Re: gEDA-user: Opamp Model vs Open Loop Gain

[John Doty <jpd@xxxxxxxxxxxxx>]

On Sat, 13 Aug 2005, User Tomdean wrote:

> I was seeing unexpected results from simulating a circuit and decided
> to replicate the datasheet.
> 
> For the LF353/NS, openloop gain is 115db to 20Hz with a linear drop to
> zero at 5MHz, more or less.  Vs=+-15, Rl=2K.
> 
> I assumed this to be with the non-inverting input grounded, AC to the
> inverting input, and 2K to ground from the output.

With a real physical circuit like this, your first check would be to take
a voltmeter and see if you have a sane DC operating point. The answer
would likely be no for this circuit. Checking a simulated circuit isn't so
different, but I bet you skipped that step.

With most opamps, the product of the input voltage offset and the voltage
gain exceeds the output voltage range. An open loop test thus needs an
adjustable DC bias to get the output voltage to a reasonable level. I
suggest doing a DC sweep with Vin, and setting its DC value to the level
that yields 0 at V(6). Then, do your AC analysis.
 
> 
> I see 15db to 100KHz with a drop to zero at 5MHz.  I expected the
> model to be similar to the datasheet.
> 
> The circuit is 
> 
> * LF353 MODEL TEST vs DATA SHEET
> Vin 2 0 DC 0 AC .001 0
> Vp  4 0 +15
> Vn  5 0 -15
> R3  6 0 2K
> X1  0 2 4 5 6 LF353/NS
> .include ../opamp/lf353_sp3.mod
> 
> Spice> run
> Spice> ac dec 100 1 10Meg
> Spice> plot db(v(6)/v(2))
> 
> What am I doing wrong?
> 
> tomdean
> 


John Doty          "You can't confuse me, that's my job."
MIT-related mail:                       jpd@xxxxxxxxxxxxx
Other mail:                             jpd@xxxxxxxxxxxxx