mbps mode is identical
to the 100 Mbps mode.
2.4.2 Manchester Decoder - 10 Mbps
In Manchester coded data, the first half of the data bit
contains the complement of the data, and the second half
of the data bit contains the true data. The 84221
Manchester decoder converts the single data stream from
the TP receiver into NRZ data for the controller interface
by decoding the data and stripping off the SOI pulse.
Since the clock and data recovery block has already
separated the clock and data from the TP receiver, the
Manchester decoding process to NRZ data is inherently
performed by that block.
2.6 SCRAMBLER
2.6.1 100 Mbps
100BaseTX requires scrambling to reduce the radiated
emmisions on the twisted pair. The 84221 scrambler
takes the encoded data from the 4B5B encoder,
scrambles it per the IEEE 802.3 specifications, and sends
it to the TP transmitter. The scrambler circuitry of the
84221 is designed so that none of the individual scrambler
sections on-chip will be synchronous with the others to
minimize EMI issues.
2.5 CLOCK AND DATA RECOVERY
2.5.1 Clock Recovery - 100 Mbps
2.6.2 10 Mbps
Clock recovery is done with a PLL. If there is no valid data
present on the receive inputs, the PLL is locked to the
CLKIN clock. When valid data is detected on the receive
inputs with the squelch circuit and when the adaptive
A scrambler is not used in 10 Mbps mode.
16
MD400184/A
84221
The waveform generator eliminates the need for any
external filters on the TP transmit output. The line driver
converts the shaped and smoothed waveform to a current
output that can drive 100 meters of category 5 unshielded
twisted pair cable or 150 ohm shielded twisted pair cable.
2.7 DESCRAMBLER
2.7.1 100 Mbps
The 84221 descrambler takes the scrambled data from the
data recovery block, descrambles it per the IEEE 802.3
specifications, aligns the data on the correct 5B word
boundaries, and sends it to the 4B5B decoder.
2.8.2 10 Mbps
The TP transmitter operation in 10 Mbps mode is much
different than the 100 Mbps transmitter. Even so, the
transmitter still consists of a waveform generator and line
driver.
The algorithm for synchronization of the descrambler is
the same as the algorithm outlined in the IEEE 802.3
specification. Once the descrambler is synchronized, it
will maintain synchronization as long as enough
descrambled idle pattern 1’s are detected within a given
interval. To stay in synchronization, the descrambler
needs to detect at least 25 consecutive descrambled idle
pattern 1’s in a 1 mS interval. If 25 consecutive
descrambled idle pattern 1’s are not detected within the 1
mS interval, the descrambler goes out of synchronization
and restarts the synchronization process.
The purpose of the waveform generator is to shape the
output transmit pulse. The waveform generator consists
of a ROM, DAC, clock generator, and filter. The DAC
generates a stair-stepped representation of the desired
output waveform. The stairstepped DAC output then
goes through a low pass filter in order to "smooth" the
DAC output and remove any high frequency components.
The DAC values are determined from the ROM output; the
ROM outputs are chosen to shape the pulse to the
desired template and are clocked into the DAC at high
speed by the clock generator. In this way, the waveform
generator preshapes the output waveform to be
transmitted onto the twisted pair cable to meet the pulse
template requirements outlined in IEEE 802.3 Clause 14
and also shown in Figure 4. The waveshaper replaces and
eliminates external filters on the TP transmit output.
2.7.2 10 Mbps
A descrambler is not used in 10 Mbps mode.
2.8 TWISTED PAIR TRANSMITTER
The line driver converts the shaped and smoothed
waveform to a current output that can drive 100 meters of
category 3/4/5 100 Ohm unshielded twisted pair cable or
150 Ohm shielded twisted pair cable, without any external
filters. During the idle period, no output signal is
transmitted on the TP outputs (except link pulse).
2.8.1 100 Mbps
The TP transmitter consists of an MLT3 encoder,
waveform generator and line driver.
The MLT3 encoder converts the NRZI data from the
scrambler into a three level code required by IEEE 802.3.
MLT3 coding uses three levels and converts 1's to
transitions between the three levels, and converts 0's to no
transitions or changes in level.
2.9 TWISTED PAIR RECEIVER
The purpose of the waveform generator is to shape the
transmit output pulse. The waveform generator takes the
MLT3 three level encoded waveform and uses an array of
switched current sources to control the shape of the
twisted pair output signal in order to meet IEEE 802.3
requirements. The output of the switched current sources
then goes through a low pass filter in order to "smooth"
the output and remove any high frequency components.
In this way, the waveform generator preshapes the output
waveform transmitted onto the twisted pair cable to meet
the pulse template requirements outlined in IEEE 802.3.
2.9.1 Receiver - 100 Mbps
The TP receiver detects input signals from the twisted pair
input and converts it to a digital data bit stream ready for
clock and data recovery. The receiver can reliably detect
data from a 100Base-TX compliant transmitter that has
been passed through 0-100 meters of 100 Ohm category
5 UTP.
17
MD400184/A
84221
The device stays in the unsquelch state until loss of data
is detected. Loss of data is detected if no alternating
polarity unsquelch transitions are detected during any
10 uS interval. When the loss of data is detected, the
receive squelch level is re-established.
The 100 Mbps receiver consists of an adaptive equalizer,
baseline wander correction circuit, comparators, and MLT3 decoder. The TP inputs first go to an adaptive equalizer.
The adaptive equalizer compensates for the low pass
characteristic of the cable, and it has the ability to adapt
and compensate for 0-100 meters of category 5, 100 Ohm
UTP. The baseline wander correction circuit restores the
DC component of the input waveform that was removed by
external transformers. The comparators convert the
equalized signal back to digital levels and are used to
qualify the data with the squelch circuit. The MLT-3
decoder takes the three level MLT-3 digital data from the
comparators and converts it back to normal digital data to
be used for clock and data recovery.
2.9.4 Squelch - 10 Mbps
The TP squelch algorithm for 10 Mbps mode is identical
to the 100 Mbps mode, except:
• The 10 Mbps squelch algorithm is not used for link
integrity, but to sense the beginning of a packet.
• The receiver goes into the unsquelch state if the input
voltage exceeds the squelch levels for three bit times
with alternating polarity within a 50-250 nS interval.
2.9.2 Receiver - 10 Mbps
• The receiver goes into the squelch state when SOI is
detected.
The 10 Mbps mode receiver is much simpler than the 100
Mbps mode receiver and is identical to the 100 Mbps
receiver except:
• Unsquelch detection has no affect on link integrity, link
pulses are used for that in 10 Mbps mode.
• The adaptive equalizer is disabled and bypassed.
• Start of packet is determined when the receiver goes
into the unsquelch state and CRS is asserted.
• The baseline wander correction circuit is disabled.
• The 10 Mbps receiver is able to detect input signals
from the twisted pair cable that are within the template
specified in IEEE 802.3 Clause 14 and shown in Figure
5.
• The receiver meets the squelch requirements defined in
IEEE 802.3 Clause 14.
• The output of the squelch comparator is used for
squelch, link pulse detect, SOI detect, reverse polarity
detect.
The receiver squelch and unsquelch levels can be
lowered by 4.5 dB by setting the receive level adjust bit in
the MI serial port Channel Configuration register. By
setting this bit, the device can support cable lengths
exceeding 100 meters.
2.9.5 Receive Level Adjust
• The data comparator is a zero crossing comparator
whose output is used for clock and data recovery.
2.9.3 Squelch - 100 Mbps
The squelch block determines whether the input contains
valid data. The 100 Mbps TX squelch is one of the criteria
used to determine link intergrity. The squelch comparators
compare the TX inputs against fixed positive and negative
thresholds, called squelch levels.
2.10 COLLISION
The collision function is not present in the 84221. The
MAC should be able to reliably detect collision based on
the TXEN and CRS_DV signals.
The output from the squelch comparator goes to a digital
squelch circuit, which determines whether the receive
input data on that channel is valid. If the data is invalid,
the receiver is in the squelched state. If the input voltage
exceeds the squelch levels at least four times with
alternating polarity within a 10 uS interval, the data is
considered to be valid by the squelch circuit and the
receiver now enters into the unsquelch state.
2.11 START OF PACKET
2.11.1 100 Mbps
Start of packet for 100 Mbps mode is indicated by a
unique Start of Stream Delimiter (SSD). The SSD pattern
consists of the two /J/K/ 5B symbols inserted at the
beginning of the packet in place of the first two preamble
symbols, as defined in IEEE 802.3 Clause 24 and shown
in Table 2 and Figure 2.
In the unsquelch state, the receive threshold level is
reduced by approximately 30% for noise immunity reasons
and is called the unsquelch level. When the receiver is in
the unsquelch state the input signal is considered valid.
18
MD400184/A
84221
The transmit ESD is generated by the 4B5B encoder and
the /T/R/ symbols are inserted by the 4B5B encoder after
the end of the transmit data packet, as shown in Figure 2.
The transmit SSD is generated by the 4B5B encoder and
the /J/K/ symbols are inserted by the 4B5B encoder at the
beginning of the transmit data packet in place of the first
two 5B symbols of the preamble, as shown in Figure 2.
The receive ESD pattern is detected by the 4B5B
decoder by examining groups of 10 consecutive code bits
(two 5B words) from the descrambler during valid packet
reception to determine whether there is an ESD.
The receive pattern is detected by the 4B5B decoder by
examining groups of 10 consecutive code bits (two 5B
words) from the descrambler. Between packets, the
receiver will be detecting the idle pattern, which is 5B /I/
symbols. While in the idle state, CRS_DV is deasserted.
If the 10 consecutive code bits from the receiver during
valid packet reception consist of the /T/R/ symbols, the
end of packet is detected, data reception is terminated,
CRS_DV is deasserted, and /I/I/ symbols are substituted
in place of the /T/R/ symbols.
If the receiver is in the idle state and 10 consecutive code
bits from the receiver consist of the /J/K/ symbols, the start
of packet is detected, data reception is begun, CRS_DV is
asserted, and /5/5/ symbols are substituted in place of the
/J/K/ symbols.
If 10 consecutive code bits from the receiver during valid
packet reception do not consist of /T/R/ symbols, but
consist of /I/I/ symbols instead, the packet is considered
to have been terminated prematurely and abnormally.
When this premature end of packet condition is detected,
RXER remains asserted for the nibble associated with the
first /I/ symbol detected and then RXER and CRS_DV are
all deasserted.
If the receiver is in the idle state and 10 consecutive code
bits from the receiver consist of a pattern that is neither /I/
I/ nor /J/K/ symbols but contains at least 2 non-contiguous
0's, then activity is detected but the start of packet is
considered to be faulty and a False Carrier Indication (also
referred to as bad SSD) is signalled to the controller
interface. When False Carrier is detected while CRS_DV
is asserted, RXD[1:0]=11, 10 for two di-bits while RXER is
asserted. Once a False Carrier Event is detected, the idle
pattern (two /I/I/ symbols) must be detected before any
new SSD’s can be sensed.
2.12.2 10 Mbps
The end of packet for 10 Mbps mode is indicated with the
SOI (Start of Idle) pulse. The SOI pulse is a positive
double wide pulse containing a Manchester code violation
inserted at the end of every packet.
If the receiver is in the idle state and 10 consecutive code
bits from the receiver consist of a pattern that is neither /I/
I/ nor /J/K/ symbols but does not contain at least 2 non
contiguous 0's, the data is ignored and the receiver stays
in the idle state.
The transmit SOI pulse is generated by the TP transmitter
and inserted at the end of the data packet after TXEN is
deasserted. The transmitted SOI output pulse at the TP
output is shaped by the transmit waveshaper to meet the
pulse template requirements specified in IEEE 802.3
Clause 14 and shown in Figure 6.
2.11.2 10 Mbps
Since the idle period in 10 Mbps mode is defined to be no
data on the TP inputs, then the start of packet for 10 Mbps
mode is detected when valid data is detected by the TP
squelch circuit. When start of packet is detected, CRS is
asserted as described in the Controller Interface Section.
Refer to the TP Squelch - 10 Mbps Section for the
algorithm for valid data detection.
The receive SOI pulse is detected by the TP receiver by
sensing missing data transitions. Once the SOI pulse is
detected, data reception is ended and CRS and RXDV are
deasserted.
2.13 LINK INTEGRITY & AUTONEGOTIATION
2.12 END OF PACKET
2.13.1 General
2.12.1 100 Mbps
The 84221 can be configured to implement either the
standard link integrity algorithms or the AutoNegotiation
algorithm.
End of packet for 100 Mbps mode is indicated by an End
of Stream Delimiter (referred to as ESD). The ESD pattern
consists of the two /T/R/ 4B5B symbols inserted after the
end of the packet, as defined in IEEE 802.3 Clause 24 and
shown in Table 2 and Figure 2.
19
MD400184/A
84221
N
B
1.0
0.8
P
H
VOLTAGE (V)
0.6
I
D
0.4
O
E
C
0.2
0.0
Q
A
M
F
R
J
–0.2
S
–0.4
U
LK
–0.6
W
V
–0.8
G
–1.0
0
10
20
30
40
50
60
70
80
TIME (ns)
Reference
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U
V
W
Time (ns)
Internal MAU
0
15
15
25
32
39
57
48
67
89
74
73
61
85
100
110
111
111
111
110
100
110
90
Voltage (V)
0
1.0
0.4
0.55
0.45
0
-1.0
0.7
0.6
0
-0.55
-0.55
0
1.0
0.4
0.75
0.15
0
-0.15
-1.0
-0.3
-0.7
-0.7
Figure 4. TP Output Voltage Template - 10 Mbps
20
MD400184/A
90
100
T
110
84221
2.13.3 100Base-TX Link Integrity Algorithm
The standard link integrity algorithms are used solely to
establish an active link to and from a remote device.
There are different standard link integrity algorithms for 10
and 100 Mbps modes. The AutoNegotiation algorithm is
used for two purposes:
Since 100Base-TX is defined to have an active idle signal,
then there is no need to have separate link pulses like
those defined for 10Base-T. The 84221 uses the squelch
criteria and descrambler synchronization algorithm on the
input data to determine if the device has successfully
established a link with a remote device (called Link Pass
state). Refer to IEEE 802.3 for details on both algorithms.
• To automatically configure the device for either 10/100
Mbps and Half/Full Duplex modes
• Establish an active link to and from a remote device
2.13.4 AutoNegotiation Algorithm
The standard link integrity and AutoNegotiation algorithms
are described below.
As stated previously, the AutoNegotiation algorithm is
used for two purposes:
2.13.2 10Base-T Link Integrity Algorithm
• To automatically configure the device for either 10/100
Mbps and Half/Full Duplex modes
The 84221 uses the same 10Base-T link integrity
algorithm that is defined in IEEE 802.3 clause 14. This
algorithm uses normal link pulses, referred to as NLP’s
and transmitted during idle periods, to determine if a
device has successfully established a link with a remote
device (called Link Pass state). The transmit link pulse
meets the template defined in IEEE 802.3 Clause 14 and
shown in Figure 7. Refer to IEEE 802.3 Clause 14 for
more details if needed.
• To establish an active link to and from a remote device
The AutoNegotiation algorithm is the same algorithm that
is defined in IEEE 802.3 Clause 28. AutoNegotiation uses
a burst of link pulses, called fast link pulses and referred to
as FLP’s, to pass up to 16 bits of signaling back and forth
between the 84221 and a remote device. The transmit
a. Short Bit
3.1 V
Slope 0.5 V/ns
585 mV
585 mV sin (
* t/PW)
PW
0
b. Long Bit
3.1 V
Slope 0.5 V/ns
585 mV
585 mV sin (
* t/PW)
585 mV sin [2
0
PW/4
(t - PW2)/PW]
3PW/4
PW
Figure 5. TP Input Voltage Template - 10 Mbps
21
MD400184/A
84221
FLP pulses meet the template specified in IEEE 802.3 and
shown in Figure 7. A timing diagram contrasting NLP’s
and FLP’s is shown in Figure 8.
The AutoNegotiation algorithm is initiated by any of the
following events:
• Powerup
• Device Reset
• AutoNegotiation Reset
• Entering the Link Fail state
have to be high to enable AutoNegotiation. When the
AutoNegotiation algorithm is enabled, the device halts all
transmissions including link pulses for 1200-1500 mS,
enters the Link Fail state, and restarts the negotiation
process. When the AutoNegotiation algorithm is disabled,
the selection of 100 Mbps or 10 Mbps mode is determined
by the speed select bit in the MI serial port Control
register, and the selection of Half or Full Duplex is
determined by the duplex select bit in the MI serial port
Control register.
2.13.8 AutoNegotiation Reset
Once a negotiation has been initiated, the 84221 first
determines if the remote device has AutoNegotiation
capability. If the device is not AutoNegotiation capable
and is just transmitting either a 10Base-T or 100Base-TX
signal, the 84221 will sense that and place itself in the
correct mode. If the 84221 detects FLP’s from the remote
device, then the remote device is determined to have
AutoNegotiation capability and the device then uses the
contents of the MI serial port AutoNegotiation
Advertisement register and FLP’s to advertise it’s
capabilies to a remote device. The remote device does
the same, and the capabilities read back from the remote
device are stored in the MI serial port AutoNegotiation
Remote End Capability register. The 84221 negotiation
algorithm then matches it’s capabilities to the remote
devices capabilities and determines to what mode the
device should be configured according to the priority
resolution algorithm defined in IEEE 802.3 Clause 28.
Once the negotiation process is completed, the 84221
then configures itself for either 10 or 100 Mbps mode and
either Full or Half Duplex modes (depending on the
outcome of the negotiation process), and it switches to
either the 100Base-TX or 10Base-T link integrity
algorithms (depending on which mode was enabled by
AutoNegotiation). Refer to IEEE 802.3 Clause 28 for more
details.
2.13.5 AutoNegotiation Outcome Indication
The outcome or result of the AutoNegotiation process is
stored in the speed detect and duplex detect bits in the MI
serial port Status Output register.
The AutoNegotiation algorithm can be initiated at any time
by setting the AutoNegotiation reset bit in the MI serial
port Control register.
2.13.9 Link Indication
Receive link detect activity can be monitored through the
link detect bit in the MI serial port Status and Status
Output registers or it can also be programmed to appear
on LED status pins by appropriately setting the
programmable LED select bits in the MI serial port
Configuration 2 register as shown in Table 3. Whenever
the LED Status pins are programmed to be a link detect
output, these pins are asserted low whenever the device
is in the Link Pass state.
2.13.10 Link Disable
The link integrity function can be disabled by setting the
link disable bit in the MI serial port Configuration 1
register. When the link integrity function is disabled, the
device is forced into the Link Pass state, configures itself
for Half/Full Duplex based on the value of the duplex bit in
the MI serial port Control register, configures itself for 100/
10 Mbps operation based on the values of the speed bit in
the MI serial port Control register, and continues to
transmit NLP’s or TX idle patterns, depending on whether
the device is in 10 or 100 Mbps mode.
2.14 JABBER
2.14.1 100 Mbps
The jabber function is disabled in the 100 Mbps mode.
2.13.6 AutoNegotiation Status
2.14.2 10 Mbps
The status of the AutoNegotiation process can be
monitored by reading the AutoNegotiation Acknowledgement Bit in the MI serial port Status register.
A jabber condition occurs when the transmit packet
exceeds a predetermined length. When jabber is
detected, the TP transmit outputs are forced to the idle
state, collision is asserted, and jabber register bits in the
MI serial port Status and Channel Status Output registers
are set.
2.13.7 AutoNegotiation Enable
The AutoNegotiation algorithm can be enabled (or
restarted) by setting the AutoNegotiation enable bit in the
MI serial port Control register or by asserting the ANEG
pin. The AutoNegotiatiopn enable bit and ANEG pin both
22
MD400184/A
84221
2.17 10/100 MBPS SELECTION
2.15 RECEIVE POLARITY CORRECTION
2.15.1 100 Mbps
2.17.1 General
No polarity detection or correction is needed in 100 Mbps
mode.
The device can be forced into either the 100 or 10 Mbps
mode, or the device can detect 100 or 10 Mbps
capability from a remote device and automatically place
itself in the correct mode.
2.15.2 10 Mbps
The polarity of the signal on the TP receive input is
continuously monitored. If one SOI pulse indicates
incorrect polarity on the TP receive input, the polarity is
internally determined to be incorrect, and the reverse
polarity bit is set in the MI serial port Channel Status
Output register.
The device can be forced into either the 100 or 10 Mbps
mode by either setting the speed select bit in the MI serial
port Control register or by setting the SPEED pin with
AutoNegotiation disabled. Both the speed select bit and
SPEED pin need to be set to the same speed (10 or 100)
for the device to be properly configured. The speed select
bit and SPEED pin are ignored if AutoNegotiation is
enabled.
The 84221 will automatically correct for the reverse
polarity condition provided that the autopolarity feature is
not disabled.
The device can automatically configure itself for 100 or 10
Mbps mode by using the AutoNegotiation algorithm to
advertise and detect 100 and 10 Mbps capabilities to and
from a remote device. Refer to the LINK INTEGRITY &
AUTONEGOTIATION Section for more details on
AutoNegotiation.
2.16 FULL DUPLEX MODE
2.16.1 100 Mbps
Full Duplex mode allows transmission and reception to
occur simultaneously.
The device can be either forced into Half or Full Duplex
mode, or the device can detect either Half or Full Duplex
capability from a remote device and automatically place
itself in the correct mode.
2.17.2 10/100 MBPS Indication
The device speed (100/10 Mbps) can be monitored
through the speed bit in the MI serial port Channel Status
Output register.
Each channel can be forced into the Full or Half Duplex
modes by either setting the duplex bit in the MI serial port
Control register or asserting the DPLX pin for the
corresponding channel with AutoNegotiation disabled.
2.18 RESET
The 84221 is reset when either:
(1) VDD is applied to the device,
The device can automatically configure itself for Full or
Half Duplex modes by using the AutoNegotiation algorithm
to advertise and detect Full and Half Duplex capabilities to
and from a remote terminal. For detailed information, refer
to the LINK INTEGRITY & AUTONEGOTIATION Section.
(2) the reset bit is set in the MI serial port Control
register, or
(3) the RESET pin is asserted active low.
When the reset is initiated by either (1) or (2), an internal
power-on reset pulse is generated which resets all internal
circuits, forces the MI serial port bits to their default
values, and latches in new values for the MI address. After
the power-on reset pulse has finished, the reset bit in the
MI serial port Control register is cleared and the device is
ready for normal operation. The device is guaranteed to
be ready for normal operation 50 mS after the reset was
initiated.
2.16.2 10 Mbps
Full Duplex in 10 Mbps mode is identical to the 100 Mbps
mode.
2.16.3 Full Duplex Indication
Full Duplex detect activity can be monitored through the
duplex detect bit in the MI serial port Channel Status
Output register.
When the reset is initiated by (3), the identical procedure
takes place as in (1) and (2), except the device stays in
reset until the RESET pin is deasserted high.
Full Duplex detect activity also appears on the LED1 pin
by default.
23
MD400184/A
84221
2.19 POWERDOWN
Table 4. LED Event Definition
The 84221 can be powered down by setting the
powerdown bit in the MI serial port Control register. In
powerdown mode, the TP outputs are in high impedance
state, all functions are disabled except the MI serial port,
and the power consumption is reduced to a minimum. The
device will be ready for normal operation 50 mS after
powerdown is deasserted.
Symbol
ACT
COL
LINK100
LINK10
LINK
LINK+ACT
2.20 CLOCK
The 84221 requires a 50 MHz reference frequency for
internal signal generation in RMII mode. This reference
frequency must be applied to the CLKIN pin.
FDX
2.21 LED DRIVERS
10/100
Definition
Activity Occurred, Stretch Pulse to
100 mS
Collision Occurred, Stretch Pulse to
100 mS
100 Mb Link Detected
10 Mb Link Detected
100 Mb or 10 Mb Link Detected
LED on if Link Detected (10 or 100).
LED Blinks if Activity Determined,
Stretch Pulse to 100 mS
Full Duplex Mode Detected with Link
Pass
10 Mb Mode Enabled (High) or
100 Mb Mode Enabled (Low) with
Link Pass
The LED[3:0] outputs can drive LED’s tied to either VDD
or GND.
The LED definitions assume that the LED
outputs are active low. If the LED Anodes are tied to the
positive power supply (through limiting resistors), the LED
will indicate the event as shown in Table 3. If the LED
Cathodes are tied to ground and the Anodes to the 84221
Driver output, they will indicate the respective
complementary events.
The 84221 has one predefined repeater mode which can
be enabled by asserting the REPEATER pin.
The LED[3:0] outputs can also drive other digital inputs.
2.23 MI SERIAL PORT
2.22 REPEATER MODE
2.23.1 Signal Description
The MI serial port has five pins, MDC, MDIO, and
PHYAD[4:2]. MDC is the serial shift clock input. MDIO is a
bidirectional data I/O pin. PHYAD[4:2] are physical
address pins.
Table 3. LED Function Definition
LEDDEF
LED3
LED2
LED1
LED0
1
LINK +
ACT
COL
FDX
10/100
0
LINK
100
ACT
FDX
LINK10
Notes:
Default = 000 When pin LEDDEF = 0
24
MD400184/A
84221
0 BT
4.5 BT
3.1 V
0.5 V/ns
0.25 BT
2.25 BT
585 mV
6.0 BT
+50 mV
–50 mV
45.0 BT
585 mV sin(2 * * (t/1BT))
0 t 0.25 BT and
2.25 t 2.5 BT
–3.1 V
2.5 BT
4.5 BT
Figure 6. SOI Output Voltage Template - 10 Mbps
1.3 BT
0 BT
3.1 V
0.5 V/ns
585 mV
0.5 BT
0.6 BT
2.0 BT
300 mV
4.0 BT
+50 mV
+50 mV
–50 mV
200 mV
0.25 BT
–50 mV
4.0 BT
42.0 BT
–3.1 V
0.85 BT
2.0 BT
Figure 7. Link Pulse Output Voltage Template - 10 Mbps
25
MD400184/A
84221
Pins PHYAD[4:2] set the three most significant bits of the
PHY address. The two least significant bits of the PHY
address are set internally to match the channel number,
as shown in Table 5.
high impedance state. When the MI serial port is in the
idle state, a 01 pattern on the MDIO pin initiates a serial
shift cycle. Data on MDIO is then shifted in on the next 14
rising edges of MDC (MDIO is high impedance). If the
register access mode is not enabled, on the next 16 rising
edges of MDC, data is either shifted in or out on MDIO,
depending on whether a write or read cycle was selected
with the bits READ and WRITE. After the 32 MDC cycles
have been completed, one complete register has been
read/written, the serial shift process is halted, data is
latched into the device, and MDIO goes into high
impedance state. Another serial shift cycle cannot be
initiated until the idle condition (at least 32 continuous 1's)
is detected.
Table 5. PHYAD[1:0] Settings
PHYAD1
PHYAD0
Channel 3
1
1
Channel 2
1
0
Channel 1
0
1
Channel 0
0
0
2.23.2 Timing
2.23.3 Bit Types
Figure 10 shows a timing diagram for a MI serial port
cycle.
Since the serial port is bidirectional, there are many types
of bits. The bit type definitions are summarized in Table 6.
The MI serial port is idle when at least 32 continuous 1's
are detected on MDIO and remains idle as long as
continuous 1's are detected. During idle, MDIO is in the
Write bits (W) are inputs during a write cycle and are high
impedance during a read cycle. Read bits (R) are outputs
during a read cycle and high impedance during a write
cycle. Read/Write bits (R/W) are actually write bits that
a.) Normal Link Pulse (NLP)
TX_DI±
b.) Fast Link Pulse (FLP)
TX_DI±
D0
Clock
D1
Clock
Data
D2
Clock
Data
Data
D3
Clock
D14
Clock
Data
D15
Clock
Data
Clock
Data
Figure 8. NLP vs. FLP Link Pulse
26
MD400184/A
84221
can be read out during a read cycle. R/WSC bits are R/W
bits that are self clearing after a set period of time or after
a specific event has completed. R/LL bits are read bits that
latch themselves when they go low, and they stay latched
low until read. After they are read, they are reset high. R/
LH bits are the same as R/LL bits, except that they latch
high. R/LT are read bits that latch themselves whenever
they make a transition or change value, and they stay
latched until they are read. After R/LT bits are read, they
are updated to their current value. The R/LT bits can also
be programmed to assert the interrupt function as
described in the Interrupt section.
2.23.4 Frame Structure
The structure of the serial port frame is shown in Table 7
and a timing diagram is shown in Figure 10. Each serial
port access cycle consists of 32 bits (or 720 bits if multiple
register access is enabled and REGAD[4:0]=11111),
exclusive of idle. The first 16 bits of the serial port cycle
are always write bits and are used for addressing. The last
16/704 bits are from one/all of the 4 x 11 data registers.
The first 2 bits in Table 7 and Figure 10 are start bits and
need to be written as a 01 for the serial port cycle to
continue. The next 2 bits are read and write bits which
determine whether the accessed data register bits will be
read or write. The next 5 bits are device addresses. The 3
most significant bits must match the values on pins
PHYAD[4:2] and the 2 least significant bits select one of
four channels for access. The next 5 bits are register
address select bits which select one of the eleven
registers for access. The next 2 bits are turnaround bits
which are not an actual register bits but extra time to
switch MDIO from write to read if necessary. The final 16
bits of the MI serial port cycle (or 704 bits if multiple
register access is enabled and REGAD[4:0]=11111) come
from the specific data register designated by the register
address bits REGAD[4:0].
Table 6. MI Register Bit Type Definition
Symbol
Name
W
Write
Input
No Operation,
Hi Z
R
Read
No Operation, Hi
Z
Output
R/W
Read/Write
Input
Output
R/
WSC
Read/
Write
Self Clearing
Input
Clears Itself
After Operation Completed
Output
R/LL
Read/
Latching
Low
No Operation,
Hi Z
Output
When Bit Goes
Low,
Bit Latched.
When Bit Is
Read, Bit
Updated.
R/LH
Read/
Latching
High
No Operation,
Hi Z
Output
When Bit Goes
High,
Bit Latched.
When Bit Is
Read, Bit
Updated.
R/LT Read/
Latching
on Transition
No Operation,
Hi Z
Output
When Bit
Transitions,
Bit Latched And
Interrupt Set
When Bit Is
Read, Interrupt
Cleared And Bit
Updated.
Definition
Write Cycle
Read Cycle
2.23.5 Register Structure
The 84221 has eleven 16 bit registers for each channel.
All eleven registers are available for setting configuration
inputs and reading status outputs. A map of the registers
is shown in Table 8. The eleven registers consist of six
registers that are defined by IEEE 802.3 specifications
(Registers 0-5) and five registers that are unique to the
84221 (Registers 16-20).
27
MD400184/A
84221
The structure and bit definition of the Control Register is
shown in Table 9. This register stores various configuration
inputs and its bit definition complies with the IEEE 802.3
specifications.
The structure and bit definition of the Channel
Configuration Register is shown in Table 16. This register
stores various configuration inputs unique to each
channel.
The structure and bit definition of the Status Register is
shown in Table 10. This register contains device
capabilities and status output information and its bit
definition complies with the IEEE 802.3 specifications.
The structure and bit definition of the Channel Status
Output Register is shown in Table 17. This register
contains output status information from each channel.
The structure and bit definition of the Global Interrupt
Mask Register is shown in Table 18. This register is
common for all four channels. Bit 7 is the interrupt
indication. The 7 least significant bits are the Mask bits for
the R/LT status bits in the Channel Status Output
Register.
The structure and bit definition of the PHY ID Register 1
and PHY ID Register 2 is shown in Table 11 and Table 12,
respectively. These registers contain an identification code
unique to the 84221 and their bit definition complies with
the IEEE 802.3 specifications.
Register 20 in Table 19 is reserved for factory use only. All
bits must be set to the pre-set default states shown for
normal operation.
The structure and bit definition of the Auto Negotiation
Advertisement and Auto Negotiation Remote End
Capability registers is shown in Table 13 and Table 14,
respectively. These registers are used by the Auto
Negotiation algorithm and their bit definition complies with
the IEEE 802.3 specifications.
2.23.6 Invalid Registers
The registers in locations 6-15 and 21-31 are not
implemented on the device, hence unused. When an
unused register is read, the value returned can be
configured to be either all 0s or all 1s by appropriately
pinstrapping the REGDEF pin.
The structure and bit definition of the Global Configuration
Register is shown in Table 15. This register is common for
all four channels. It stores various configuration inputs.
28
MD400184/A
MD400184/A
0
0
0
0
READ CYCLE
MDIO
MDC
MDIO
MDC
WRITE CYCLE
ST
1
1
ST
1
1
29
OP
OP
1
3
1
3
P4
4
P4
4
P3
5
P3
5
PHYAD
P2
6
PHYAD
P2
6
P1
7
P1
7
P0
8
P0
8
R4
9
R4
9
REGAD
R2
11
R1
12
R0
13
1
14
TA
0
15
D15
16
R3
10
R1
12
REGAD
R2
11
R0
13
Z
14
0
TA
D15
15
D14
16
WRITE BITS
PHY CLOCKS IN DATA ON RISING EDGES OF MDC WITH t
R3
10
WRITE BITS
PHY CLOCKS IN DATA ON RISING EDGES OF MDC WITH t s = 10nS MIN t h = 10ns MIN
0
2
0
2
D13
18
D13
D12
18
= 10nS MIN t
17
s
D14
17
D11
19
D10
20
= 10ns MIN
D11
20
D9
21
D10
21
D8
22
D9
22
DATA
D7
D6
24
D7
24
DATA
23
D8
23
D5
25
D6
25
D4
26
D5
26
D3
27
D4
27
D2
28
D3
28
READ BITS
PHY CLOCKS OUT DATA ON RISING EDGES OF MDC WITH t d = 20nS MAX
h
D12
19
D1
29
D2
29
D0
30
D1
30
31
D0
31
84221
Figure 10. MI Serial Port Frame Timing Diagram
84221
INTERNAL
INTERRUPT
MDC
MDIO
MDIO HI-Z
PULLED HIGH EXTERNALLY
INTERRUPT
PULSE
MDIO HI-Z
PULLED HIGH EXTERNALLY
INTERRUPT
PULSE
MDIO HI-Z
PULLED HIGH EXTERNALLY
INTERNAL
INTERRUPT
MDC
MDIO
B1
B0
LAST TWO BITS
OF READ CYCLE
MDIO HI-Z
PULLED HIGH EXTERNALLY
Figure 11. MDIO Interrupt Pulse
30
MD400184/A
84221
3.0 REGISTER DESCRIPTION
Table 7. MI Serial Port Structure
<Idle>
IDLE
<Start>
ST[1:0]
<Read>
READ
<Write>
WRITE
<PHY Addr.>
PHYAD[4:0]
<Reg. Addr.>
REGAD[4:0]
<Turnaround>
TA[1:0]
<Data>
D[15:0]
MI Registers - Address and Default Value
Default
(Hex Code)
REGAD
Name
00000
Control Register
3000
00001
Status Register
7809
00010
PHY ID 1 Register
0016
00011
PHY ID 2 Register
F840
00100
Auto Negotiation Advertisement Register
01E1
00101
Auto Negotiation Remote Capability Register
0000
10000
Reserved
0008
10001
Reserved
0002
10010
Channel Status Output Register
10011
Reserved
007F
10100
Reserved
0000
0340/0240/
0140/0040
Symbol
Name
Definition
R/W
IDLE
Idle Pattern
These bits are an idle pattern. Device will not initiate an MI cycle until it
detects at least 32 1's.
W
ST[1:0]
Start Bits
When ST[1:0]=01, a MI serial port access cycle starts.
W
READ
Read Select
1 = Read Cycle
W
WRITE
Write Select
1 = Write Cycle
W
PHYAD[4:0]
Physical Device
Address
When PHYAD[4:2] bits match the PHYAD[4:2] pins, the MI serial port is
selected for operation. PHYAD[1:0] is used for channel selection:
PHYAD [1:0]=11 For Channel 3
PHYAD [1:0]=10 For Channel 2
PHYAD [1:0]=01 For Channel 1
PHYAD [1:0]=00 For Channel 0
W
REGAD4[4:0]
Register Address
If REGAD[4:0]=00000-11100, these bits determine the specific register from
which D[15:0] is read/written. If multiple register access is enabled and
REGAD[4:0]=11111, all registers are read/written in a single cycle.
W
TA[1:0]
Turnaround Time
These bits provide some turnaround time for MDIO
When READ=1, TA[1:0]=Z0
When WRITE=1, TA[1:0]=ZZ
D[15:0]+
Data
These 16 bits contain data to/from one of the eleven registers selected by
register address bits REGAD[4:0].
31
MD400184/A
R/W
R or W
MD400184/A
Status
PHY ID #1
PHY ID #2
AutoNegot.
Advertisement
AutoNegot.
Remote
Capability
1
2
3
4
5
32
20 Reserved
19 Reserved
18 Channel Status
Output
17 Reserved
16 Reserved
Control
0
R
0
R/W
0
R/W
0
0
0
0
R/W
0
0
R/W
0
R/W
0
R
0
R
0
R/W
0
R/W
0
RPOL
0
0
0
0
R/W
0
0
R
0
R
0
R/W
ACK
NP
0
ACK
R
1
R
1
NP
OUI20
OUI19
R
0
R
1
R
0
OUI4
CAP_TXH
CAP_TXF
CAP_T4
R
0
R/W
1
R/W
0
OUI3
SPEED
LPBK
RST
R/WSC
0
R/W
0
0
0
R/W
0
R
0
0
R/W
0
0
R/W
0
0
R
0
RF
R/W
0
RF
R
1
OUI21
R
0
OUI5
R
1
x.13
x.14
x.15
R/W
0
0
0
R/W
0
R
0
0
R/W
0
0
R/W
0
0
R
0
0
R/W
0
0
R
1
OUI22
R
0
OUI6
R
1
CAP_TF
R/W
0
0
0
R/W
0
R
0
0
R/W
0
0
R/W
0
0
R
0
0
R/W
0
0
R
1
OUI23
R
0
OUI7
R
1
CAP_TH
R/W
0
PDN
ANEG_EN
R/W
1
x.11
x.12
0
R/W
0
0
0
R/W
0
R
11/10/01/00
—
R/W
0
0
R/W
0
0
R
0
T4
R/W
0
T4
R
0
PART5
R
0
OUI9
R/W
0
0
0
R/W
0
11/10/01/00
R
—
R/W
0
0
R/W
0
0
R
0
TX_FDX
R/W
1
TX_FDX
R
0
PART4
R
0
OUI10
R
0
0
R/W
0
DPLX
x.8
0
R/W
0
0
1
R/W
1
R/LT
1
LINK_FAIL
R/W
0
0
R/W
1
1
R
0
10_FDX
R/W
1
10_FDX
R
1
PART2
R
0
OUI12
R
0
CAP_SUPR
R/W
0
0
R/W
0
R
0
0
R/W
0
0
R/W
0
0
R
0
TX_HDX
R/W
1
TX_HDX
R
0
PART3
R
0
OUI11
R
0
0
0
R/WSC
0
R/W
0
0
1
R/W
1
R/LT
0
SPD_DET
R/W
0
0
R/W
0
0
R
0
0
R/W
1
0
R
0
PART1
R
0
OUI13
R
0
ANEG_ACK
0
R/W
0
0
0
1
R/W
1
1
R/W
1
R/LT
0
0
R/W
0
0
1
R/W
1
R
0
0
0
R/W
0
—
R
REV3
R
0
OUI15
R
1
CAP_ANEG
R
0
0
x.3
R/W
R/LT
0
DPLX_DET
R/W
0
0
0
R/W
0
R
0
0
0
R/W
0
R
0
PART0
R
1
OUI14
R/LH
0
REM_FLT
R
0
0
0
R
0
COLTST
R
0
x.4
x.5
x.6
x.7
Table 8. MI Serial Port Register Map
R/W
0
0
0
R/W
0
R
0
0
R/W
0
0
R/W
0
0
R
0
PAUSE
R/W
0
PAUSE
R
0
OUI24
R
0
OUI8
R
0
0
R
0
R/WSC
0
ANEG_RST
x.9
R/W
0
RMII_DIS
x.10
R/W
0
0
0
1
R/W
1
R/LT
0
0
R/W
1
1
R/W
0
0
R
0
0
R/W
0
0
R
—
REV1
R
1
OUI17
R/LH
0
JAB
R
0
0
x.1
R/W
0
1
R/W
1
R/LT
0
0
R/W
0
0
R/W
0
0
R
0
0
R/W
0
0
R
—
REV2
R
1
OUI16
R/LL
0
LINK
R
0
0
x.2
0
R/W
0
R/W
1
1
R/LT
0
0
R/W
0
0
R/W
0
0
R
0
CSMA
R/W
1
CSMA
—
R
REV0
R
0
OUI18
R
1
EXREG
R
0
0
x.0
84221
84221
Table 9. Register 0 – Control Register Definition
0.15
RST
0.14
LPBK
0.13
SPEED
0.12
ANEG_EN
0.11
PDN
0.10
RMII_DIS
0.9
ANEG_RST
0.8
DPLX
R/WSC
R/W
R/W
R/W
R/W
R/W
R/WSC
R/W
0.7
COLTST
0.6
0
0.5
0
0.4
0
0.3
0
0.2
0
0.1
0
0.0
0
R/WSC
R
R
R
R
R
R
R
Bit
Symbol
Name
Definition
R/W
Default
0.15
RST
Reset
1 = Reset, Bit Self Clearing After Reset
Completed
0 = Normal
R/WSC
0
0.14
LPBK
Loopback Enable
1 = Loopback Mode enabled
0 = Normal
R/W
0
0.13
SPEED
Speed Select
1 = 100 Mbps Selected (100BaseTX)
0 = 10 Mbps Selected (10BaseT)
Note: This Bit can be Overridden with SPEED
pin.
R/W
1
0.12
ANEG_EN
Auto Negotiation
Enable
1 = Auto Negotiation Enabled
0 = Auto Negotiation Disabled
Note: This Bit can be Overridden with ANEG pin.
R/W
1
0.11
PDN
Power Down
Enable
1 = Power Down
0 = Normal
R/W
0
0.10
RMII_DIS
RMII Interface
Disable
1 = RMII Interface Disable
0 = Normal
R/W
0
0.9
ANEG_RST
Auto Negotiation
Reset
1 = Restart Auto Negotiation Process,
Bit Self Clearing After Reset Completed
0 = Normal
R/WSC
0
0.8
DPLX
Duplex Mode
Select
1 = Full Duplex
0 = Half Duplex
Note: This Bit can be Overridden with DPLX pin.
R/W
0
0.7
COLTST
Collision Test
Enable
1 = Collision Test Enabled
0 = Normal
R/W
0
R
0
0.6
through
0.0
Reserved
Note: 0.15 Bit Is Shifted First
33
MD400184/A
84221
Table 10. Register 1- Status Register Definition
1.15
CAP_T4
1.14
CAP_TXF
1.13
CAP_TXH
1.12
CAP_TF
1.11
CAP_TH
1.10
0
1.9
0
1.8
0
R
R
R
R
R
R
R
R
1.7
0
1.6
CAP_SUPR
1.5
ANEG_ACK
1.2
LINK
1.1
JAB
1.0
EXREG
R
R
R
R/LL
R/LH
R
1.4
REM_FLT
1.3
CAP_ANEG
R/LH
Bit
Symbol
Name
Definition
1.15
CAP_T4
100BaseT4
Capable
1.14
CAP_TXF
1.13
R
R/W
Default
0 = Not Capable of 100BaseT4 Operation
R
0
100BaseTX Full
Duplex Capable
1 = Capable of 100BaseTX Full Duplex
R
1
CAP_TXH
100BaseTX Half
Duplex Capable
1 = Capable of 100BaseTX Half Duplex
R
1
1.12
CAP_TF
10BaseT Full
Duplex Capable
1 = Capable of 10BaseT Full Duplex
R
1
1.11
CAP_TH
10BaseT Half
Duplex Capable
1 = Capable of 10BaseT Half Duplex
R
1
Reserved
R
0
1.10
through
1.7
1.6
CAP_SUPR
MI Preamble
Suppression
Capable
0 = Not Capable of Accepting MI Frames with
Preamble Suppression
R
0
1.5
ANEG_ACK
Auto Negotiation
Acknowledgment
1 = Auto Negotiation Acknowledgment Process
Complete
0 = Auto Negotiation Not Complete
R
0
1.4
REM_FLT
Remote Fault
Detect
1 = Remote Fault detect. This bit is set when
Remote Fault Bit 5.13 is set.
0 = No Remote Fault
R/LH
0
1.3
CAP_ANEG
Auto Negotiation
Capable
1 = Capable of Auto Negotiation
R
1
1.2
LINK
Link Status
1 = Link Detect (Same As Bit 18.6 Inverted)
0 = Link Not Detect
R/LL
0
1.1
JAB
Jabber Detect
1 = Jabber Detect
0 = Normal
R/LH
0
1.0
EXREG
Extended Register 1 = Extended Registers Exist
Capable
R
1
Note: 1.15 Bit Is Shifted First
34
MD400184/A
84221
Tabel 11. Register 2 - PHY ID Register 1 Definition
2.15
OUI3
2.14
OUI4
2.13
OUI5
2.12
OUI6
2.11
OUI7
2.10
OUI8
2.9
OUI9
2.8
OUI10
R
R
R
R
R
R
R
R
2.7
OUI11
2.6
OUI12
2.5
OUI13
2.4
OUI14
2.3
OUI15
2.2
OUI16
2.1
OUI17
2.0
OUI18
R
R
R
R
R
R
R
R
Bit
Symbol
Name
Definition
R/W
2.15
2.14
2.13
2.12
2.11
2.10
2.9
2.8
2.7
2.6
2.5
2.4
2.3
2.2
2.1
2.0
OUI3
OUI4
OUI5
OUI6
OUI7
OUI8
OUI9
OUI10
OUI11
OUI12
OUI13
OUI14
OUI15
OUI16
OUI17
OUI18
Company ID,
Bits 3-18
SEEQ OUI = 00-A0-7D
Note: 2.15 Bit Is Shifted First
35
MD400184/A
R
Default
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
0
84221
Table 12. Register 3 - PHY ID Register 2 Definition
3.15
OUI19
3.14
OUI20
3.13
OUI21
3.12
OUI22
3.11
OUI23
3.10
OUI24
3.9
PART5
3.8
PART4
R
R
R
R
R
R
R
R
3.7
PART3
3.6
PART2
3.5
PART1
3.4
PART0
3.3
REV3
3.2
REV2
3.1
REV1
3.0
REV0
R
R
R
R
R
R
R
R
Bit
Symbol
Name
Definition
R/W
Default
3.15
3.14
3.13
3.12
3.11
3.10
OUI19
OUI20
OUI21
OUI22
OUI23
OUI24
Company ID,
Bits 3-18
SEEQ OUI = 00-A0-7D
R
1
1
1
1
1
0
3.9
3.8
3.7
3.6
3.5
3.4
PART5
PART4
PART3
PART2
PART1
PART0
Manufacturer’s
Part Number
04
R
0
0
0
1
0
0
3.3
3.2
3.1
3.0
REV3
REV2
REV1
REV0
Manufacturer’s
Revision Number
R
–
–
–
–
Note: 3.15 Bit Is Shifted First
36
MD400184/A
84221
Table 13. Register 4 - AutoNegotiation Advertisement Register Definition
4.15
NP
4.14
ACK
4.13
RF
4.12
0
4.11
0
4.10
PAUSE
4.9
T4
4.8
TX_FDX
R/W
R
R/W
R/W
R/W
R/W
R/W
R/W
4.7
TX_HDX
4.6
10_FDX
4.5
0
4.4
0
4.3
0
4.2
0
4.1
0
4.0
CSMA
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit
Symbol
Name
Definition
R/W
4.15
NP
Next Page Enable
0 = No Next Page
R/W
0
4.14
ACK
Acknowledge
1 = Auto Negotiation Word Recognized
0 = Not Recognized
R
0
4.13
RF
Remote Fault
1 = Auto Negotiation Remote Fault Detect
0 = No Remote Fault
R/W
0
Reserved
R/W
0
4.12
through
4.11
Default
4.10
PAUSE
PAUSE Frame
Capable
1 = Capable of Transmitting and Receiving
Pause Frames
0 = Not Capable
R/W
0
4.9
T4
100BaseT4
Capable
1 = Capable of 100BaseT4
0 = Not Capable
R/W
0
4.8
TX_FDX
100BaseTX Full
Duplex Capable
1 = Capable of 100BaseTX Full Duplex
0 = Not Capable
R/W
1
4.7
TX_HDX
100BaseTX Half
Duplex Capable
1 = Capable of 100BaseTX Half Duplex
0 = Not Capable
R/W
1
4.6
10_FDX
10BaseTX Full
Duplex Capable
1 = Capable of 10BaseTX Full Duplex
0 = Not Capable
R/W
1
4.5
10_HDX
10BaseTX Half
Duplex Capable
1 = Capable of 10BaseTX Half Duplex
0 = Not Capable
R/W
1
Reserved
R/W
0
1 = Capable of 802.3 CSMA Operation
0 = Not Capable
R/W
1
4.4
through
4.1
4.0
CSMA
CSMA 802.3
Capable
Note: 4.15 Bit Is Shifted First
37
MD400184/A
84221
Table 14. Register 5 - AutoNegotiation Remote Capability Definition
5.15
NP
5.14
ACK
5.13
RF
5.12
0
5.11
0
5.10
PAUSE
5.9
T4
5.8
TX_FDX
R
R
R
R
R
R
R
R
5.7
5.6
5.5
5.4
5.3
5.2
5.1
5.0
TX_HDX
10_FDX
0
0
0
0
0
CSMA
R
R
R
R
R
R
R
R
Bit
Symbol
Name
Definition
R/W
Default
5.15
NP
Next Page Enable
1 = Next Page Exists
0 = No Next Page
R
0
5.14
ACK
Acknowledge
1 = Received Auto Negotiation Word
Recognized
0 = Not Recognized
R
0
5.13
RF
Remote Fault
1 = Auto Negotiation Remote Fault Detect
0 = No Remote Fault
R
0
Reserved
R
0
5.12
through
5.11
5.10
PAUSE
PAUSE Frame
Capable
1 = Capable of Transmitting and Receiving
Pause Frames
0 = Not Capable
R
0
5.9
T4
100BaseT4
Capable
1 = Capable of 100BaseT4
0 = Not Capable
R
0
5.8
TX_FDX
100BaseTX Full
Duplex Capable
1 = Capable of 100BaseTx Full Duplex
0 = Not Capable
R
0
5.7
TX_HDX
100BaseTX Half
Duplex Capable
1 = Capable of 100BaseTx Half Duplex
0 = Not Capable
R
0
5.6
10_FDX
10BaseTX Full
Duplex Capable
1 = Capable of 10BaseTx Full Duplex
0 = Not Capable
R
0
5.5
10_HDX
10BaseTX Half
Duplex Capable
1 = Capable of 10BaseTx Half Duplex
0 = Not Capable
R
0
Reserved
R
0
1 = Capable of 802.3 CSMA Operation
0 = Not Capable
R
0
5.4
through
5.1
5.0
CSMA
CSMA 802.3
Capable
Note: 5.15 Bit Is Shifted First
38
MD400184/A
84221
Table 15. Register 16 - Reserved
16.15
0
16.14
0
16.13
0
16.12
0
16.11
0
16.10
0
16.9
0
16.8
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
16.7
0
16.6
1
16.5
0
16.4
0
16.3
1
16.2
0
16.1
0
16.0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit
16.15
thru
16.0
Symbol
Name
—
—
Definition
Reserved for factory use. Must be written with
default values specified above for normal operation
39
MD400184/A
R/W
Default
—
—
84221
Table 16. Register 17 - Reserved
17.15
0
17.14
0
17.13
0
17.12
0
17.11
0
17.10
0
17.9
0
17.8
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
17.7
0
17.6
0
17.5
0
17.4
0
17.3
0
17.2
0
17.1
1
17.0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit
Symbol
17.15
thru
17.0
—
Name
—
Definition
Reserved for factory use. Must be written with
default values specified above for normal
operation
40
MD400184/A
R/W
Default
—
—
84221
Table 17. Register 18 - Channel Status Output Register Definition
18.15
RPOL
18.14
18.13
18.12
18.11
18.10
18.9
18.8
R
R
R
R
R
R
R
R
18.7
18.6
LINK_FAIL
18.5
SPD_DET
18.4
DPLX_DET
18.3
18.2
18.1
18.0
JAB
R
R/LT
R/LT
R/LT
R/LT
R/LT
R/LT
R/LT
Bit
Symbol
Name
Definition
18.15
RPOL
Reversed Polarity
Detect
1 = Reversed Polarity Detect
0 = Normal
R/W
Default
R
0
18.14
Reserved
0
18.13
through
18.10
Reserved
R
0
0
0
0
18.9
18.8
Reserved
R
11/10/
01/00
18.7
Reserved
R
0
18.6
LINK_FAIL
Link Fail Detect
1 = Link Not Detected
0 = Normal
R/LT
1
18.5
SPD_DET
100/10 Speed
Detect
1 = Device in 100BaseTx Mode
0 = Device in 10 BaseT Mode
R/LT
0
18.4
DPLX_DET
Duplex Detect
1 = Device in Full Duplex Mode
0 = Device in Half Duplex Mode
R/LT
0
18.3
Reserved
R/LT
0
18.2
Reserved
R/LT
0
18.1
Reserved
R/LT
0
1 = Jabber Detected
0 = Normal
R/LT
0
18.0
JAB
Jabber Detect
Note: 18.15 Bit Is Shifted First
41
MD400184/A
84221
Table 18. Register 19 - Reserved
19.15
0
19.14
0
19.13
0
19.12
0
19.11
0
19.10
0
19.9
0
19.8
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
19.7
0
19.6
1
19.5
1
19.4
1
19.3
1
19.2
1
19.1
1
19.0
1
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit
Symbol
Name
—
—
19.15
thru
19.
Definition
Reserved for factory use. Must be written with
default values specified above for normal
operation
42
MD400184/A
R/W
Default
—
—
84221
Table 19. Register 20 - Reserved Register
20.15
0
20.14
0
20.13
0
20.12
0
20.11
0
20.10
0
20.9
0
20.8
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
20.7
0
20.6
0
20.5
0
20.4
0
20.3
0
20.2
0
20.1
0
20.0
0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Bit
Symbol
20.15
Thru
20.0
Name
Definition
R/W
Default
Reserved for Factory Use. Must be written to 0
for Normal Operation
R/W
0
Note: 20.15 Bit Is Shifted First
43
MD400184/A
84221
4.0 APPLICATION INFORMATION
For best cross-talk and system performance, it is
suggested that stacked connector and transformer
configurations be used.
Alternately, non-stacked
connectors may be used with stacked transformer types
and the crossover wiring can be put on the pc board using
a ground plane to reduce impedance mismatch.
4.1 EXAMPLE SCHEMATICS
A typical example of the 84221 used for a switching hub
application in twisted pair mode is shown in Figure 12; an
example of the 84221 used in fiber mode is shown in
Figure 13.
The transmit output needs to be terminated with two
external termination resistors in order to meet the output
impedance and return loss requirements of IEEE 802.3.
It is recommended that these two external resistors be
connected from VDD to each of the TPOP/N outputs, and
their value should be chosen to provide the correct
termination impedance when looking back through the
transformer from the twisted pair cable, as shown in
Figure 12. The value of these two external termination
resistors depends on the type of cable driven by the
device. Refer to the Cable Selection Section for more
details on choosing the value of these resistors.
4.2 TP INTERFACE
4.2.1 Transmit Interface
The interface between the TP outputs on TPOP/N and
the twisted pair cable is typically transformer coupled and
terminated with the two resistors as shown in Figure 12.
The transformer for the transmitter is recommended to
have a winding ratio of 1:1 with the center tap of the
primary winding tied to VDD, as shown in Figure 12. The
specifications for such a transformer are shown in Table
20. Sources for quad transformers compatible with the
84221 are listed in Table 21. Note that both “Stacked” and
“Non-Stacked pin out types are listed. The Stacked and
Non-Stacked designation refers to the type of RJ-45
connector used on the secondary side (line side) of the
transformer. The pinout of these types differ slightly so
that traces to the magnetics may be kept as short and
direct as possible.
To minimize common mode output noise and to aid in
meeting radiated emissions requirements, it may be
necessary to add a common mode choke on the transmit
outputs as well as add common mode bundle termination.
The transformers listed in Table 21 all contain common
mode chokes on both the transmit and receive sides, as
shown in Figure 12. Common mode bundle termination is
achieved by tying the unused pairs in the RJ45 to chassis
ground through 75 ohm resistors and a 0.01 uF capacitor,
as shown in Figure 12.
The "non-stacked" RJ-45 (also referred to as harmonica)
is a traditional horizontally oriented connector consisting
of four RJ-45 jacks in a single in-line assembly.
To minimize noise pickup into the transmit path in a
system or on a PCB, the loading on TPOP/N should be
minimized and both outputs should always be loaded
equally.
The newer "stacked" connector consists of a two-over-two
configuration so that four RJ-45 jacks are located in the
footprint area of two side by side connectors. This is a
significant improvement for higher density multi-port
applications and smaller system form-factors.
Table 20. TP Transformer Specification
Specification
Parameter
The SEEQ 84221 pin-out has been optimized for
connection to transformers and connectors designed for
the higher density stacked configuration. The SEEQ
Quad Transceiver will also operate with non-stacked
connectors using either transformers that map the pin-out
to the non-stacked configuration, or by using trace
"crossovers" on the pc board layout.
Receive
1:1 CT
1:1
Inductance, (uH Min)
350
350
Leakage Inductance, (uH)
0.2
0.2
Turns Ratio
The transformers listed with "non-stacked" pin-outs use
crossover connections inside the part to map the stacked
pin-out of the 84221 to non-stacked RJ-45 connectors.
Crossovers internal to the transformer are not made in a
controlled impedance environment, so this can impact,
somewhat, the cross-talk performance of the system.
44
MD400184/A
Transmit
Capacitance (pF Max)
15
15
DC Resistance (Ohms Max)
0.4
0.4
84221
VDD
50
1%
TXD1_3
TXD0_3
TXEN_3
RXD1_3
RXD0_3
CRS_DV_3
RXER_3
7
50
1%
TPOP_3
1:1
1
2
4
5 RJ45
7
8
3
TPON_3
25
TPIP_3
TPIN_3
25
25
1:1
0.01
6
25
75
84221
75
0.01
50
1%
SWITCH
FABRIC
OR
REPEATER
QUAD
100/10 MB
ETHERNET
CONTROLLER
TPOP_2
TXD1_2
TXD0_2
TPON_2
TXEN_2
RXD1_2
RXD0_2
25
CRS_DV_2
TPIP_2
25
RXER_2
25
TPIN_2
25
7
50
1%
1:1
1
2
4
5 RJ45
7
8
3
1:1
0.01
6
75
75
0.01
TXD1_1
TXD0_1
TXEN_1
RXD1_1
RXD0_1
CRS_DV_1
RXER_1
7
50
1%
50
1%
TPOP_1
1:1
1
2
4
5 RJ45
7
8
3
TPON_1
25
1:1
TPIP_1
25
25
TPIN_1
25
0.01
6
75
75
0.01
TXD1_0
TXD0_0
TXEN_0
RXD1_0
RXD0_0
CRS_DV_0
RXER_0
7
50
1%
TPOP_0
MDIO, MDC
1:1
1
2
4
5 RJ45
7
8
3
TPON_0
TPIP_0
2
50
1%
TPIN_0
25
25
25
1:1
0.01
6
25
16
LED[3:0]_[3:0]
PINSTRAP
REGDEF
DPLX[3:0]
SPEED[3:0]
ANEG
REPEATER
REXT
75
75
10K
0.01
AD_REV
LEDDEF
SYSTEM RESET
RESET
SYSTEM CLOCK
CLKIN
GND
Figure
Figure12.
1. Typical Switching Hub Port Schematic Using the 84221 in Twisted Pair Mode
45
MD400184/A
84221
Table 21. TP Transformer Sources
Part Number
Pin Out Type
Pulse
Vendor
H1062
Stacked
bel
S5585999B47
Stacked
nano pulse
6931-30
Stacked
Valor
ST6179
Stacked
Halo
TG110S453NX
Stacked
Pulse
H1053
NonStacked
bel
S558-5999J5
NonStacked
nano pulse
6949-30
Non Stacked
Valor
ST6403P
NonStacked
Halo
TG110S456NX
NonStacked
To minimize common mode input noise and to aid in
meeting susceptibility requirements, it may be necessary
to add a common mode choke on the receive input as
well as add common mode bundle termination. The
transformers listed in Table 24 contain common mode
chokes on both the transmit and receive sides, as shown
in Figure 12. Common mode bundle termination is
achieved by tying the receive secondary center tap and
the unused pairs in the RJ45 to chassis ground through
75 ohm resistors and a 0.01 uF capacitor, as shown in
Figure 12.
In order to minimize noise pickup into the receive path in
a system or on a PCB, the loading on TPIP/N should be
minimized and both inputs should be loaded equally.
4.3 TP TRANSMIT OUTPUT CURRENT SET
The TPOP/N output current level is set by an external
resistor tied between REXT and GND. This output
current is determined by the following equation where R is
the value of REXT:
Iout = (10K/R) * Iref
Where Iref = 40 mA (100 Mbps, UTP)
= 32.6 mA (100 Mbps, STP)
4.2.2 Receive Interface
Receive data is typically transformer coupled into the
receive inputs on TPIP/N and terminated with
an
external resistor as shown in Figure 12.
The transformer for the receiver is recommended to have
a winding ration of 1:1, as shown in Figure 12. The
specifications for such a transformer are shown in Table
20. Sources for the transformer are listed in Table 21.
The receive input needs to be terminated with the correct
termination impedance to meet the input impedance and
return loss requirements of IEEE 802.3. In addition, the
receive TP inputs need to be attenuated.
It is
recommended that both the termination and attenuation
be accomplished by placing four external resistors in
series across the TPIP/N inputs as shown in Figure 12.
The resistors should be 25%/25%/25%/25% of the total
series resistance, and the total series resistance should
be equal to the characteristic impedance of the cable
(100 ohms for UTP, 150 Ohms for STP). For 100 Ohm
twisted pair the resistor string values should be 25 Ohms
each (1%). It is also recommended that a 0.1uF capacitor
be placed between the center of the series resistor string
and Vcc in order to provide an AC ground for attenuating
common mode signal at the input. This capacitor is also
shown in Figure 12.
= 100 mA (10 Mbps, UTP)
= 81.6 mA (10 Mbps, STP)
For 100 Ohm UTP, REXT should be typically set to 10K
ohms and REXT should be a 1% resistor in order to meet
IEEE 802.3 specified levels. Once REXT is set for the
100 Mbps and UTP modes as shown by the equation
above, Iref is then automatically changed inside the
device when the 10 Mbps mode or UTP120/STP150
modes are selected as described in the Twisted Pair
Characteristics Transmit Section.
Keep resistor REXT as close to pins REXT and GND as
possible in order to reduce noise pickup into the
transmitter.
Since the TP output is a current source, capacitive and
inductive loading can reduce the output voltage level from
the ideal. Thus, in an actual application, it might be
necessary to adjust the value of the output current to
compensate for external loading. The TP output level can
be adjusted by changing the value of the external resistor
tied to RXT.
46
MD400184/A
84221
4.4 TRANSMITTER DROOP
4.6 REPEATER APPLICATIONS
The IEEE 802.3 specification has a transmitter output
droop requirement for 100BaseTX. Since the 84221 TP
output is a current source, it has no perceptible droop by
itself. However, the open circuit inductance of the
transformer added to the device transmitter output as
shown in Figure 12 will cause droop to appear at the
transmit interface to the TP wire. If the transformer
connected to the 84221 outputs meets the requirements
in Table 20, the transmit interface to the TP cable will meet
the IEEE 802.3 droop requirements.
4.6.1 RMII Based Repeaters
The 84221’s RMII interface can interface to any repeater
controller with a RMII interface. The 84221 does add
additional latency since the internal data path goes
through the unexposed MII interface. So any repeater
implementation has to carefully consider the repeater bit
budget impact due to the additional latency.
4.7 SERIAL PORT
4.5 CONTROLLER INTERFACE
4.7.1 General
4.5.1 General
The 84221 has a MI serial port to set all of the devices's
configuration inputs and read out the status outputs. Any
external device that has an IEEE 802.3 compliant MI
interface can connect directly to the 84221 without any
glue logic, as shown in Figure 12 and Figure 13.
The 84221 interfaces seemlessly to any standard
controller that has a RMII interface that fully complies with
the specification created by the RMII consortium.
4.5.2 Clocks
As described earlier, the MI serial port consists of five
lines: MDC, MDIO, and PHYAD[4:2]. However, only 2
lines, MDC and MDIO, are needed to shift data in and out.
The 84221 requires a 50 MHz clock that must be applied
to the CLKIN input. Data is clocked in on the rising edge
of CLKIN.
PHYAD[4:2] define the three most significant bits of the
PHY address, as described in the Section 4.7.2, Serial
Port Addressing Section.
4.5.3 RMII Disable
The RMII outputs can be placed in the high impedance
state and inputs disabled by setting the RMII disable bit in
the MI serial port Control register. When this bit is set to
the disable state, the TP outputs are both disabled and
transmission is inhibited. The default value of this bit
when the device powers up or is reset is dependent on
the device address. If the device address latched into
PHYAD[4:0] at reset is 11111, it is assumed that the
device is being used in applications where there maybe
more than one device sharing the RMII bus, like external
PHY’s or adapter cards, so the device powers up with the
RMII interface disabled. If the device address latched into
PHYAD[4:0] at reset is not 11111, it is assumed that the
device is being used in an application where it is the only
device on the RMII bus, like hubs, so the device powers
up with the RMII interface enabled.
4.7.2 Serial Port Addressing
The device address for the MI serial port is selected by
connecting the PHYAD[4:2] pins to the desired value. The
PHYAD[1:0] addresses are internally hardwired for each
channel as shown in both Tables 5 and 7.
4.8 UNMANAGED PORT CONFIGURATION
The 84221 has configuration inputs which can “over-ride”
the default configuration state obtained on POWER-UP or
RESET of the device. Use of these pins ANEG,
SPEED_[3:0], and DPLX_[3:0] allow selection of Global
Autonegotiation, Individual Port Speed (10/100), and
Individual Port Duplex (Full/Half), by properly strapping
47
MD400184/A
84221
these pins to VDD or VSS as shown in Table 23. Note
that these pins SHOULD NOT FLOAT, but must be
connected either High or Low for proper operation.
In order to obtain the “Default Mode of Operation”, ie:
Auto-negotiation enabled, 100MBs, and Half Duplex; the
ANEG, SPEED_[3:0], and DPLX_[3:0] pins should be set
to 1,1,0 respectively.
Table 23. Hardware Configuration
Configuration
State
AutoNegotiate
Normal
(POC/RESET)
Enabled
Config Pins
Speed
Duplex
Advertise
10/100
Advertise
Full/Half
ANEG=1
SPEED_
[3:0]=1
DPLX_
[3:0]=0
Complement
State
Disabled
10MBs
Full
Config Pins
ANEG=0
SPEED_
[3:0]=0
DPLX_
[3:0]=1
4.9 LONG CABLE
IEEE 802.3 specifies that 10BaseT and 100BaseTX
operate over twisted pair cable lengths from 0 to 100
meters. The squelch levels can be reduced by 4.5 dB if
the receive level adjust bit is appropriately set in the MI
serial port Channel Configuration register, which will allow
the 84221 to operate with up to 150 meters of twisted pair
cable. The equalizer is already designed to accomodate
between 0 to 150 meters of cable.
4.10 CLOCK
The 84221 requires a 25 MHz reference frequency for
internal signal generation in MII mode, and 50 MHz in
RMII mode. The appropriate reference frequency must
be applied to the CLKIN pin.
The LEDDEF pin determines the default settings for
LED[3:0]. If LEDDEF = 0, the default functions for
LED[3:0] are Link 100, Activity, Full Duplex, and Link 10,
respectively. If LEDDEF = 1, the LED functions for
LED[3:0] are forced to LINK + ACTIVITY, Collision, Full
Duplex and 10/100 Mbps operation, respectively. Table 5
defines the LED functions. Table 4 defines the LED
events.
The LED[3:0] outputs can also drive other digital inputs.
Thus, LED[3:0] can also be used as digital outputs whose
function can be user defined and controlled through the
MI serial port.5V Compatible I/O Operation.
4.12 5V COMPATIBLE I/O OPERATION
The input and output pins of the 84221 are tolerant of
signal levels up to a maximum of 5.5V (including
overshoot etc.). This allows the transceiver to be
operated with 5V controllers that have TTL I/O
characteristics (0.8 to 2.0V Input levels) without the use of
levelshifters or other interfaces.
Controllers and other system components may be
operate with 5V supplies and all inter-chip signals may be
connected directly to the 84221. All required external
logic levels must retain TTL compatability since the 84221
outputs are not guaranteed to achieve higher than 2.3V
with a load of 10ma. However, the inputs of the 84221 will
tolearte TTL or CMOS logic levels being driven into the
device.
This should make replacement of the Physical Layer
transceivers in existing designs quite simple since any 5V
devices do not need to be changed.
4.13 POWER SUPPLY DECOUPLING
All VDD's on each individual side should be connected
together (grouped) and tied to a power plane, as close as
possible to the 84221 supply pins. If the VDD's vary in
potential by even a small amount, noise and latchup can
result. The 84221 VDD pins should be kept to within 50
mV of each other.
4.11 LED DRIVERS
The LED[3:0] outputs can all drive LED’s tied to VDD as
shown in Figure 12 and Figure 13. In addition, the
LED[3:0] outputs can drive LED’s tied to GND as well.
The LED definitions assume that the LED outputs are tied
to VDD, active low signals (otherwise the LED outputs will
indicate their respective opposite events.)
All GND's should be connected as close as possible to
the device with a large ground plane. If the GND's vary in
potential by even a small amount, noise and latchup can
result. The GND pins should be kept to within 50 mV of
each other.
48
MD400184/A
84221
A 0.01-0.1uF decoupling capacitor should be connected
between the VDD group and GND on each of the 4 sides
of the 84221 as close as possible to the device pins,
preferably within 0.5 in. The value should be chosen
depending on whether the noise from VDD-GND is high
or low frequency. A conservative approach would be to
use two decoupling capacitors on each side, one 0.1uf for
low frequencys, and one 0.001 uf for high frequency noise
on the power supply.
The PCB layout and power supply decoupling discussed
above should provide sufficient decoupling to achieve the
following when measured at the device:
The VDD connection to the transmit transformer center
tap shown in Figures 12 and 13 must be well decoupled in
order to minimize common mode noise injection from the
supply into the twisted pair cable. It is recommended that
a 0.01 uF decoupling capacitor be placed between the
transformer center tap VDD connection and the 84221
GND plane. This decoupling capacitor should be
physically placed as close as possible to the transformer
center tap, preferably within 0.5 in.
(3) all GND’s should be within 50 mVpp of each other.
(1) the resultant AC noise voltage measured across
each VDD/GND set should be less than 100 mVpp,
(2) all VDD’s should be within 50 mVpp of each other,
and
49
MD400184/A
84221
5.0 SPECIFICATIONS
5.1 ABSOLUTE MAXIMUM RATINGS
Absolute maximum ratings are limits beyond which may cause permanent damage to the device or affect device
reliability. All voltages are specified with respect to GND, unless otherwise specified.
VDD Supply Voltage ........................................................................................................................... -0.3V to +4.0V
All Inputs and Outputs .......................................................................................................................... -0.3V to 5.5V
Package Power Dissipation ........................................................................................................... 3.0 Watt @ 70 oC
Storage Temperature ......................................................................................................................... -65 to +150 oC
Temperature Under Bias...................................................................................................................... -10 to +80 oC
Lead Temperature (Soldering, 10 Sec).......................................................................................................... 260 oC
Body Temperature (Soldering, 30 Sec) ......................................................................................................... 220 oC
Note that all inputs and outputs are 5V tolerant
5.2 DC ELECTRICAL CHARACTERISTICS
Unless otherwise noted, all test conditions are as follows:
1. TA= 0 to +70 oC
2. VDD = 3.3V +5%
3. 25 MHz +0.01%
4. REXT = 10K +1%, no load
LIMIT
SYM
VIL
PARAMETER
MIN
TYP
Input Low Voltage
VIH
Input High Voltage
IIL
Input Low Current
MAX
UNIT
0.8
Volt
2
CONDITIONS
Volt
±1
uA
VIN = GND
All Except RESET
10
IIH
Input High Current
50
uA
±1
uA
VIN = GND RESET
VIN = 5V for all except REGDEF
VIN = VDD for REGDEF
VOL
VOH
0.4
Volt
IOL = -4 mA, Except LED[3:0]
1
Volt
IOL = -20 mA, LED[3:0]
VDD
-1.0
Volt
IOH = 4 mA
All Except LED[3:0]
VDD
Volt
IOH = 10 mA, LED[3:0]
Volt
IOH = 10 uA
Output Low Voltage
Output High Voltage
-1.0
2.4
CIN
Input Capacitance
IDD
VDD Supply Current
IGND
GND Supply Current
5
pF
450
mA
Transmitting 100%, 100 Mbps
450
mA
Transmitting 100%, 10 Mbps
700
mA
Transmitting 100%, 100 Mbps, Note 1
700
mA
Transmitting 100%, 10 Mbps, Note 1
200
uA
Powerdown
Note 1. IGND includes current flowing into GND from the external resistors and transformer on TPOP/
TPON as shown in Figure 12.
50
MD400184/A
84221
Twisted Pair Characteristics, Transmit
Unless otherwise noted, all test conditions are as follows:
1. TA= 0 to +70 oC
2. VDD = 3.3 V +5%
3. 25 MHz +0.01%
4. REXT=10K +1%, no load
5. TPOP/N Loading Shown in Figure 12 or Equivalent
LIMIT
SYM
TOV
PARAMETER
TP Differential Output Voltage
MIN
0.950
TYP
1.000
MAX
1.050
UNIT
V pk
CONDITIONS
100 Mbps, UTP Mode, 100 Ohm Load
1.165
1.225
2.2
2.5
1.285
V pk
100 Mbps, STP Mode, 150 Ohm Load
2.8
V pk
2.694
3.062
10 Mbps, UTP Mode, 100 Ohm Load
3.429
V pk
10 Mbps, STP Mode, 150 Ohm Load
TOVS
TP Differential Output Voltage
Symmetry
98
102
%
100 Mbps, Ratio of Positive And
Negative Amplitude Peaks on TPOP/N
TORF
TP Differential Output Rise
And Fall Time
3.0
5.0
nS
100 Mbps
TORFS
TP Differential Output Rise
And Fall Time Symmetry
+0.5
nS
100 Mbps, Difference Between Rise
And Fall Times on TPOP/N
TODC
TP Differential Output Duty
Cycle Distortion
+0.25
ns
100 Mbps, Output Data=0101...
NRZI Pattern Unscrambled,
Measure At 50% Points
TOJ
TP Differential Output Jitter
+0.7
ns
100 Mbps, Output Data=Scrambled /H/
TOO
TP Differential Output
Overshoot
5.0
%
100 Mbps
TOVT
TP Differential Output Voltage
Template
See Figure 4
10 Mbps
TSOI
TP Differential Output SOI
Voltage Template
See Figure 6
10 Mbps
TLPT
TP Differential Output Link
Pulse Voltage Template
See Figure 7
10 Mbps, NLP and FLP
TOIV
TP Differential Output Idle
Voltage
TOIA
TP Output Current
TOIR
TP Output Current Adjustment
Range
TORA
TP Output Current TLVL Step
Accuracy
TOR
TP Output Resistance
TOC
TP Output Capacitance
50
38
40
42
31.06
32.66
34.26
88
100
112
71.86
81.64
91.44
10 Mbps, Measured on Secondary Side
of Xfmr in Figure 12
mA pk 100 Mbps, UTP with TLVL[3:0]=1000
mA pk 100 Mbps, STP with TLVL[3:0]=1000
mA pk 10 Mbps, UTP with TLVL[3:0]=1000
mA pk 10 Mbps, STP with TLVL[3:0]=1000
0.80
1.2
Adjustable with REXT,
Relative to TOIA with REXT=10K
0.86
1.16
Adjustable with TLVL[3:0].
See Section 4.3.
Relative to Value at TLVL[3:0]=1000.
+50
%
10K
Ohm
15
pF
51
MD400184/A
mV
Relative to Ideal Values in
Table 2. Values Relative to Output with
TLVL[3:0]=1000.
84221
Twisted Pair Characteristics, Receive
Unless otherwise noted, all test conditions are as follows:
1. TA= 0 to +70°C
2. VDD = 3.3V +5%
3. 25 MHz +0.01%
4. REXT = 10K +1%, no load
5. 62.5/10 MHz Square Wave on TP inputs in 100/10 Mbps
SYM
RST
RUT
PARAMETER
TP Input Squelch
Threshold
TP Input Unsquelch
Threshold
ROCV
TP Input Open
Circuit Voltage
RCMR
TP Input Common
Mode Voltage Range
RDR
TP Input Differential
Voltage Range
RIR
TP Input Resistance
RIC
TP Input Capacitance
MIN
166
LIMIT
TYP
MAX
500
UNIT
mV pk
CONDITIONS
100 Mbps, RLVL=0
310
540
mV pk
10 Mbps, RLVL=0
60
200
mV pk
100 Mbps, RLVL=1
217
378
mV pk
10 Mbps, RLVL=1
100
300
mV pk
100 Mbps, RLVL=0
186
324
mV pk
10 Mbps, RLVL=0
60
180
mV pk
100 Mbps, RLVL=1
130
227
mV pk
10 Mbps, RLVL=1
VDD- 2.4
Voltage on Either TPIP or TPIN with
Respect to GND
Volt
Voltage on Either TPIP or TPIP with
Respect to GND
± 0.2
ROCV
±0.25
VDD
5K
Volt
ohm
10
pF
52
MD400184/A
Volt
84221
AC Test Timing Conditions
Unless otherwise noted, all test conditions are as follows:
1. TA= 0 to +70 oC
2. VDD = 3.3 V +5%
3. 25 MHz +0.01%
4. REXT = 10K +1%, no load
5. Input conditions:
All Inputs:
tr,tf <= 5 nS, 20-80%
6. Output Loading
TPOP/N:
Same as Figure 12 or Equivalent, 10 pF
REGDEF:
1K Pullup, 50 pF
All Other Digital Outputs:
25 pF
7. Measurement Points:
TPOP/N, TPIP/N:
0 V During Data, +0.3 V at start/end of packet
All other inputs and outputs:
1.4 V
5.7 Clock Timing Characteristics
Refer To Figure 16 For Timing Diagram
LIMIT
SYM
t1
PARAMETER
CLKIN Period
MIN
19.996
TYP
20
MAX
20.002
UNIT
nS
t2
CLKIN High Time
7
nS
t3
CLKIN Low Time
7
nS
t1
t2
CLKIN
Figure 16. Output Timing
53
MD400184/A
CONDITIONS
t3
84221
Transmit Timing Characteristics
Refer To Figures 17-18 For Timing Diagram
LIMIT
SYM
t15
TXEN Setup Time
PARAMETER
MIN
4
t16
TXEN Hold Time
2
t19
TYP
MAX
UNIT
nS
nS
TXD Setup Time
4
nS
t20
TXD Hold Time
2
nS
t23
Transmit Propagation
Delay
60
t24
Transmit Output Jitter
t25
Transmit SOI Pulse Width
To 0.3V
t26
Transmit SOI Pulse Width
to 40 mV
t27
LEDn Delay Time
t28
LEDn Pulse Width
140
nS
100 Mbps
600
nS
10 Mbps
+0.7
nS pk-pk
100 Mbps
+ 5.5
nS pk-pk
10 Mbps
nS
10 Mbps
4500
nS
10 Mbps
25
mS
Activity
105
mS
Activity
250
80
54
MD400184/A
CONDITIONS
84221
CLK_IN
t 15
t 16
TX_EN
t 19 t 20
TXD [1:0]
D0
D1
D2
D3
t 24
t 23
TPO±
IDLE
IDLE
/J/K/
DATA
t 27
t 28
LEDn
Figure 17. Transmit Timing - 100 Mbps
55
MD400184/A
/T/R/
IDLE
84221
CLK_IN
t
t16
15
TXEN
t
TXD [1:0]
D0
D1
t
19 20
D2
D3
[1]
t 26
t
TPO±
t 24
23
PREAMBLE
PREAMBLE
t
27
t 25
DATA
t
28
LEDn
[1] = Each Di-Bit is present on RXD[1:0] for ten consecutive CLKIN cycles.
Figure 18. Transmit Timing - 10 Mbps
56
MD400184/A
DATA
SOI
84221
Receive Timing Characteristics
Refer To Figures 19-22 For Timing Diagram
LIMIT
SYM
PARAMETER
t31
Start of Packet to
CRS_DV Assert Delay
t32
End of Packet to
CRS_DV Deassert
Delay
t33
t34
MIN
130
Start of Packet to
Valid RXD[1:0] Delay
End of Packet
to Valid RXD[1:0]
Delay
TYP
MAX
UNIT
200
nS
CONDITIONS
100 Mbps
400
nS
10 Mbps
240
nS
100 Mbps
1000
nS
10 Mbps
240
nS
100 Mbps
3600
ns
10 Mbps
360
ns
100 Mbps
2800
ns
10 Mbps
t37
CLKIN to RXD[1:0],
RXER Delay
2
6
nS
100 Mbps and 10 Mbps
t40
SOI Pulse Minimum
Width Required for
Idle Detection
125
200
nS
10 Mbps
t41
Measured TPIP/N from last zero cross
to 0.3 V point.
Receive Input Jitter
t43
LEDn Delay Time
t44
LEDn Pulse Width
t45
RXD, CRS_DV
RXER Output
Rise and Fall Times
60
±2.0
nS pk-pk
100 Mbps
±13.5
nS pk-pk
10 Mbps
25
mS
Activity
105
mS
Activity
5
nS
57
MD400184/A
84221
TPI±
IDLE
J
K
DATA DATA
DATA DATA
DATA
DATA
DATA DATA
DATA
DATA
DATA DATA
DATA
DATA
DATA DATA
DATA
DATA
t 41
t 31
CRS_DV
CLKIN
t 33
t 37
RXD [1:0]
PREAMBLE PREAMBLE PREAMBLE PREAMBLE PREAMBLE
t 37
t 37
RXER
Figure 19. Receive Timing, Start of Packet - 100 Mbps
58
MD400184/A
84221
TPI±
DATA
T
R
I
I
I
I
I
I
I
I
I
I
I
I
I
t 32
CRS_DV
t 34
t 37
t 37
CLKIN
t 37
RXD [1:0]
DATA
DATA
DATA
DATA
DATA
DATA
DATA
Figure 20. Receive Timing, End of Packet - 100 Mbps
59
MD400184/A
I
I
I
I
I
I
84221
t 41
TPI±
DATA
DATA
t 31
CRS_DV
CLKIN
t 37
t 33
RXD [1:0]
PREAMBLE
PREAMBLE
DATA
DATA
DATA
[1]
t 43
t 44
LEDn
[1]: Each Di-Bit is present on RXD[1:0] for 10 consecutive CLKIN cycles.
Figure 21. Receive Timing, Start of Packet - 10 Mbps
60
MD400184/A
DATA
84221
t 41
TPI±
DATA
DATA
DATA
DATA
DATA
SOI
t 40
t 32
CRS_DV
t 34
t 37
CLKIN
t 37
RXD [3:0]
DATA
DATA
DATA
DATA
DATA
DATA
DATA
[1]
[1]: Each Di-Bit is present on RXD[1:0] for 10 consecutive CLKIN cycles.
Figure 22. Receive Timing, End of Packet - 10 Mbps
61
MD400184/A
84221
Link Pulse Timing Characteristics
Refer To Figures 23-24 For Timing Diagram
LIMIT
SYM
PARAMETER
MIN
TYP
MAX
t61
NLP Transmit Link Pulse
Width
t62
NLP Transmit Link Pulse
Period
8
t63
NLP Receive Link Pulse
Width Required For
Detection
50
t64
NLP Receive Link Pulse
Minimum Period Required
For Detection
6
7
mS
link_test_min
t65
NLP Receive Link Pulse
Maximum Period
Required For Detection
50
150
mS
link_test_max
t66
NLP Receive Link Pulses
Required To Exit Link Fail
State
3
t67
FLP Transmit Link Pulse
Width
100
t68
FLP Transmit Clock Pulse
To Data Pulse Period
55.5
t69
FLP Transmit Clock Pulse
To Clock Pulse Period
111
t70
FLP Transmit Link Pulse
Burst Period
8
t71
FLP Receive Link Pulse
Width Required For
Detection
50
t72
FLP Receive Link Pulse
Minimum Period Required
For Clock Pulse Detection
5
25
uS
flp_test_min_timer
t73
FLP Receive Link Pulse
Maximum Period
Required For Clock Pulse
Detection
165
185
uS
flp_test_max_timer
t74
FLP Receive Link Pulse
Minimum Period Required
For Data Pulse Detection
15
47
uS
data_detect_min_timer
See Figure 8
CONDITION
ns
24
mS
nS
link_loss
3
3
Link
Pulses
150
nS
62.5
69.5
uS
125
139
uS
22
mS
lc_max
interval_timer
transmit_link_burst_timer
nS
62
MD400184/A
UNIT
84221
LINK PULSE TIMING CHARACTERISTICS (Continued)
Refer To Figures 23-24 For Timing Diagram
LIMIT
SYM
PARAMETER
MIN
t75
FLP Receive Link Pulse
Maximum Period
Required For Data Pulse
Detection
MAX
UNIT
78
100
uS
FLP Receive Link Pulses
Required To Detect Valid
FLP Burst
17
17
Link
Pulses
t77
FLP Receive Link Pulse
Burst Minimum Period
Required For Detection
5
7
mS
nlp_test_min_timer
t78
FLP Receive Link Pulse
Burst Maximum Period
Required For Detection
50
150
mS
nlp_test_max_timer
t79
FLP Receive Link Pulses
Bursts Required To Detect
AutoNegotiation Capability
3
3
Link Pulse
Bursts
t80
FLP Receive Acknowledge Fail Period
1200
1500
mS
t81
FLP Transmit Renegotiate Link Fail Period
1200
1500
mS
break_link_timer
t82
NLP Receive Link Pulse
Maximum Period
Required For Detection
After FLP Negotation Has
Completed
750
1000
mS
link_fail_inhibit_timer
t76
TYP
3
63
MD400184/A
CONDITION
data_detect_max_timer
84221
TPO±
t 61
t 62
a.) Transmit NLP
TPI±
t 63
t 64
t 66
t 65
PLEDn
b.) Receive NLP
Figure 23. NLP Link Pulse Timing
64
MD400184/A
84221
CLK
DATA
CLK
DATA
CLK
CLK
DATA
TPO±
t 67
t 68
t 69
t 70
a.) Transmit FLP and Transmit FLP Burst
CLK
DATA
CLK
DATA
TPI±
31.25
t 71
62.5
93.75
125
156.25
t 72
t 73
t 74
t 75
b.) Receive FLP
TPI±
t 79
t 77
t 78
LEDn
c.) Receive FLP Burst
Figure 24. FLP Link Pulse Timing
65
MD400184/A
84221
Jabber Timing Characteristics
Refer To Figure 25 For Timing Diagram
LIMIT
SYM
PARAMETER
MIN
MAX
UNIT
t91
Jabber Activation Delay
Time
50
TYP
100
mS
CONDITIONS
10 Mbps
t92
Jabber Deactivation
Delay Time
250
750
mS
10 Mbps
TXEN
t 91
t 92
TPO±
Figure 25 Jabber Timing
66
MD400184/A
84221
LED Driver Timing Characteristics
Refer To Figure 26 For Timing Diagram
LIMIT
SYM
PARAMETER
MIN
t96
LED[3:0] On Time
t97
LED[3:0] Off Time
TYP
MAX
UNIT
80
105
mS
LED[3:0] Programmed to Blink
80
105
mS
LED[3:0] Programmed to Blink
t 96
LED[5:0]
Figure 26. LED Driver Timing
67
MD400184/A
CONDITIONS
t 97
84221
MI Serial Port Timing Characteristics
Refer To Figure 27 For Timing Diagram
PARAMETER
LIMIT
TYP
SYM
t101
MDC High Time
MIN
20
MAX
UNIT
nS
CONDITIONS
t102
MDC Low Time
20
nS
t103
MDIO Setup Time
10
nS
Write Bits
t104
MDIO Hold Time
10
nS
Write Bits
t105
MDC To MDIO Delay
20
nS
Read Bits
t106
MDIO Hi-Z To Active
Delay
20
nS
Write-Read Bit Transition
t107
MDIO Active To HI-Z
Delay
20
nS
Read-Write Bit Transition
t108
Frame Delimiter (Idle)
t110
MDC To MDIO Interrupt
Pulse Assert Delay
100
nS
t111
MDC To MDIO Interrupt
Pulse Deassert Delay
100
nS
32
Clocks
t 101
MDC
0
t 103
1
13
14
t 104
# of Consecutive MDC Clocks With
MDIO = 1
t 102
15
16
t 106
17
30
31
t 105
t 107
MDIO
(READ)
ST1
REGAD0
TA0
D15
D14
D0
TA1
t 103
MDIO
(WRITE)
ST0
ST1
t 104
ST0
REGAD0
TA1
TA0
D15
D1
D0
t 109
MDINT
Figure 27. MI Serial Port Timing
68
MD400184/A
84221
Ordering Information
Q
Q
84221
/
Z
PACKAGE
TYPE
TEMPERATURE
RANGE
PART TYPE
PRODUCT REV
Q = Plastic Quad
Flat Pack
Q – 0°C to +70°C
Quad 100 Base-TX/FX10 Base-T
Physical Layer Device (PHY)
Revision History
99191
Document Revision changed to MD400184/A
Global: All references to FX or 100BaseFX have been deleted.
Page 16: Section 2.4.1 4B5B Decoder - 100 Mbps
- Paragraph #2 has had a copy change.
- Section 2.6.3 Scambler Bypass, has been deleted.
Page 17: Section 2.7.1 100 Mbps
- Paragraph #3 has been deleted.
- Section 2.7.3 has been deleted.
Page 19: Section 2.11.1 100 Mbps
- Paragraph #5 has had a copy change.
- Section 2.12.1 100 Mbps has had a copy change.
Page 32: Table 8. MI Serial Port Register Map
- 18 Channel Status Output , Bit X.14 is now 0, Bit X.9 and X.8 are now —, Bits X.3, X.2, X.1, X.0 are now 0.
Page 67: Figure 25. Jabber Timing
- Timing COL,and CRS have been deleted.
Page 41: Table 17. Register 18 - Channel Status Output Register Definition
- Bits 18.14, 18.9, 18.8, 18.3, 18.2, 18.1 are now blank.
- Definitions 18.14, 18.9, 18.8, 18.3, 18.2, 18.1, have been changed to Reserved.
- Symbol and name for 18.14, 18.9, 18.8, 18.3, 18.2, 18.1 are now blank.
This document is subject to changed without notice. SEEQ Technology Incorporated assumes no
responsibility for any errors contained herein.
Copyright by SEEQ Technology Incorporated, All Rights Reserved.
69
MD400184/A
84221
128 Pin PQFP
23.20 ±0.30
20.00 ±0.20
18.50 ref.
12.50 ref.
Technology Incorporated
14.00 ±0.20
17.20 ±0.30
0.18 ± 0.05
84221
0.10 MAX
#128
QQ84220
0.75 ref.
#1
0.50 BSC
0.75 ref.
0.13 min.
0.28 max.
3.40 MAX
0.25 min.
See Detail A.
0.13 R min.
Gauge Plane
0° - 7°
0.13 R min.
0.88 ± 0.15
Detail A.
1. All dimensions are in (millimeters).
70
MD400184/A