Hi Walter,
We had another issue that with the text in the PAM4 BIRD rev 5:
It is mentioned on page 6 that
"If these parameters are declared as Usage InOut or Out, the algorithmic model
is expected to output
values from the AMI_Init call for the EDA tool to use during waveform and eye
processing."
i.e. the threshold values are supposed to be output during Init call. This is
restrictive as the VGA bound settings themselves can change and adapt and in
the case of higher order PAM that means the thresholds will change over time.
As such, if the model decides to communicate the threshold values it should be
able to do it during AMI_Getwave too.
Thanks,
Ambrish.
From: ibis-macro-bounce@xxxxxxxxxxxxx [mailto:ibis-macro-bounce@xxxxxxxxxxxxx]
On Behalf Of Ambrish Varma
Sent: Monday, April 06, 2015 2:06 PM
To: Muranyi, Arpad; IBIS-ATM
Subject: [ibis-macro] Re: Updated PAM4 BIRD rev 5
Hi Arpad,
I am sorry - but I am not quite certain what you are trying to say here. Page
171 and 180 are saying the same things. The threshold is a value used by the
EDA tool to determine a 0 or 1 (for NRZ) for the output waveform.
All I am saying is that either the AMI model tell us what that threshold is OR
let the EDA tool determine that.
Thanks,
Ambrish.
From: Muranyi, Arpad [mailto:Arpad_Muranyi@xxxxxxxxxx]
Sent: Monday, April 06, 2015 1:03 PM
To: Ambrish Varma; IBIS-ATM
Subject: RE: Updated PAM4 BIRD rev 5
Ambrish,
On pg. 171 I see the following sentence:
"10. The output waveform of the receiver AMI_GetWave function represents the
voltage waveform at the decision point of the receiver. The EDA tool completes
the simulation/analysis with this waveform."
To me this says that the decision point is after the AMI model
in the Rx, i.e. the input to the core logic on the chip...
The statement you quoted is preceded by "The output waveform is
expected to be the waveform at the decision point of the receiver" and this
output waveform is also the output of the GetWave function,
i.e. after the AMI Rx model, i.e. the input to the core logic.
Thanks,
Arpad
===============================================================
From: ibis-macro-bounce@xxxxxxxxxxxxx<mailto:ibis-macro-bounce@xxxxxxxxxxxxx>
[mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Ambrish Varma
Sent: Friday, April 03, 2015 11:04 PM
To: IBIS-ATM
Subject: [ibis-macro] Re: Updated PAM4 BIRD rev 5
Arpad,
As per IBIS ver 6.0 (page 180), the decision point is "the point in the
receiver where the choice is made as to whether the data bit is a "1" or a "0""
- thereby requiring a threshold.
Thanks,
Ambrish.
From: ibis-macro-bounce@xxxxxxxxxxxxx<mailto:ibis-macro-bounce@xxxxxxxxxxxxx>
[mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Muranyi, Arpad
Sent: Friday, April 03, 2015 6:25 PM
To: IBIS-ATM
Subject: [ibis-macro] Re: Updated PAM4 BIRD rev 5
Ambrish,
I thought that the "decision point" is at the core logic input,
i.e. after the Rx and all of its signal processing bells and
whistles, while these thresholds are on the input side (pad) of
the Rx... Or am I missing something?
Thanks,
Arpad
=================================================================
From: ibis-macro-bounce@xxxxxxxxxxxxx<mailto:ibis-macro-bounce@xxxxxxxxxxxxx>
[mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Ambrish Varma
Sent: Friday, April 03, 2015 4:22 PM
To: wkatz@xxxxxxxxxx<mailto:wkatz@xxxxxxxxxx>; IBIS-ATM
Subject: [ibis-macro] Re: Updated PAM4 BIRD rev 5
Hello Walter,
I have a quick question on the threshold parameters 'PAM4_UpperThreshold,
PAM4_CenterThreshold, PAM4_LowerThreshold':
If these parameters are not present, can't the EDA tool get the threshold
values from the waveform at the Rx decision point?
Why should the EDA tool be "expected to use the following values:
PAM4_UpperThreshold = 0.333
PAM4_CenterThreshold = 0.0
PAM4_Lower_Thresholf = -0.333" ?
There is a good chance that the waveform at the Rx decision point will be
distorted due to ISI and if the threshold parameters are absent, the EDA tool's
expectation of 0.333, 0 and -0.333 will be off.
If the parameters are absent, the EDA tool should be able to come up with the
threshold values.
Thanks.
Ambrish.
From: ibis-macro-bounce@xxxxxxxxxxxxx<mailto:ibis-macro-bounce@xxxxxxxxxxxxx>
[mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Walter Katz
Sent: Tuesday, March 31, 2015 4:31 PM
To: IBIS-ATM
Subject: [ibis-macro] Updated PAM4 BIRD rev 5
All,
I am enclosing an update PAM4 BIRD (now at rev5). Please review carefully. I
plan on submitting this formally to the IBIS Open Forum meeting on April 10.
Walter
Walter Katz
wkatz@xxxxxxxxxx<mailto:wkatz@xxxxxxxxxx>
Phone 303.449-2308
Mobile 303.335-6156
