8.12 LVDS Timing at Different Sampling Frequencies — 1-Wire Interface, 14x-Serialization, Decimation by 2 Filter Enabled
See (3)(1)(2)
| ADC CLK Frequency (MSPS) |
Set-up Time (tsu), ns |
Hold Time (tH), ns |
tPROG = (6 / 7) × T + tdelay, ns(4)
|
| Fs = 1 / T |
Data Valid to Zero-Crossing of LCLKP
(both edges) |
Zero-Crossing of LCLKP to Data Becoming Invalid
(both edges) |
tPROG = delay from input clock zero-cross rising edge to frame clock zero cross (rising edge) |
| MIN |
TYP |
MAX |
MIN |
TYP |
MAX |
MIN |
TYP |
MAX |
| 80 |
0.43 |
|
|
0.54 |
|
|
7.5 |
9 |
10.5 |
| 60 |
0.54 |
|
|
0.9 |
|
|
7.5 |
9 |
10.5 |
| 40 |
1.1 |
|
|
1.45 |
|
|
7.5 |
9 |
10.5
|
(1) The LVDS timing depends on the state of the internal PLL. Use
Table 3 to configure the PLL when decimation by two is enabled..
(2) For any given ADC input clock frequency, TI recommends to use the highest PLL state to get the best set-up time. The timing numbers are specified under this condition. For example, for a 40-MSPS input clock frequency, use PLL state 3 to get set-up time ≥ 1.1 ns. PLL state 2 can also be used at 40 MSPS, however, the set-up time degrades by 100 to 200 ps (while the hold time improves by a similar amount).
(3) Bit clock and Frame clock jitter has been included in the Set-up and hold timing.
(4) Values below correspond to tdelay, not tPROG