Hi,

There are many ways to simulate the stability.
The most common one for your BG is to run an AC analysis and look at the open loop gain and phase along with the gain and phase margins. All what you need is a testbench where you put your design and a stimulus with 1V AC DC source (1V to directly read the gain without any divisions). In the old days, we used to open the AC loop and leave the DC loop closed in order to keep all the devices at the right operating point around which you exercise the small signal anaylsis. Nowadays (with Spectre at least), you don't really need to

My advice for you is:

1. Look at your Analog Design book to learn things about feedback analysis, gain, phase ... etc
2. Look at the Hspice doc for more details about the AC analysis. I'm a Spectre user as far as I'm concerned.

If you are really in a hurry, then run a quick transient and if it the output oscillates then it is instable  :-[

Cheers,
Riad.

Hi,all
I have met a similar  problem is ,the amplifier used in my bandgap is self_biased. In detail ,the output of the amplifier is the input(gate voltage) of the bias mosfet. In this case ,I am puzzled  with how to break the loop.

Regards
glacier

hi glacier,
                i guess self bias means all bias voltages will be generated by internally, so whats the problem while breaking? because while analyzing stability through breaking we should not disturb bias voltage.

Thanks,
rajasekhar.

Hi,rajasekhar
  The self bias means the Gate Voltage of the Pmos used as the current bias for the amplifier is generated by the amplifier itself ,in details ,the voltage is the output of the amplifier.
  In this case,can I break at the output of the amplifier ?


Regards
Glacier

hi Glacier ,
                 yes, you can break because breaking the loop means you are not breaking for dc only for ac signal you are breaking.are you using any start up?  Hope you understand.

thanks,
rajasekhar.

hi Aaron,
              what i want to convey is when you are ding stability analysis for ckts have both +ve and -ve feedback break the loop at the common point of both loop so that at a time we can break both loops.

thanks,
rajasekhar.