[802.3_NGBASET] (Rx CMNR) Ad Hoc Meeting
Again - more comments for the adhoc.
Bryan
-----Original Message-----
From: Moffitt, Bryan
Sent: Tuesday, September 15, 2015 7:47 AM
To: 'Larry Cohen'; 'Cibula, Peter R'; 'Jim Bauer'; 'German Feyh'; 'George Zimmerman'; 'Hossein Sedarat'
Subject: RE: Revised report on CMNR test with directional coupler
One more point -
Although there is an attempt to provide a completely generic Annex, it seems to me as though there are only two "goals" if I can call them such;
1) driving a fixed level into the clamps for anechoic comparison and for dealing with the shielded cabling which has a less clearly defined common mode dominance across a broader frequency range.
2) driving a level that attempts to maintain a fixed common mode for common mode rejection tests (this is clearly limited to unshielded testing in the lower frequency range), which would involve a separate calibration step.
Is this consistent with your views?
Bryan
-----Original Message-----
From: Moffitt, Bryan
Sent: Tuesday, September 15, 2015 7:27 AM
To: Larry Cohen; Cibula, Peter R; Jim Bauer; German Feyh; George Zimmerman; Hossein Sedarat
Subject: RE: Revised report on CMNR test with directional coupler
I still support this work and the overall issue of maintaining the input drive level as opposed to the clamp output level, however still feel it may be too early to insert coupler details into the draft. We need to prove in operation out to 2 GHz, including confirmation that results will be consistent with anechoic results.
If I'm properly understanding the coupler proposal, test operation is not actually improved with the coupler, the coupler just provides a means of monitoring input level. This is certainly helpful but it should be noted that this comes with additional complexity that also needs to be documented. This points out a shortcoming of the current draft that should be addressed with or without the coupler - drive level calibration, which I did in my chamber comparison but failed to impress in my draft comments (my apologies). I'm expecting that the results should be the same with or without the coupler if this calibration is followed, although this too should be explored.
Attached is the very first step that should be documented (with either approach), the calibration of the signal source. The calibration result would then be used to control the source level in the actual test. Note that the calibration with the coupler is somewhat more complicated (although easy with a 4 port network analyzer).
I also feel that Larry's earlier proposals are more important than the coupler since they bear on the actual quality of the test result:
1) include a loss pad to minimize reflection effects on the signal generator (the coupler does do this)
2) minimize switching transients during the test sweep
Perhaps it can go either way at this point and further improvement can be done in the next go around? Did I understand George correctly that since the addition of 25G to bq, the entire document is open for comment after this?
Regards,
Bryan Moffitt
CommScope Systems Engineering
________________________________________
From: Larry Cohen [Larry.Cohen@xxxxxxxxxxxx]
Sent: Monday, September 14, 2015 9:19 PM
To: Cibula, Peter R; Moffitt, Bryan; Jim Bauer; German Feyh; George Zimmerman; Hossein Sedarat
Subject: Revised report on CMNR test with directional coupler
To all:
It has been pointed out to me by Jim Bauer that the main result in slides 7 and 8 does not clearly specify the how the power at the clamp AE port is derived from the measured coupler power output (the red line on the plot in slide #8). The attached revised report has revised slides #7 and #8 that will hopefully clarify the result. Sorry about any confusion this error might have caused.
Thanks,
Larry