RELEASED
PM7383 FREEDM-32A256
DATASHEET
PMC-2010336
ISSUE 1
FRAME ENGINE AND DATA LINK MANAGER 32A256
common clock and frame pulse. For T1/J1/E1 channelised traffic or for
unchannelised traffic, each link is independent and has its own associated clock.
For each link, the RCAS256 performs a serial to parallel conversion to form data
bytes. The data bytes are multiplexed, in byte serial format, for delivery to the
Receive HDLC Processor / Partial Packet Buffer block (RHDL256) at SYSCLK
rate. In the event where multiple streams have accumulated a byte of data,
multiplexing is performed on a fixed priority basis with link #0 having the highest
priority and link #31 the lowest.
From the point of view of the RCAS256, links configured for H-MVIP traffic
behave identically to links configured for T1/J1/E1 channelised or unchannelised
traffic in the back end, only differing on the link side as described herein. First,
the number of time-slots in each frame is programmable to be 32 or 128 and has
an associated data clock frequency that is double the data rate. This provides
more bandwidth per link for applications requiring higher data densities on a
single link. Second, H-MVIP links reference the start of each frame with a frame
pulse, thereby avoiding having to gap the link clock during the framing bits/bytes
of each frame. The frame pulse is provided by an H-MVIP bus master and
ensures that all agents sharing the H-MVIP bus remain synchronized. When
configured for operation in 2.048 Mbps mode, the frame pulse is sampled using
the same clock which samples the data. When configured for operation in 8.192
Mbps H-MVIP mode, the frame pulse is sampled using a separate frame pulse
clock provided by an H-MVIP bus master. The frame pulse clock has a
synchronous timing relationship to the data clock. Third, not all links are
independent. When configured for operation in 2.048 Mbps H-MVIP mode, each
group of 8 links share a clock and a frame pulse. Links 0 through 7, 8 through
15, 16 through 23 and 24 through 31 each share a clock and a frame pulse. Not
all 8 links within each group need to be configured for operation in 2.048 Mbps
H-MVIP mode. However, any link within each logical group of 8 which is
configured for 2.048 Mbps H-MVIP operation will share the same clock and frame
pulse. When configured for operation in 8.192 Mbps H-MVIP mode, links 4m
(0?m?7) share a frame pulse, a data clock and a frame pulse clock. Again, not
all eight 4m (0?m?7) links need to be configured for operation in 8.192 Mbps
H-MVIP mode, however, any link which is configured for 8.192 Mbps H-MVIP
operation will share the same frame pulse, data clock and frame pulse clock. If
link 4m is configured for 8.192 Mbps H-MVIP operation, then data transferred on
that link is “spread” over links 4m, 4m+1, 4m+2 and 4m+3 from a channel
assigner point of view. Accordingly, when link 4m is configured for operation in
8.192 Mbps H-MVIP mode, links 4m+1, 4m+2 and 4m+3 must also be configured
for operation in 8.192 Mbps H-MVIP mode. In the back end, the RCAS256
extracts and processes the time-slots in the same way as channelised T1/J1/E1
traffic.
PROPRIETARY AND CONFIDENTIAL
46