Subject: RE: SpectraQuest timing measurement
From: Todd Westerhoff (twester@hhnetwk.com)
Date: Fri Jan 25 2002 - 09:09:02 PST
Matt,
You ALWAYS want to measure the driver output into the test load at Vmeas,
because you're using that measurement to compensate for the loading
condition the semiconductor vendor uses to measure Tco. SPECCTRAQuest calls
that measurement "Buffer Delay", other tools call it other things.
SPECCTRAQuest ALWAYS measures Buffer Delay to the Vmeas threshold, no matter
how you set the "measure delays at" option.
"Measure delays at" refers to the way the simulator measures flight times,
BEFORE the buffer delay adjustment is made. How you set this depends on
methodology and philosophy.
Nominally, you'd measure flight times to Vil/Vih and subtract the buffer
delay measured to Vmeas. That is the "hybrid" option you refer to, and how
the SPECCTRAQuest software works by default.
There is another school of thought in SI - that simply measuring flight
times to Vil/Vih is too conservative. It's based on the observation that
setup/hold specs for a part are measured the same way Tco is - with
reference to a specific threshold. Therefore, the reasoning goes, you can
measure flight times to the point where the receiver crosses the reference
threshold, and, as long as the signal is within certain quality (i.e.
non-monotonic) and slew-rate limits, the flight time measurement is valid.
The required slew-rate is (or should be) part of the receiving device's
input spec. If the input signal does NOT meet the quality/slew rate
requirements, well, then you still have to use Vil/Vih, or in some cases, a
derating formula.
SPECCTRAQuest allows you to use either technique, and that's what the
"measure delays to" switch is for. SPECCTRAQuest makes the assumption that
the Vmeas voltage is the same as the input reference voltage (in the "single
point" case) - not a bad assumption, and it's the only "single point"
measurement voltage in the IBIS model anyway.
Other tools support the two different methodologies in their own ways - I
don't happen to remember. The issue is a generic one, though, and that's
why I've posted it here.
Which technique you use is up to you and your assessment of your design,
suppliers and the accuracy of your models. Using the "measure delays to
Vmeas" method certainly makes your timing equations look better, as it
removes uncertainty in the flight times. However, given the number of
issues SI tools typically DON'T model (like, power/ground noise), you may to
use the conservative approach, to have a little timing margin left over for
some of those things that go "bump" in the night.
Hope that helps,
Todd.
Todd Westerhoff
SI Engineer - Hammerhead Networks
5 Federal Street - Billerica, MA - 01821
email:twester@hhnetwk.com - ph: 978-671-5084
============================================
"Oh, but ain't that America, for you and me
Ain't that America, we're something to see
Ain't that America, Home of the Free
Little pink houses, for you and me"
- John Mellencamp
-----Original Message-----
From: owner-ibis-users@server.eda.org
[mailto:owner-ibis-users@server.eda.org]On Behalf Of ruston, matt
Sent: Thursday, January 24, 2002 4:35 PM
To: ibis-users@server.eda.org
Subject: SpectraQuest timing measurement
All:
Hi. I have a specific question regarding Spectraquest SigXplorer. Under
analysis/measurement modes, they give me the option to "measure delays at"
either Input Thresholds or Vmeasure.
It seems to me that I want a hybrid of these two options. I want to measure
the driver output at the Vmeasure point of the driver to correlate to
standard load timing. I then want to use the Input threshold of the receiver
(either Vil or Vih, depending on low-to-high or high-to-low transition).
Am I mis-reading what my options are? Does what I want make sense? Any and
all thoughts and experiences are welcome.
Regards,
Matt
This archive was generated by hypermail 2b28 : Fri Jan 25 2002 - 09:50:36 PST