Re: [SI-LIST] : HSPICE Start Up conditions

About this list Date view Thread view Subject view Author view

From: Umesh Painaik ([email protected])
Date: Wed Mar 22 2000 - 12:16:32 PST


I had faced simmilar problem where the spice deck didnot accurately
represent the actual chip
Some things I did to clear hspice of any charges were
1) Use .option dvdt=2 (if using hspive97.2 etc) this should enable
hspice's most accurate algorithm
2) see if the sim spec has any gmindc specifications (it needs to be
less than 1e-5 for accurate op point simulation)
3) set up a simple sim spec
what i had to do eventually was was give the spice deck back to the
manufacturer with the test case which I used which showed that it didnot
work. They ended up remeasuring the process corner parameters.
Umesh Painaik
VCS

Kai Keskinen wrote:

>
>
> I have been given an encrypted driver model. It is a PECL driver. The
> app notes say to use 4.5k pulldowns. The sim works now but it is
> extremely sensitive to initial charge values on those caps. I'm told
> the physical chip works with the recommended pull downs and coupling
> caps so is this a problem with the model or something related to
> HSPICE? I suspect the model but I can't really do much with it.
>
> Kai Keskinen
> Equipment and Network Interconnect
> Nortel Subsystems and Performance Networks (NSPaN)
> (613)-765-3506 (ESN 395)
> [email protected]
>
> -----Original Message-----
> From: Tom Dagostino [SMTP:[email protected]]
> Sent: Wednesday, March 22, 2000 1:50 PM
> To: si-list
> Subject: RE: [SI-LIST] : HSPICE Start Up conditions
>
> One thing I would look at is the output impedance of the driver.
> With the 4.5K pull down there is little standing current keeping
> Zout high (I'm assuming and ECL like driver). During the falling
> transition the transistor is off and the 4.5K is the source of
> current.
>
>
>
> Tom Dagostino
> ICX Modeling Group
> [email protected]
> 503-685-1613
>
>
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]]On Behalf Of Kai
> Keskinen
> Sent: Wednesday, March 22, 2000 7:46 AM
> To: '[email protected]'
> Subject: [SI-LIST] : HSPICE Start Up conditions
>
>
> Hi SI Folk:
>
> I had a lot of problems getting a spice model of a gigabit
> ethernet chip to run properly. The differential outputs are
> tied to ground with 4.5kOhm resistors and then coupled to
> the long T-lines (W-element) with ~1nF caps. They are
> terminated with 100ohms across the differential pairs just
> before the receiver. It turned out that if we did not set
> the initial charging conditions on the coupling caps to a
> very narrow range of values, the simulation never actually
> worked. One or both of the differential outputs did not at
> all resemble what the chip should transmit even for very
> long simulation times.
>
> My questions are: Why did the circuit not charge up the caps
> and then work as would happen in reality? Is this a very
> common effect when running HSPICE or other sims?
>
> Thanks,
>
> Kai Keskinen
> Equipment and Network Interconnect
> Nortel Subsystems and Performance Networks (NSPaN)
> (613)-765-3506 (ESN 395)
> [email protected]
>

**** To unsubscribe from si-list or si-list-digest: send e-mail to [email protected]. In the BODY of message put: UNSUBSCRIBE si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu
****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Thu Apr 20 2000 - 11:35:50 PDT