CDLF05C (Actual measurement)
|
|
|
|
10Gbps PRBS Response |
|
Differential Transmission / Reflection
|
 |
|
 |
| [X-axis: 20ps/Div, Y-axis: 400mV/Div] |
|
|
|
|
|
|
|
12.5Gbps PRBS Response |
|
Differential Group Delay
|
 |
|
 |
| [X-axis: 16ps/Div, Y-axis: 400mV/Div] |
|
|
|
|
|
|
CDLF10C (Actual measurement)
|
|
|
|
10Gbps PRBS Response
|
|
Differential Transmission / Reflection
|

|
|
 |
| [X-axis: 20ps/Div, Y-axis: 400mV/Div] |
|
|
|
|
|
|
|
12.5Gbps PRBS Response
|
|
Differential Group Delay
|
 |
|
 |
| [X-axis: 16ps/Div, Y-axis: 400mV/Div] |
|
|
|
|
|
|
CDLF15C (Actual measurement)
|
|
|
|
10Gbps PRBS Response
|
|
Differential Transmission / Reflection
|

|
|
 |
| [X-axis: 20ps/Div, Y-axis: 400mV/Div] |
|
|
|
|
|
|
| 12.5Gbps PRBS Response |
|
Differential Group Delay
|
 |
|
 |
| [X-axis: 16ps/Div, Y-axis: 400mV/Div] |
|
|
|
|
|
|
CDLF20C (Actual measurement)
|
|
|
| 10Gbps PRBS Response |
|
Differential Transmission / Reflection
|
 |
|
 |
| [X-axis: 20ps/Div, Y-axis: 400mV/Div] |
|
|
|
|
|
|
|
12.5Gbps PRBS Response |
|
Differential Group Delay
|
 |
|
 |
| [X-axis: 16ps/Div, Y-axis: 400mV/Div] |
|
|
|
|
|
|
CDLF30C (Actual measurement)
|
|
|
| 10Gbps PRBS Response |
|
Differential Transmission / Reflection
|
 |
|
 |
| [X-axis: 20ps/Div, Y-axis: 400mV/Div] |
|
|
|
|
|
|
| 12.5Gbps PRBS Response |
|
Differential Group Delay
|
 |
|
 |
| [X-axis: 16ps/Div, Y-axis: 400mV/Div] |
|
|
|
|
Within tens of picoseconds of delay time adjustment, a substrate trace can be used for the return circuit. For a 100ps adjustment, either a substrate trace of a coupled differential pair, as shown in Fig.1, or a substrate trace of an uncoupled differential pair, as shown in Fig.2 can be assumed. However, the trace area in either case will become very large. Moreover, as shown in Fig. 1, a time lag arises between the differential pairs due to the imbalance at the corner. Common mode noise will occur at the mid-point, as will the fear of noise radiation. It is very likely that high-quality 10GHz+ band width characteristics cannot be obtained for the circuit shown in Fig.2.
Fig.1 and Fig.2 are verified with an electromagnetic simulator and the characteristics are compared to the results utilizing the CDLF10C. The published Recommended Land Pattern is assumed for the substrate: W=0.38mm, t=0.3mm, and εr=4.1. The underside of the substrate is GND plane.
|
|