Hi All,

I was wondering if anyone can point me to a link or help explain the strengths of an inverting summing op-amp circuit over an non-inverting summing circuit. I've read in multiple places that the inverting circuit is preferable though never with a solid explanation of why that is the case. Running some PSpice simulations seems to show that the non-inverting circuit is actually more accurate (probably do to the extremely high input resistance on the summing pin). It seems it may have something to do with the inverting circuit having a "virtual ground" but I don't see how this is really prefferable unless the voltage sources being summed can only source, not sink current. Anyone have any insight into this? Any comments appreciated.

 

The things you mention are not relevant. All op-amps have high input impedance and they all use the "virtual ground" concept although the "virtual ground" may not be at actual ground. I can think of several things that may decide the choice of circuit to use. It all depends on your requirements.

1. Non-inverting amplifiers cannot have a gain less than 1
2. It takes two inverting stages to keep the phase the same
3. Inverting circuits don't like single supply operation, unless a synthetic ground at Vcc/2 is used.

 

The point about op amps is that they try to keep the potential at both inputs the same.
Most configurations are based on this idea.

However non inverting amps are subject to common mode errors and you must take care that the input does not exceed the common mode spec, which may be considerably less than with inverting amps.

Non inverting amps have inherently higher input impedance as the feedback is applied in series and raises the input impedance.

 

Thanks for the replies. I've looked into it a bit more and I think the idea is that as the input impedance of an op-amp pin can be dependent on the voltage on that pin, the non-inverting summing amp circuit's input pin can be exposed to varying voltages depending on the input (versus the inverting summing amplifier circuit who's pin is held around 0) and this is seen as a drawback. But with the extremely high impedance of inputs it seems pretty ridiculous to me. I can't imagine it having much an effect on the accuracy of the circuit.

I think that was also a good point about the common mode voltage between the two.

One other criticism I've read about the non-inverting summing circuit is that the more lines your are adding the smaller each sum component is, which I guess can be valid for a design with many input lines. Overall I guess it depends on the particular design (which is better) but it seems the general "avoid the non-inverting summing amplifier" rule of thumb -at least I've been exposed to- really isn't that valid.

 

....it seems the general "avoid the non-inverting summing amplifier" rule of thumb -at least I've been exposed to- really isn't that valid....

It sounds to me that you understand the situation. But, if you do, I'm surprised that you don't understand why this is a valid rule of thumb. The basic fact that the input impedance varies based on the other input voltage levels shows that there can be strong interactions between various input stages, which could bring on unforseen results in multistage designs.

You would generally like each input stage (before the summing stage) to have minimal interactions between each other. This is a basic principle in system design. The more you can isolate stages and prevent interactions, the easier it is to design and the less likely it is to have unforseen failures. It is sometimes difficult to forsee what interactions, particularly at higher frequencies, might do. Hence, the noninverting summer is less desirable, as a rule.

Of course, all rules have exceptions. So, if you see a potential benefit, and are willing to carefully analyze the entire system, including all interactions, over the entire circuit bandwidth and over all possible input voltage ranges, then it is ok to ignore the rule of thumb. Rules of thumb are really just there to make you be careful when you need to be.

 

I think I see where you are going. This could be especially concerning when summing voltages from sources with high impedance outputs.

Typically, whenever I've used a summing circuit it has been summing the outputs of other op-amps at generally low frequency, in which case I don't think (unless I'm wrong) the generally high impedance of the summing input would be a big deal.

Either way I definitely have a better understanding now, Thanks.