TRAN analysis in SPICE

[hyc]

.tran analysis

Hi, All:

This may seem very obvious to many of you, but I am somewhat confused on what we are trying to verify with TRAN analysis in SPICE.

Take for example, the typical spec for OpAmp are

DC Gain,
Gain Bandwidth
Input Common Mode Range
Slew Rate
Output Voltage Swing Range
CMRR
PSRR
....

All these spec, once designed could be verified with running DC and AC analysis alone, couldn't they? So why do we ever need TRAN analysis, there doesn't seem to have a delay time spec to meet like in digital ckt, Can anyone explain what stuff we have to and can only verify with .TRAN analysis? Also if we find something wrong from running .TRAN, analysis were they correlated to any of above mentioned Spec.

Thanks in advance.

 

[LvW]

how 2 find cmrr in pspice

The last four of the parameters mentioned by you can only be simulated using the TRAN analysis feature since they are no small signal parameters.

 

[akhilrathore]

pspice vcm an vs

hi,
.tran or transient analysis is needed only when you want to see how your signal is varying with time, foe slew rate you will need transient analysis.

hope it will help

 

[FvM]

pspice tran parameter

CMRR and PSRR are defined as small signal (.AC) parameters in my opinion, slew rate is .TRAN as said, voltage swing can be either a statical (.DC) or a large signal dynamic (.TRAN) parameter.

 

[LvW]

settling time pspice

CMRR and PSRR are defined as small signal (.AC) parameters in my opinion, slew rate is .TRAN as said, voltage swing can be either a statical (.DC) or a large signal dynamic (.TRAN) parameter.

For my opinion, also CMRR should be simulated under large signal conditions because it depends on frequency as well as on signal amplitudes. And, more than that, I think also the maximum voltage swing should be tested via Tran analysis because of its frequency dependence.

 

[FvM]

small signal analysis hspice

For my opinion, also CMRR should be simulated under large signal conditions because it depends on frequency as well as on signal amplitudes.

The same could be said about gain parameters. If you check any OP data sheet, no voltage magnitude is mentioned for CMRR and PSRR, which indicates a small signal measurement in my opinion. Also the standard literature, e.g. Grey/Meyer Analysis and Design is discussing CMRR as small signal phenomenon, see particularly chapter 6.2.5.

 

[eecs4ever]

transient analysis in pspice

You should also check the settling time using a transient sim when using an OP-AMP in feedback. This is important in switch capacitor circuits.

 

[LvW]

spice tran

.............
If you check any OP data sheet, no voltage magnitude is mentioned for CMRR and PSRR, which indicates a small signal measurement in my opinion. Also the standard literature, e.g. Grey/Meyer Analysis and Design is discussing CMRR as small signal phenomenon, see particularly chapter 6.2.5.

Yes, I know that in textbooks CMRR is handled as a fixed parameter for each opamp. And it´s true, that normally CMRR is specified in data sheets without mentioning any test voltage. (By the way: How often data sheets tell us not the truth?).
However, from my own experience I know that CMRR is (somewhat) dependent on the amount of Vcm.
And, more than that, NATIONAL has specified in some cases different CMRR values - for different Vcm ranges.
But, in general I agree that the dependence of CMRR on amplitudes can be neglected.

 

[FvM]

pspice settling time

I know that CMRR is (somewhat) dependent on the amount of Vcm.

Yes, but this can be analyzed in .AC analysis.

 

[hyc]

spice tran analysis

Thank you for everyone's reply. Although all people agree the slew rate and the settling time require TRAN analysis, however, can't slew rate be easily verified with the derived equation ID/Cc ? Also, isn't settling time closely related to the bandwidth and the phase margin, which are AC parameters? In another word, are there any relationship beween frequency response and the settling time?

 

[LvW]

cmrr spice

Hi HYC !
1.) Regarding settling time: The situation is not as easy as it seems. The dependence between frequency response and settling time is more involved (e. g. because of zeros and pole-zero-doublets of the transfer function).

2.)A general remark regarding TRAN analysis (as you doubt if it makes sense or really is necessary):
If you like to simulate instead of measuring, the TRANSIENT analysis is the most important analysis and gives by far more realistic results than AC or DC runs.
AC analyses give a small signal approximation of the circuits behaviour neglecting any non-linearities - and the DC analysis neglects all capacitive and induktive influences. Thus, even in cases where a DC analysis seems to be sufficient, it can be advantageous to run a TRAN analysis with a linear increasing input voltage.
The same applies to the AC case. When you have time enough and are patient, a TRAN analysis for a lot of different frequencies gives you more realistic information about the frequency behaviour than an AC analysis.
But in many cases the results of AC runs are sufficient to evaluate the basic frequency properties (but don´t forget: it´s only small signal !).
Regards

Added after 2 minutes:

Not to forget: TRAN analysis gives you information on transient behaviour starting at the switch-on time !

 

[hyc]

cmrr measurement in spice

LvW:

Thank you very much for the enlightening explanation.

Best Regards,