[802.3_ISAAC] Suggested insertion loss limit for 802.3dm

[Ragnar Jonsson <rjonsson@xxxxxxxxxxx>]

All,

 

In their Annapolis presentation, David Cliber and Bert Bergner suggested to use 3m of CX174e and 12m of CX31a as a reference for insertion loss limit for 802.3dm. I think that this is an excellent idea.

 

To create the 802.3dm insertion loss limit, I suggest to use the cable combination from David and Bert, using 100C reference insertion loss with additional 15% design margin.

 

Based on these ideas I like to suggest the following insertion loss limit for 802.3dm:

                IL_limit(f) = 1.5 + 0.002*f + 0.3*sqrt(f)

where IL is in dB/m and f is in MHz.

 

The enclosed plot shows the resulting insertion loss limit as well as the insertion loss for the 3m+12m reference cable at different temperatures. Note that this suggested limit is similar to the SDP insertion loss limit that can be found in IEEE 2977-2021 (A-PHY) and the SDP insertion loss limit in the document that was in a liaison to 802.3 from ASA. The suggested limit allows slightly more insertion loss than what is specified for coax cables in the A-PHY and ASA documents.

 

The proposed limit was derived based on the following calculations:

1. The CX31 and CX174 are defined in ISO 19642-11 as a limit on insertion loss, based on the commonly used formula IL = a + b*sqrt(f) + c*f
2. The parameters for the CX31a cables at 20C are approximately a=6e-2, b=1.4e-2, and c=7.4e-5, where IL is in dB/m and f is in MHz (note that these parameters are approximations to the ISO 19642-11 limits, based on curve fitting)
3. The parameters for the CX174e cables at 20C are approximately a=0.11, b=2e-2, and c=2e-4, where IL is in dB/m and f is in MHz (note that these parameters are approximations to the ISO 19642-11 limits, based on curve fitting)
4. The parameters for the total insertion loss for 12m of CX31a + 3m of CX174e at 20C is simply the linear combination:
• a = 12*6e-2 + 3*0.11 = 1.05
• b = 12*1.4e-2 + 3*2e-2 = 0.228
• c = 12*7.4e-5 + 3*2e-4 = 1.488e-3
5. To change from 20C to 100C the parameters can be scaled up by 15% giving
• a = 1.05 * 1.15 = 1.2075
• b = 0.228 * 1.15 = 0.2622
• c = 1.488e-3 * 1.15 = 1.711e-3
6. Adding 15% design margin gives
• a = 1.05 * 1.15 * 1.15 = 1.3886
• b = 0.228 * 1.15 * 1.15 = 0.3015
• c = 1.488e-3 * 1.15 * 1.15 = 1.968e-3
7. Finally rounding the parameter values to “nice” numbers gives the proposed insertion loss limit
• a = 1.5
• b = 0.3
• c = 0.002

 

I would love to hear some feedback on the suggested insertion loss limit.

 

Ragnar

---

To unsubscribe from the STDS-802-3-ISAAC list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-ISAAC&A=1

Attachment: e240530b_CX_12m_3m.jpg
Description: e240530b_CX_12m_3m.jpg