[ibis-quality] Re: I/V voltage sweep requirements
- From: "Mike LaBonte (milabont)" <milabont@xxxxxxxxx>
- To: <ibis-quality@xxxxxxxxxxxxx>
- Date: Tue, 12 Jul 2005 14:22:30 -0400
I also think the 1 amp limit may be low, but it's what the group agreed
on. The real reason for the limit may have something to do with limits
in the dyanamic range of existing curve viewers, wich sometimes are
unable to zoom in on the finer details when the overall voltage range is
large. But with higher device speeds higher currents seem possible.
Taking it to the extreme, the movement of one electron would be one amp
if the time period were short enough.
About zero slopes, I don't see a problem. In fact, distributing on-die
termination resistance across two clamp curves with zero-slope endings
is a fairly standard technique. All that matters is the combined I/V
curves, which admiteddly should have no zero slopes. It's easy to
believe that simulators might "pad" the slope of the combined curves for
convergence, if necessary.
Maybe we should add an explanation for VCC, something like: "... VCC, as
given by [Pullup Reference] or [Voltage Range]". Maybe it should also be
specific that the range of typ/min/max I/V tables must be compared
against typ/min/max VCC, respectively.
Mike
-----Original Message-----
From: ibis-quality-bounce@xxxxxxxxxxxxx
[mailto:ibis-quality-bounce@xxxxxxxxxxxxx] On Behalf Of Tom Dagostino
Sent: Tuesday, July 12, 2005 1:06 PM
To: ibis-quality@xxxxxxxxxxxxx
Subject: [ibis-quality] Re: I/V voltage sweep requirements
Mike a few comments
I've seen buffers that were capable of sinking over an amp of current in
the active region. The one amp limit seems low to me.
Won't zero slope areas on curves cause problems? I know some simulators
will add some slope to IV curves if required and it is unlikely that any
simulation would get to that current level.
Define Vcc, is it Vcctyp? Vccmin or Vccmax? I ask this because I'm
occasionally asked to model a buffer over a very wide voltage range: min
= 2.5, typ = 3.3 and max = 3.6V. Even for normal 3.0/3.3/3.6 Volt
modeling what is Vcc?
Tom Dagostino
Teraspeed(R) Labs
13610 SW Harness Lane
Beaverton, OR 97008
503-430-1065
tom@xxxxxxxxxxxxx
www.teraspeed.com
Teraspeed Consulting Group LLC
121 North River Drive
Narragansett, RI 02882
401-284-1827
-----Original Message-----
From: ibis-quality-bounce@xxxxxxxxxxxxx
[mailto:ibis-quality-bounce@xxxxxxxxxxxxx]On Behalf Of Mike LaBonte
(milabont)
Sent: Tuesday, July 12, 2005 8:09 AM
To: ibis-quality@xxxxxxxxxxxxx
Subject: [ibis-quality] I/V voltage sweep requirements
I have an action item to revise I/V voltage sweep requirement checks.
There is a little controversy, so I'd like to propose it by showing
fragments with commentary here. There are 4 separate checks for I/V
sweep range. The [Pullup] check looks like this:
4.3.4 {LEVEL 1} [Pullup] voltage sweep range is correct
The sweep for [Pullup] should be made between -vcc to 2*vcc.
Pullup tables are relative to Pullup reference. The Pullup table
combined with the power and ground clamp tables should be monotonic.
Non-monotonic combined I-V tables must be documented.
+++ IBISCHK NOTES
+++++++++++++++++++++++++++++++++++++++++++++++++++++
Proposal: ibischk caution message if the lower endpoint of the
[Pullup] voltage column is greater than -vcc, or if the upper
endpoint
is less than 2vcc.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
My suggested change to that one is:
4.3.4 {LEVEL 1} [Pullup] voltage sweep range is correct
The sweep for [Pullup] should be at least from -vcc to 2*vcc.
Pullup tables are relative to Pullup reference. The Pullup table
combined with the power and ground clamp tables should be monotonic.
Non-monotonic combined I-V tables must be documented.
For ECL devices [Pullup] data must cover the range of vcc to vcc -
2.2.
An exception is that any [Pullup] table may be truncated at points
where current reaches the limit specified in 4.3.3. In this case there
must be two points with the same current at the end of the table, so
that the final slope of the curve is flat.
+++ IBISCHK NOTES
+++++++++++++++++++++++++++++++++++++++++++++++++++++
Proposal: ibischk caution message if the lower endpoint of the
[Pullup] voltage column is greater than -vcc, or if the upper
endpoint
is less than 2vcc.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
The above:
a) Allows voltage ranges greater than required.
b) Handles ECL differently
c) Allows narrower sweep ranges as long as table reaches 1A and is flat.
We never discussed c), are there any objections?
Also, I discovered that we did cover ECL in a separate check:
4.3.15 {LEVEL 1} ECL models I-V tables swept from -Vdd to +2 Vdd.
I-V tables in ECL models should be swept from -Vdd to +2 Vdd, even
though the operating range is narrower.
Obviously this is easy to miss, written into a separate check a page or
two away from the others. I think we should change 4.3.4 through 4.3.7
to cover ECL and eliminate 4.3.15.
Lastly, I would like to propose that [GND Clamp] and [Power Clamp]
tables would be allowed to end at the point where current cross the zero
point in the case of buffers with on-die termination. The final slope of
these would have to be zero to avoid double counting.
Mike
---------------------------------------------------------------------
IBIS Quality website: http://www.eda.org/pub/ibis/quality_wip/
IBIS Quality archives: //www.freelists.org/archives/ibis-quality
To unsubscribe send an email:
To: ibis-quality-request@xxxxxxxxxxxxx
Subject: unsubscribe
---------------------------------------------------------------------
IBIS Quality website: http://www.eda.org/pub/ibis/quality_wip/
IBIS Quality archives: //www.freelists.org/archives/ibis-quality
To unsubscribe send an email:
To: ibis-quality-request@xxxxxxxxxxxxx
Subject: unsubscribe
---------------------------------------------------------------------
IBIS Quality website: http://www.eda.org/pub/ibis/quality_wip/
IBIS Quality archives: //www.freelists.org/archives/ibis-quality
To unsubscribe send an email:
To: ibis-quality-request@xxxxxxxxxxxxx
Subject: unsubscribe
Other related posts:
