Continue to Site

# Non inverting opamp, with inverted signal

#### cupoftea

Hi,
Would you agree, the opamp in fig2, page 8, of the LT1509 datasheet, is a non inverting opamp?, just that its input is inverted.
So when the transfer function is made out, it should be multiplied by "-1", because of the inverted signal?

Its a non inverting integrator. The inverting fdbk input is grounded. The
NI input is the drop across the current sense R.

Regards, Dana.

Thanks, and so the transfer function Vout/vin = (1 + Zfb/Zin)...and so we should multiply this by minus one because the non inverting input always goes negative of ground?

This is operation :

Regards, Dana.

Thanks, i appreciate it looks like a non inverting opamp...but it cannot be, the transfer function for noninv is 1+ZFB/ZIN...and this is always >1. That cannot be so for an error amplifier in an SMPS...so it in fact, must be a inverting amplifier, with transfer function, -ZFB/ZIN.

Gain falls to 0 if we go far enough out in freq :

This circuit is a NI Integrator :

Regards, Dana.

Thanks , if its a non inverting opamp, then gain is still >=1 even if the freq is high enough to make all the caps shorts.

The G goes away because the interior G elements run out of gas. To wit why the
sim shows this, accurately. Of course the DC case quite different.

And if we use 100% fdbk, a follower, the G is still < 1 because of finite Aol, even at DC.

Regards, Dana.

Last edited:
I believe you are talking at cross-purposes.
1. It's true that the non-inverting circuit has not exactly the inverted transfer function of the inverting circuit but Hni(s) = 1 - Hi(s).
2. With an ideal OP, |Hni(s)| >= 1, as stated.
3. OP open loop gain modifies the transfer function, not specific for this circuit.