Is the supply for a differential op amp compulsory or can we operate those without a supply? I just want to amplify the difference of inputs, I'm not in need of any supply voltage.

  • \$\begingroup\$ The transistors inside the op-amp needs to be biased properly for amplification. You need to provide supply for that. \$\endgroup\$
    – nidhin
    Jul 12 '14 at 10:07
  • \$\begingroup\$ The range of generally overlapping but certainly not totally equivalent answers and comments shows that the question is useful. It would be a shame if the downvotes caused it to end up being closed or discouraged the asker. \$\endgroup\$
    – Russell McMahon
    Jul 12 '14 at 11:42
  • \$\begingroup\$ A transformer with more turns on the secondary is a voltage amplifier with a fixed gain and a loss in power available at output compared to input. \$\endgroup\$
    – Russell McMahon
    Jul 12 '14 at 11:43
  • 1
    \$\begingroup\$ @RussellMcMahon, a transformer is not an amplifier. \$\endgroup\$ Jul 12 '14 at 12:00
  • 1
    \$\begingroup\$ @AlfredCentauri (1.) Note that I did not make an unqualified statement to that effect at any time. I have pointed out, because it is useful to do so in this context, that in a limited but useful range of cases, a transformer is functionally identical to an amplifier. (3). Please see my added "gain block" example and diagram in my answer. If you disagree with the important technical point made would you please explain why in technical terms. \$\endgroup\$
    – Russell McMahon
    Jul 12 '14 at 14:30

Is the supply for a differential op amp compulsory or
can we operate those without a supply?
I just want to amplify the difference of inputs, I'm not in need of any supply voltage.

  • Generally, no, you cannot operate them without an external power supply.

  • In VERY special cases, yes, voltage (but not power) amplifiers that do not require an external power supply can be designed.

You ask specifically about an "op amp" = "operational amplifier".
Using the Wikipedia definition this is necessarily a powered device with high gain, and it would be difficult to build one that did not require power supply separate from the input signal. While it is conceptually possible to imagine an amplifier which derives power from its input signals and amplifies eg voltage but at reduced power level, this would be extremely unusual, would have significant limitations in what it could do and would usually be more an exercise in perversity than a useful device.

In the general case the answer is short - see below.

Most of the material below relates to special cases - in the "special cases" section below I describe devices which break the general "rule".


In the general case, the answer is no.

An opamp without a supply is like a car without a motor. *
It can't do anything, it won't go anywhere and it has absolutely no power output.

So, no, you MUST power an opamp to allow it to amplify signals.

How you power it MAY be creative, although you need to know what you are doing to have any chance if success. ie in SOME cases you may be able to obtain power for an opamp or other IC from portions of your circuit that are not formal power supplies. But, you are still supplying it with power and it is necessary to do so.


Very specialised subsets of amplifiers exist which can be powered by their input and so do not need separate power supplies. These cannot be power amplifiers but can amplify eg voltage or current.

For example, if you want an AC amplifier and are prepared to accept certain significant limitations then you can describe eg a transformer with a step up voltage ratio as an amplifier, and it will not require any external power supplies. eg a 110 VAC to 230 VAC step up transformer can be used in the USA by visitors from eg Britain or NZ to operate 230 VAC equipment from US 110 VAC mains. This transformer IS a voltage amplifier with a gain of ABOUT 230/110 =~~ 2.091. It has a current gain of less than 1 (approximately 1/2.091 ~= 0.478 and is somewhat lossy for power as some of the input power is lost in the windings and laminations.

This is a real amplifier in that if you decreased or increased the input voltage by a limited percentage of its initial value the output voltage would decrease or increase by about the same amount. I say "about" as things like magnetisation current and 'matching' of input and load will mean that the Vout/Vin voltage ratio is somewhat dependant than more than just the transformer turns ratio. However, to a first approximation it functions as a voltage amplifier.

In the diagram below we have two "gain blocks".
In the upper block we have a transformer with a 1:3 turns ratio.
In the bottom gain block we have unknown circuitry.
In both cases, for a range of frequencies and loads and certain other conditions, Vout ~= 3 x Vin in both cases. If Vout/Vin = voltage gain is the only aspect which is of interest to us then the transformer can be considered to be "amplifying" the voltage. The transformer is, for the parameters of interest, not functionally different to a "formal" amplifier.

enter image description here

As pointed out by Alfred, unpowered transformers cannot, by definition, have power gain. Otherwise put: Where power amplification is required an unpowered transformer fed with only the input "signal" cannot deliver more power to the output than is provided to the input.

An unusual device does exist which uses"magnetics" (eg windings on a 'steel' core or cores) + external power input to provide overall power gain. The device may use multiple semi independent windings and not be a true transformer or may incorporate true transformer action as part of its operation. Such a device is commonly called a "magnetic amplifier". It works by making use of the non linear magnetic saturation aspects of the magnetic core material and uses a low power signal to change the degree of magnetic saturation in a core shared by a high power winding. The change in core saturation produces an impedance swing in the high power winding which produces changes in power level much greater that what is required to cause them.
See section at end - "Powered transformers and magnetic amplifiers".

A hybrid amplifier: Transformer voltage gain & MOSFET power gain

The website of the person who designs these amplifiers, an IEE member, makes interesting reading. Their comments and terminology leave no doubt about what they think about provision of voltage gain without powering.
[Susan Parker, MIEE, Zeus Power Amplifier - A Zero Feedback Power Amplifier, for Audio and Other Applications.

Susan's Audiophonics site.
She says - "I use transformers by design as the primary building blocks and gain stages
and either tube or solid state devices as followers. This topology is further explained on this site."

Here is an interesting example of a high quality audio amplifier - one of a family - which uses transformers to provide voltage gain, and semiconductor voltage followers, with a voltage gain of less than 1, to provide power gain.
Here the actual voltage amplifier is wholly unpowered and provides the voltage gain for the whole system. The amplifier as a whole could not work with only the power amplifier or only the voltage amplifier.

Zeus Power Amplifier - note that the two MOSFETs are voltage followers with a voltage gain of less than unity.

enter image description here

Powered Transformers & Magnetic Amplifiers:

If you define a transformer as being an unpowered device (ie no energy other than signal fed into it) and you define an amplifier as a device which MUST have power gain then you have by definition made it that a transformer is not an amplifier.

However, there is no reason that a transformer cannot if desired be provided with a power supply so that "magnetics" are used to manipulate energy flows.
In such cases the "powered magnetic device" can have both voltage and power gain - so much so that when this arrangement is used the term "transformer" is replaced by the term "magnetic amplifier".

A magnetic amplifier is a specially designed transformer (or electromagnetic device consisting of windings on a (usually) steel core, with a DC power supply that can be used to provide voltage and or power gain. There are variants on this theme with small level DC controlling higher power AC, AC signals producing AC power out (functionally essentially indistinguishable from an AC amplifier implemented by other means) and more.

Magnetic amplifiers often use a slowly varying DC signal to vary the power of an AC signal, but may be and are used as true AC in AC out "powered" amplifiers.

Lots of examples here - Magnetic amplifiers .
And here are DIY examples using ordinary off the shelf 12V transformers to control significant AC power with a small DC (varying) signal. Homemade magnetic amplifiers
Magnetic amplifiers MAY have a single winding but often have multiple windings and power gain.

Wikipedia - magnetic amplifiers

Mgnetic amplifiers 99 pages - 1960. Excellent. Thanks to Dan Stahlke's stack exchanghe answer hetre

Cute 1951 US Navy related paper Magnetic Amplifiers - another lost technology

Same maybe ? http://teslapress.com/magamp.pdf

From whence this image (and many like unto it)

enter image description here

*"All models are wrong. Some models are useful" - George Box.

  • \$\begingroup\$ The phantom downvoter strikes again. Hopefully (for his sake) it's not Alfred. I'd be rather sad for anyone who thinks this answer is "not useful". \$\endgroup\$
    – Russell McMahon
    Jul 12 '14 at 16:14
  • \$\begingroup\$ Russell, you really should consider deleting the comment above because, honestly, I think it diminishes you. Rest assured, I leaves feedback on those very rare occasions I cast a downvote and frankly, I can't quite find the interest to read your answer much less downvote it. \$\endgroup\$ Jul 13 '14 at 2:21
  • \$\begingroup\$ @AlfredCentauri NB - :-) -> That sounds like it's worth -1 for technical reasons :-) | All of us "are" the vector sum of a complex mix of things - I meant what I said and I think it's positive. I genuinely hoped it was not you and I genuinely pleased to hear it wasn't. People who downvote answers like this one are either USUALLY misguided technically or trying to convey a personal grievance, and if a downvote does not genuinely reflect the technical does state of an answer it does others harm who see a lesser score and may not spend the time learning something new and interesting. ... \$\endgroup\$
    – Russell McMahon
    Jul 13 '14 at 4:01
  • \$\begingroup\$ ... You and I have been disagreeing over a tautology which has little to do with the technical merits of the point raised and that is a shame. I wholly agree with you that an unpowered transformer cannot have power gain. I drew attention to the fact that a "powered transformer" solution does exist - an unusual device in this age but still much used and highly useful. The term "magnetic amplifier" reflects the powered + magnetics aspects and is the exception that proves your point. A shame we can't just grin ruefully about the path taken to reach a mutual agreement. \$\endgroup\$
    – Russell McMahon
    Jul 13 '14 at 4:07

I just want to amplify the difference of inputs

From the Wikipedia article "Amplifier"

An electronic amplifier, amplifier, or (informally) amp is an electronic device that increases the power of a signal.

Thus, the need for a power supply to the op-amp. The additional power of the amplified signal must come from somewhere since energy is conserved thus, power must be supplied to the op-amp in addition to the signals to be amplified.

A transformer can be used to increase the voltage of a signal (the secondary voltage can be a multiple of the primary voltage) but not the power of a signal (ideally, power in equals power out) so it isn't an amplifier and, thus, doesn't require a power supply.

  • \$\begingroup\$ Alfred - you can validly have a voltage amplifier, or a current amplifier or even unusual things like an impedance amplifier, that do not need power amplification, and which can produce far less power in the load than is dissipated in the input circuitry. \$\endgroup\$
    – Russell McMahon
    Jul 12 '14 at 11:40
  • \$\begingroup\$ @RussellMcMahon, a voltage amplifier has very low (ideally zero) output impedance and very high (ideally infinite) input impedance and thus, can provide enormous power gain. The fact that, if the load is an open circuit, there is no power gain does not take away from the essential fact that there is available power gain. Also, see, for example: van.physics.illinois.edu/qa/listing.php?id=17627 \$\endgroup\$ Jul 12 '14 at 12:04
  • \$\begingroup\$ Stop now :-). An ideal voltage amplifier may have power gain by default but even a circuit using an ideal voltage amplifier can be arranged to be ideal or not overall or to have power gain or not. While it may perhaps seem pedantic to make these points its desirable to do so as otherwise "what's normally done" may obscure the basic principles being conveyed. | Motorcars have 4 road wheels and a steering wheel. Yes? Marvellous J.A.P. engined Morgan 3 wheelers are then not motor cars. Nor are various Reliant products. Some may agree :-). Some of the early vehicles which would ... \$\endgroup\$
    – Russell McMahon
    Jul 12 '14 at 12:12
  • \$\begingroup\$ ... usually be described as early motorcars had tiller steering. Excluding the latter may not cause too many problems but excluding 3 wheelers from the ranks of cars will bring a host of pugnacious Morgan enthusiasts down on your head. My not really whimsical point is that constraining out definitions either to 'what is usually done' or to secondary characteristics of an ideal example is liable to 'cause problems'. || A voltage amplifier has voltage gain. Anything else is secondary and demanding inclusion of secondary characteristics obscures the point. \$\endgroup\$
    – Russell McMahon
    Jul 12 '14 at 12:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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