[BP] Broadband interference levels
The nice thing about using Broadband (noise like) interference in the
Rx interference tolerance test is that it is easy to provide a lower
bound on the level which is tied to the (informative) channel spec.
Going back to moore_02_0605 and using the same method, with the current
channel specs, i get the total crosstalk:
Aggressor unequalized Aggressor (-1/8,3/4,-1/8)
800mV aggressor 6.6mV (RMS) 4.9mV (RMS)
1.2V aggressor 10.0mV (RMS) 7.4mV (RMS)
By the way, the enclosed plot shows the crosstalk spectrum for a 800mV
aggressor, in dB(V RMS in 1Hz).
Looking at spanga_c1_0406, i infer from page 12 that ICR_+3p6 is about
at the spec limit of crosstalk, and from pages 13 and 11 that its
crosstalk has about the effect of 12Vp sinusoidal interference, when
the aggressor is 800mVp-p, equalized. On page 15 fulvio shows two
channels, one getting close to the spec limit, the other just over.
Looking at page 16, and assuming that things sum RSS, i see that
for the two channels the "self interference" is equivalent to
9-12mVp sinusoidal. I will choose 11mV since the worse channel is
closer to the edge than the better on.
Translating this 11mVp sinusoidal to RMS noise, and considering that
12mVp sinusoidal is equivalent to 4.9mV RMS noise gives max "self
interference" of 4.5mV.
RSSing this into the previous table gives:
Agressor unequalized Agressor (-1/8,3/4,-1/8)
800mV aggressor 8.0mV (RMS) 6.7mV (RMS)
1.2V aggressor 10.9mV (RMS) 8.7mV (RMS)
I do not think that either of the limits are reasonable specs, we need
something in between. So i suggest that the Rx interference tolerance
test be run with a noise-like signal with 9.0mV RMS amplitude.
charles
--
|--------------------------------------------------------------------|
| Charles Moore
| Avago Technologies
| ISD
| charles.moore@avagotech.com
| (970) 288-4561
|--------------------------------------------------------------------|
