1
\$\begingroup\$

I'm trying to build a non-inverting amplifier with a gain of two, with a little bit of filtering. I need to amplifier a signal from 0Vdc-to-1.65Vdc to 0Vdc-to-3.3Vdc, and ideally have a low-pass cut-off around 1.3kHz

Looking through the net I can't really understand how to do an AC analysis on a single supply opamp circuit, there are a million-and-one examples of dual-supply OpAmp filters, but I've not seen any demonstrating a single supply OpAmp filter AC analysis for frequency response - can anybody help?

enter image description here


I think I found the solution - and that is to correct (for the AC Analysis only) the bias points, so this means the AC source is offset by mid-rail (in this case 1.65) as is the bottom side of R1

However - I would still love to make sure this approach is correct Cheers!

\$\endgroup\$

3 Answers 3

1
\$\begingroup\$

There isn't really much difference for AC analysis for a single-supply op-amp. The input needs to be offset to accommodate the op-amp's input range limits.

In your example, add a 833mV DC supply between the AC1 generator (-) pin and GND. This will provide an offset for the signal. The DC offset will be amplified by 2 to yield 1.66V at the output with AC = 0.

Once you've done that, make sure your offset AC swing is still within the range of the op-amp common-mode range. This can be quite a bit smaller than you expect. The AD8605 supports rail-to-rail input. You'll need to reduce the AC component so that the output isn't going to clip. Suggest maybe 100mV.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks! Think I've got it. Cheers! \$\endgroup\$
    – Marcos
    Commented Oct 12, 2023 at 6:24
1
\$\begingroup\$

In general, opamps don't "know or care" if they have a single or dual supply. It's just a question of what you call the 0V reference. There are techniques to generate an artificial mid rail "0V" (it's called a "virtual ground") if you are constrained to a single supply, and even specialist chips to do this, but the op amp operation is unaffected, so long as it works works well.

There are op amps like this one, billed as "single supply" opamps, but in this case the description really means that the output can operate all the way to 0V (the "negative" supply). This is not the general case.

In short, and op amp can be "single supply" if you ensure that it is biassed somewhere sensible between its supply pins. In a "split supply" situation, we call that point "0V", but it could just as well be +2.5V in an app with 5V supply only. What we need to take care of is that the op amp output will get close enough to the supply pins for our intended application.

From this, it is clear that SPICE analysis of the opamp circuit itself does not differ at all between the two cases. It's just a question of what we label as 0V.

\$\endgroup\$
0
\$\begingroup\$

You need to DC-bias the input so your amplifier is not saturated. Normally you'd put it in the middle of the output range. Simulate that first (.op). You can stack a DC supply on the AC input.

You should also ensure that with your desired signal level the amplifier does not clip or unduly distort. You'll need a transient simulation for that, with appropriate bias and input signal. This is not necessary for the AC analysis but the AC analysis is not useful if the amplifier is going to be clipping.

Don't bother changing the input level for the AC analysis, it only scales the output results and adds potential confusion. AC analysis replaces the nonlinear components with linearized equivalents about the operating point, so nonlinear effects are ignored.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.