Fangyi,
One way of adding slow random jitter is passing a uniformly distributed random
process through a LF filter then converting it into phase. But what is "slow"
depends on the signal rate and receiver characteristics. Typically, CDR can
easily eliminate jitter whose spectral density is already small at signal rate.
Then, it is not slow jitter, but medium frequency (MF) and HF jitter that
affects the results.
How can we separate jitter into LF, MF and HF portions and make them follow the
signal rate of the devise and its CDR LF filter cut-off frequency?
Vladimir
From: ibis-macro-bounce@xxxxxxxxxxxxx [mailto:ibis-macro-bounce@xxxxxxxxxxxxx] ;
On Behalf Of Fangyi Rao
Sent: Friday, July 19, 2019 6:09 PM
To: michael.mirmak@xxxxxxxxx; IBIS-ATM (ibis-macro@xxxxxxxxxxxxx)
<ibis-macro@xxxxxxxxxxxxx>
Subject: [ibis-macro] Re: Draft BIRD for separating high- and low-frequency Tx
Rj parameters
Thanks for the clarification, Mike.
"For question (2), am I correct in thinking you are referring to repeater-based
simulations?"
No. I was asking about point-to-point channels. In time domain simulations we
need to add signal waveforms from Tx1 and Tx2 at Rx input. Tx1 signal has a
phase offset relative to Tx2 signal due to Tx_Rj_LF in Tx1. This phase offset
is constant from bit to bit because Tx_Rj_LF is defined as low frequency
jitter. How do we add randomness to this phase offset? Let the tx1 phase drift
slowly (but at what frequency?), or do a sweep on the Tx1 phase?
Thanks,
Fangyi
From: ibis-macro-bounce@xxxxxxxxxxxxx<mailto:ibis-macro-bounce@xxxxxxxxxxxxx>
<ibis-macro-bounce@xxxxxxxxxxxxx<mailto:ibis-macro-bounce@xxxxxxxxxxxxx>> On
Behalf Of Mirmak, Michael
Sent: Friday, July 19, 2019 4:48 PM
To: IBIS-ATM (ibis-macro@xxxxxxxxxxxxx<mailto:ibis-macro@xxxxxxxxxxxxx>)
<ibis-macro@xxxxxxxxxxxxx<mailto:ibis-macro@xxxxxxxxxxxxx>>
Subject: [ibis-macro] Re: Draft BIRD for separating high- and low-frequency Tx
Rj parameters
[EXTERNAL]
Fangyi,
Thanks for the note. The focus for the change was on statistical simulation,
so my answers will be a bit limited until I can do more research; my initial
reaction is that the phrase is really "added to the jitter accounting budget at
the receiver".
For question (2), am I correct in thinking you are referring to repeater-based
simulations?
- MM
From: Fangyi Rao <fangyi_rao@xxxxxxxxxxxx<mailto:fangyi_rao@xxxxxxxxxxxx>>
Sent: Friday, July 19, 2019 4:02 PM
To: dmarc-noreply@xxxxxxxxxxxxx<mailto:dmarc-noreply@xxxxxxxxxxxxx>; Mirmak,
Michael <michael.mirmak@xxxxxxxxx<mailto:michael.mirmak@xxxxxxxxx>>; IBIS-ATM
(ibis-macro@xxxxxxxxxxxxx<mailto:ibis-macro@xxxxxxxxxxxxx>)
<ibis-macro@xxxxxxxxxxxxx<mailto:ibis-macro@xxxxxxxxxxxxx>>
Subject: RE: Draft BIRD for separating high- and low-frequency Tx Rj parameters
Question 2 is for time-domain simulations.
From: ibis-macro-bounce@xxxxxxxxxxxxx<mailto:ibis-macro-bounce@xxxxxxxxxxxxx>
<ibis-macro-bounce@xxxxxxxxxxxxx<mailto:ibis-macro-bounce@xxxxxxxxxxxxx>> On
Behalf Of Fangyi Rao
Sent: Friday, July 19, 2019 3:58 PM
To: michael.mirmak@xxxxxxxxx<mailto:michael.mirmak@xxxxxxxxx>; IBIS-ATM
(ibis-macro@xxxxxxxxxxxxx<mailto:ibis-macro@xxxxxxxxxxxxx>)
<ibis-macro@xxxxxxxxxxxxx<mailto:ibis-macro@xxxxxxxxxxxxx>>
Subject: [ibis-macro] Re: Draft BIRD for separating high- and low-frequency Tx
Rj parameters
[EXTERNAL]
Hi Mike,
I have two questions regarding sentence "which is to be added to the behavior
implemented by the EDA tool by modifying the signal at the receiver" in the
Tx_Rj_LF definition.
1. By "signal at the receiver" do you mean the input signal or the output
signal of the DLL?
2. If there are two transmitters in the channel, one has Tx_Rj_LF and the
other doesn't, how shall the signal of the receiver be modified?
Thanks,
Fangyi
From: ibis-macro-bounce@xxxxxxxxxxxxx<mailto:ibis-macro-bounce@xxxxxxxxxxxxx>
<ibis-macro-bounce@xxxxxxxxxxxxx<mailto:ibis-macro-bounce@xxxxxxxxxxxxx>> On
Behalf Of Mirmak, Michael
Sent: Thursday, July 18, 2019 4:33 PM
To: IBIS-ATM (ibis-macro@xxxxxxxxxxxxx<mailto:ibis-macro@xxxxxxxxxxxxx>)
<ibis-macro@xxxxxxxxxxxxx<mailto:ibis-macro@xxxxxxxxxxxxx>>
Subject: [ibis-macro] Draft BIRD for separating high- and low-frequency Tx Rj
parameters
[EXTERNAL]
Please find enclosed a draft BIRD adding parameters to separate Tx Rj into
high- and low-frequency components. The simple approach here is based on the
conversation held in ATM some months back.
Comments are welcome; I may not be available for the next ATM meeting but
should be able to introduce it at the following meeting.
Thanks!
- MM
