Re: [SI-LIST] : Urgent help needed on HSPICE simulation

About this list Date view Thread view Subject view Author view

From: Perry Qu (perry.qu@alcatel.com)
Date: Wed Aug 02 2000 - 06:53:01 PDT


Hi! Dr. Lynne:

Thank you very much for your reply. I would like to explain in detail what I'm
trying to simulate in HSPICE: it is a prelayout simulation for a high speed
differential driver driving a LVDS receiver through a short transmission line.
The single ended voltage should swing from 0.9 to 1.4 V at the receiver pin.
Thus I specify the DC voltage at receiver pin to be 1.2 V. I tried several
method that you mentioned to specify the DC operating point:

1. No DC operating point specified.
1. Use .nodeset to suggest DC point.
2. Use .IC with UIC in .TRAN
3. Use .IC without UIC.

For case 1, I got unsymmetrical voltage swing at the differential receiver pins,
with one from 1.2 - 1.4 and the other 0.5-1.2 V.

For case 2, the error is "no DC convergence"

Case 3: I got the following error:

 ** error**: iob_loads1:9996:
    Initialization < store small-signal parameters away >
    is not supported for Input/Output Buffers

Case 4: I got symmetrical voltage swing at differential reciever pins but wrong
amplitude: from 1.4 - 2.8 V.

Any suggestions ?

Regards

Perry

Lynne Green wrote:

> How to outsmart SPICE:
>
> Use .NODESET to SUGGEST a starting point for convergence.
> SPICE will converge starting from this voltage. Useful when
> there are multiple DC operating points possible, or when there
> is a lot of feedback in the circuit (e.g. transistor junction
> capacitances).
>
> Use .IC (together with the TRAN Use IC command) to FORCE
> the voltage at convergence. Useful when you want to start
> a transient run from a non-stable point. If you do this and let
> the simulation run to a long STOP time, with no changes on
> the input sources, you can find what DC point SPICE thinks it
> wants to converge at.
>
> There are MANY reasons for not converging, from a latch that
> has no preferred DC state, to a very large circuit with feedback,
> and many many more. SPICE books usually have an entire chapter
> on this.
>
> Large voltage swings at receiver could indicate an unterminated
> transmission line. Many other possibilities.
>
> Also, check node naming on the circuit, especially if you are
> using differential drivers or receivers. If you print out a node
> list in SPICE, you might see some unintended "shorting".
>
> Lynne
>
> Dr. Lynne Green
> Product Marketing Engineer
> HyperLynx, A Division of PADS Software, Inc.
> 14715 NE 95th St, Suite 200
> Redmond, WA 98052
> 425-497-5081
> FAX 425-881-1008
> lgreen@hyperlynx.com
> http://www/hyperlynx.com
>
> -----Original Message-----
> From: owner-si-list@silab.eng.sun.com
> [mailto:owner-si-list@silab.eng.sun.com]On Behalf Of Perry Qu
> Sent: Tuesday, August 01, 2000 11:40 AM
> To: si-list@silab.eng.sun.com
> Subject: Re: [SI-LIST] : Urgent help needed on HSPICE simulation
>
> Hi!
>
> Refering to my question 2 posted this morning, I found that the simulation
> results are
> quite different depending on how you specify and whether you specify the DC
> operating
> point for the device. Is there any general guideline on we should specify
> the DC point
> using .ic or .nodeset, etc. in HSPICE ? What could be wrong if I got the
> following error
> message:
>
> "**error**: no convergence in operating point"
>
> Thanks in advance.
>
> Best Regards
>
> Perry
>
> Perry Qu wrote:
>
> > Good morning, everyone:
> >
> > Thank you for your help on my previous question posted on the SI list. I
> tried out
> > the SPICE simulation that you guys suggested and I have the following
> problems:
> >
> > 2. In another case, I used a driver with SPICE model to drive a LVDS
> differential
> > receiver with IBIS model. The voltage swing I got at the receiver pin much
> larger
> > than the LVDS specifications. Do you have any idea what could be wrong ?
> >
> > Thank you very much for your help.
> >
> > Best Regards
> >
> > Perry
> >
> > SI Specialist
> > Alcatel CID
> > Kanata, ON, Canada
> > Tel: (613) 7846720

**** To unsubscribe from si-list or si-list-digest: send e-mail to
majordomo@silab.eng.sun.com. 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 : Wed Nov 22 2000 - 10:50:56 PST