| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
Although my field of expertise is other than data-com, I will take the risk and ask a question that might have trivial answer...
The question I have is in reference to phantom connection, when we apply power on the data pairs.
When we connect power supply between the center-tap of the Rx and Tx transformers, we create low impedance path for high frequency (the data signal) between the two channels, as the input stage of the power supply includes a capacitor across the lines . Now, I understand that this will be true for common mode signals, while the data signals are differential, But as some of the differential signals do convert to common mode, and vise versa. Would this arrangement be able to meet the cross-talk attenuation between Rx and Tx lines required?
If it is really a problem, I guess a solution can be in a form of series inductors to the DC lines, to increase the impedance to high frequency. If I am correct, the presentations presented on the last meeting did not took that into consideration, as far as addition costs, or the impact it might have on the discovery method.
Avinoam Levy
PowerDsine Inc.
avinoaml@powerdsineusa.com
Phone (631) 756 4680 Ext. 305
Fax (631) 756 4691
Web-site: www.powerdsine.com