Altera ALTDLL User Manual
Page 59

4–23
Chapter 4: Functional Description
DQS_CONFIG / IO_CONFIG Block
ALTDLL and ALTDQ_DQS Megafunctions User Guide
© February 2012
Altera Corporation
In all cases, the most significant bit (MSB) of the delay chain values is shifted in first
and the least significant bit (LSB) is shifted in last. For example, in the first four
configuration clock cycles, the first configuration clock cycle corresponds to the MSB
of the input delay chain value and the fourth configuration clock cycle corresponds to
the LSB.
The delay only takes effect when the
config_update
signal is asserted for one
configuration clock cycle, in which all the bits in the serial shift register feeds an
11–bit, parallel-loaded register. Right after the signal is deasserted, you can observe
the delay from
datain
(of the delay chain primitive) to
dataout
(of the delay chain
block).
For all delay chains, each delay setting increment adds approximately 50 ps of delay
(the actual value depends on the device speed grade); therefore, the total delay value
is equal to the number of stages in the delay chain ×50 ps. For example, if you set the
number of stages in the delay chain to five, then the total delay value is five times
50 ps, which is 250 ps.
through
are simulation examples that show the results of
varying the delay at the input delay chain (D1).
f
For more information about controlling these delay chains and how to vary the
output delay chains (D5 and D6), refer to the Design Example 1: Dynamically Changing
Delay Chains in Output Buffer of Stratix III section of the
.
Setting the Input Delay Chain (D1) to Zero Delay (default)
shows that there are no timing difference with the cursor at 70 ns when
D1 is set to zero delay. The cursor at 70 ns represents the path from the bidirectional
buffer (
bidir_dq_input_data_in
) through the input delay chain
(
bidir_dq_0_input_delay_chain_inst
). You can view the effects of the delay
chain by comparing the
datain
port and the
dataout
port of the
bidir_dq_0_input_delay_chain_inst
.