[SI-LIST] Re: hspice simulation issue in VCO
- From: "Ingraham, Andrew" <Andrew.Ingraham@xxxxxx>
- To: <mike_bihan@xxxxxxxxxxxx>
- Date: Tue, 31 Dec 2002 13:28:10 -0500
I have often seen warnings about negative MOS conductance in HSPICE. I
suspect it is something wrong with how the device .MODELs are created,
but I have no proof of this. Perhaps the device strays into a region
where the model equations are not defined as accurately as one would
like, and negative conductance happens. I have to hope that it doesn't
matter that much, i.e., that it is only slight, because I see it so
often (and there's nothing I can do to fix it, with encrypted models).
The other observations you quoted about seeing signal voltages greater
than VDD are correct; except for the one about self heating. (SPICE has
no concept whatsoever about self heating, and anyway it couldn't if we
don't also tell it about heat sinks and all other thermal paths attached
to each device.)
If all you had between the FET and VDD was a lossless inductor, then the
average (DC) voltage on the FET's drain would have to be VDD. If there
is any AC signal there at all, it would have to swing both below and
above VDD.
In an ideal class-B RF amplifier, the output signal swings between
ground and twice the supply voltage.
Why do you think the FETs should limit the voltage? The only way it
would do that, is if it has a low enough breakdown voltage. There is no
P channel FET so nothing to limit the signals to VDD.
Given a tuned circuit with sufficiently high Q, you can cause some
rather large voltages (for a parallel tank) or currents (for a series
tuned circuit) to exist, that are much greater than the stimulus
applied. All you need to do to sustain such signals is to inject a
small amount of stimulus at the right times, and let it build up.
In your HSPICE listing, I do wonder about this statement:
> .TRAN 1.0000 5.0000 START= 0.
Are you really using a 1 *second* transient interval? Assuming that
SPICE is smart enough to keep its internal timestep small, the SPICE
output, sampled at 1 second intervals, would be severely aliased.
Regards,
Andy
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
