| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
Peter – I do believe Steffen is correct. Looking at slides 3 & 9 in your presentation, I believe they correctly capture the specification as agreed – that the value of N is only dependent on the IL at 20 MHz. However, looking at slides 4-17 both in the graphs and the commentary, you appear to be varying N as the IL varies in frequency – which is the root cause of the behavior you observe – something not in our agreed spec. -george From: stds-802-3-spep2p@xxxxxxxxxxxxxxxxx <stds-802-3-spep2p@xxxxxxxxxxxxxxxxx>
On Behalf Of Steffen Graber Hi Peter, Looking on slides 4 and 5 (and also the other slides in your presentation), it seems, that you modified the “N” value in the alien noise limits depending on where the measured IL exceeds 16 dB or 21 dB and not depending on which is the
IL of the link segment at a fixed frequency of 20 MHz. In the alien noise limits I suggested the factor “N” is changed from 0 to 1, if the IL at 20 MHz is above 16 dB and changed to 2, if the IL at 20 MHz is above 21 dB. This is valid for the complete limit
curve and not adding frequency dependent “edges” into the limit curve, like shown in the graphs in your presentation.
Regarding your concerns related to AN, I agree that we have a lower plateau for the short link segments of 50 dB (for approx. up to 300 m) or 55 dB (for approx. up to 400 m) instead of 60 dB as for 10BASE-T1L.
Related to the suggested TCL values of up to 60 dB (for E1 and E2) and coupling attenuation values of up to 60 dB (for E1 and E2) or 70 dB (for E3) I personally see these high TCL/coupling attenuation values technically critical, and this
not only for e.g. the DIN rail terminals, but also for the complete system. Even if it is possible to meet these values using a good cable, from my perspective it would become pretty “hard” for the complete system to meet these requirements (e.g. a signal isolation transformer instead of capacitive coupling would
be required, there is need for low tolerance components in the power and signal coupling networks, a very carefully designed symmetric layout is needed, shielded connectors might be needed, etc.).
EFT testing for 100BASE-T1L is expected to be significantly more critical than for 10BASE-T1L, but this is mainly caused by the higher communication frequency of 100BASE-T1L compared to 10BASE-T1L, which is much closer to the ringing frequency
of an EFT impulse on the line and thus cannot easily be filtered out by the PHY receiver. Thus having 10 dB better system TCL values would lead to an improvement, but likely not prevent telegrams from being corrupted during EFT testing and on the other side
lead to significant side effects in system design (as not only the cable needs to be improved, but the whole system).
Thanks. Regards, Steffen To unsubscribe from the STDS-802-3-SPEP2P list, click the following link:
https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-SPEP2P&A=1 To unsubscribe from the STDS-802-3-SPEP2P list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-SPEP2P&A=1 |