When Die is selected, the simulator should make all timing
measurements (and corrections for the standard test loading
conditions) at the die of the device. This is useful for users
who perform timing analysis based on the static timing analyis
of an ASIC or custom device sans package. It is often the case
in the speciality asic and custom world to provide timing at the
die pads. In fact, it is much more difficult to provide worst case
timing for these devices at the pins, since that would require
much more characterization. The important thing to note here
is that device specifications such as Tco, setup, hold ... etc
do not necessarily have to be specified at the pin of a device.
This is the customary place since one can place a probe on
a pin to take an external measurement.
When Pin is selected, the simulator should make all timing
measurements and corrections at the pin of the device.
In the circuit simulations the package parameters are still
used. The measurement location parameter just specifys
where the measurement is made.
Note that SI Location parameter was especially included since
signal integrity measurements are only valid at the die of
a device. It is customary to set the Si Location to die, in order
to make valid measurements of signal integrity,
"Lalit B. Shinde" wrote:
> Dear SI-Gurus,
> IBIS spec provides Si_location and Timing_location parameters on
> Component to specify measurement location position. Currently
> these parameters can take any one of the following values "Die"
> and "Pin". However, the spec does not talk about which all
> "Timing" measurements should use these parameters.
> Flight time (Buffer Delay) of a Driver is usually calculated with
> Package pin parasitics included. When IBIS model on Driver has a
> Timing_location set to 'Die', do users calculate flight time of
> the Driver at Die Pad (excluding package pin parasitics) ?
> Biggest challenge lies in correlating this to the Common Clock
> System timing equation where
> Driver Tco(max) + Flight Time + Setup needs to be less than
> Clock Period.
> When Timing_location is set to Die, do users get/use the Tco data
> at Die Pad ? If yes, can someone share their experience ?
> If Tco data always includes the Package parasitics, should
> Timing_location parameter on a Driver be ignored for Buffer Delay
> calculations (And the spec be updated accordingly) ?
> (o) 978-262-6475
> **** To unsubscribe from si-list: send e-mail to firstname.lastname@example.org. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****
**** To unsubscribe from si-list: send e-mail to email@example.com. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****