The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> Circuit Simulators >> Spectre DC accuracy
https://designers-guide.org/forum/YaBB.pl?num=1177682783

Message started by Visjnoe on Apr 27^th, 2007, 7:06am

Title: Spectre DC accuracy
Post by Visjnoe on Apr 27^th, 2007, 7:06am

Dear all,

I'm observing really strong results when using Spectre dc analysis (operating point) on some small set of circuits. The problem resides in an LDO in power down, where there seems to flow some current both Kirchoff does not hold on a certain node.

Are there any special accuracy settings I should be looking at to improve the DC operating point calculation?

Regards

Peter

Title: Re: Spectre DC accuracy
Post by ywguo on Apr 29^th, 2007, 2:23am

Hi, Peter,

Sure you can improve the accuracy for the DC operating point calculation like other SPICE-like simulator. Choose simulation->options->analog if you use Virtuoso Analog Design Environment. You could set reltol or iabstol there, and any other options in the blank at the bottom of the poped window.

Best regards,
Yawei

Title: Re: Spectre DC accuracy
Post by sheldon on Apr 29^th, 2007, 7:53am

Peter,

Could you provide any additional information about the simulation
issue with the LDO? If Kirchoff's current law is not satisfied at a node,
does the simulaiton converge? Is there anything special about the
power down state? For example, does the circuit have hystersis in the
control loop? Are you performing a dc operating point calculation
or a dc sweep? If the simulation is just a dc operating point, you can try
using the simulator option gmethod=both. If you are running from inside
the environment, you can tirn off all the homotopy and manually set
homotopy=arclength.

Best Regards,

Sheldon