Instantaneous power of a MOS transistor

[Gerard Villar]

Hello,

I am researching in power management integrated circuits, and I am
developing my designs with the Cadence software. I was just wondering if
anybody could tell me where can I find information about how the spectre
simulator computes the instantaneous dissipated power of a MOS transistor.

I've been looking for such information throughout the spectre's reference
guides, but I have just found how to ask the simulator to compute the
dissipated power, but not how it does this, and what it takes into account.

Thank you in advance for your attention.

Gerard

 

[Joey]

Hi,

You can choose to save/plot the drain current going into the MOS
transistor and the drain-source voltage.You can then multiply them and
obtain the instant power.

All this can be done by the calculator tool, or SKILL programming.

 

[Gerard Villar]

Hi Joey,

Thanks for your help, but that is not exactly what I was looking for.
I know I can ask Spectre to compute the instantaneous power dissipated by a
MOS transistor, but I need to know what exactly computes, since it seems to
me that it is a bit more complicated than the drain current and
drain-to-source voltage product.
In fact the value of this product does not match the results provided by
Spectre.

Gerard

"Joey" <jonashat202@hotmail.com> escribió en el mensaje
news:1123178659.645320.7180@g44g2000cwa.googlegroups.com...

Hi,

You can choose to save/plot the drain current going into the MOS
transistor and the drain-source voltage.You can then multiply them and
obtain the instant power.

All this can be done by the calculator tool, or SKILL programming.

 

[Gerry Vandevalk]

Are you also looking for gate leakage and/or substrate leakage?

"Gerard Villar" <vallenwood@terra.es> wrote in message
news:dcvibf$jtp$1@defalla.upc.es...

Hi Joey,

Thanks for your help, but that is not exactly what I was looking for.
I know I can ask Spectre to compute the instantaneous power dissipated by
a MOS transistor, but I need to know what exactly computes, since it seems
to me that it is a bit more complicated than the drain current and
drain-to-source voltage product.
In fact the value of this product does not match the results provided by
Spectre.

Gerard

"Joey" <jonashat202@hotmail.com> escribió en el mensaje
news:1123178659.645320.7180@g44g2000cwa.googlegroups.com...
Hi,

You can choose to save/plot the drain current going into the MOS
transistor and the drain-source voltage.You can then multiply them and
obtain the instant power.

All this can be done by the calculator tool, or SKILL programming.

 

[Joey]

Oh, so you need to analyze where is the power being dissipated..

The only thing I can think of is what Gerry said, gate leakage and
substrate leakage, especailly if you're running at a low current value.

You can look at the values of these currents in the results browser.
Again, you can use ocean script to plot/save the transient values of
these currents.

 

[Andrew Beckett]

On 10 Aug 2005 18:55:08 -0700, "Joey" <jonashat202@hotmail.com> wrote:

Oh, so you need to analyze where is the power being dissipated..

The only thing I can think of is what Gerry said, gate leakage and
substrate leakage, especailly if you're running at a low current value.

You can look at the values of these currents in the results browser.
Again, you can use ocean script to plot/save the transient values of
these currents.

What the original question was about was how spectre computes the pwr
value (which may be turned on in the Outputs->Save All form). Not how to save
currents and voltages to calculate the power yourself.

The power calculations are performed within each model, and so I can't
comment on exactly how the calculation is done. Similarly, in a verilog-a model
you can use the $pwr() system task to report to the simulator how much power
the circuit represented by your verilog-a model is dissipating. This is
advantageous because if you measured the power by seeing how much current
is drawn from the supplies, most verilog-a models would not bother to model the
load on the supply caused by the internal circuitry that the model is
representing - it's a waste of effort to do that in most cases.

Regards,

Andrew.