Opamps working on dual supply need to be same (upto certain level) in both the polarity. In case the power supply is not similar in both polarity the dc output will drift to either side depending on the higher amplitude polarity. Sometimes the offset nulling will become difficult to achieve.
Dear Biswajit,
The operational amplifier circuit are based on constant current sources that made independent on temperature and power supply voltage as is possible. It turns out that the excess voltage is dropped on the transistors which are working in the active mode of operation where their voltage can be changed where their current is more or less constant independent of the VCE or VDS in case of MOS transistors.
Consequently it is the dynamic range of the amplifier that must suffer for limiting the value of the power supply voltage. Which means that the dynamic range will be from -VCC1 to + VCC2.
The second order effect as a consequence of the value of VCE is the small signal gain of the amplifier, as the power supply voltage decreases the small signal gain will decrease because of the Early effect.
An another effect , the transistor unity gain bandwidth FT may be lower by lowering one or both power supply voltages as a consequence of Early effect also.
Best wishes.
General op-amps very rarely have a ground node, and the positive and negative supply rails are simply the difference in voltage between the two.
Take for example the classic 741 op-amp:
https://www.ti.com/lit/ds/symlink/lm741.pdf
… where the schematic equivalent is shown on page 4 of the PDF.
Note the offset null is between the two pins and the negative supply rail (page 2).
Looking at the specifications one should consider the following:
1. Output Voltage Swing
2. Input Voltage Range
Note that with the specified voltage supply VS = ±20V
Output Voltage Swing = ± 16V with RL ≥ 10kΩ
Output Voltage Swing = ± 15V with RL ≥ 2kΩ
Input Voltage Range = ± 12V
What this actually says is that the output cannot get closer than 4 to 5 volts to the respective supply rail, and that the input should be at least 8 volts within the supply rails. There is also a maximum voltage difference between the two inputs, depending on topology.
So one could use asymmetric rails, for example +24V and -12V for a non-inverting amplifier that was going to have an output of 0 to 18V positive (where the feedback node resistor was to ground).
The 741 is a bit of an extreme example nowadays, but I do use asymmetric supplies for some of my work so it can be a perfectly reasonable thing to do in the right circumstances, even if it’s only 0.7 of a volt (rail to rail op-amps are often not quite as rail to rail as one would like).
thanks to everybody for nice explanation...
but susan could you tell me where do you use asymmetric power supply opamp , i mean i would like to know what is the application..
As one such application where one would like to use asymmetric supplies I give below an example. Consider a 12V battery operated system for measuring load. One can derive a 5V from this using a 7905 operated from 12V rail. So one gets 0,+5 and -7V supply rails. the load cell can be excited by the 5V supply and provides an output with CM of 2.5V. the -7 is obviously used to provide negative supply to the op amp. What if the 12V battery voltage drops to 10.8. the negative rail alone will see the change and negative rail will become -5.8V. So it is nice to have op amps with a large immunity against supply changes. You can simulate the 741 in circuit maker simulation tool with negative changing by 1V, positive changing by 1V and both changing in a tracking fashion and see how the output changes for each case. Choose a gain of about 100 to start with. Just because the negative is not equal to positive rail voltage, drift does not happen. You do get an offset but drift with time and temperature are not determined by the supply voltage rail values within limits. If both increase or both decrease, the change in offset will be minimal. Output starts changing when one of the rail voltages change. +ve = 5.21V say, & negative changing from -7 to -5./ So next time do not be surprised if an op amp operates off 24 and -12 as Susan has. the designer has a choice to manipulate the supply voltages depending on available input and desired output!!
The op amp will work normally assuming that the supply voltages are enough to enable all transistors to work in the active region. The maximum value of the output voltage will be +12 volts and the minimum value of the output voltage will be -10 volts
An example would be where one is driving a MOSFET which has a significant Gate Threshold Voltage offset e.g. IRF150
http://www.irf.com/product-info/datasheets/data/jantx2n6764.pdf
where Vgs is going to be around 3.5 to 4 volts.
... and one wishes to maximise the voltage swing relative to the input.
This assumes the MOSFET is single ended, and is being used as an AC follower (rather than a switch).
if the specification of the opamp is rail to rail output +12V to -12V but you use only -10V the negative output signal will be clipped.
I think each respectable researcher made sufficient explanations.
So, I don't want to repeat them.
All I want to mention is that, "It depends on the type of OpAmp you are using and on your expectations (performance) from the OpAmp".
An experienced analog circuit designer (who knows well about the OpAmp and knows well what s/he expects from the circuit), can decide easily whether any critical effect may happen and -if required- what precautions s/he should take.
The maximum value of the output voltage will be +12 volts and the minimum value of the output voltage will be -10 volts - but for RRO opa only.
No harm if you have to use different values for supplies. So far as the negative supply is adequately below common mode input (check the data sheets) and the positive supply if more than the CM input 9see data sheets), there is no harm in even providing say -2.5v as negative supply and 32V as positive supply. the effect on offset due to increase of +ve supply from17.25 as also reduction of -ve supply from 17.25 can be calculated based on PSRR. there is no necessity for the supplies to be of same order...! yes, the output range will be decided by supplies. Even for RR) one cannot get output range as -10 to 12, as a small difference is required to be maintained from output to rails. perhaps one could get -9.9 to 11.9V
Will the speed of Vout response ( towards V+ or V- )
be affected by the different supply voltages ?
Given an AC coupled circuit
I suggest that the higher voltage side will ramp up faster
and have less noise ( lower Common Mode Rejection Ratio)
due to having more available energy on that part of the OpAmp circuit.
transistor which are connected with negative supply will cutoff near -10V instead of -12V . It depends on the schematic how good of common mode input and output range and differential mode input and output range.
Dear Biswajit Singh sir,
It's a common question asked by many of the students. You can take any value positive(+Vcc) and negative power supply(-Vcc). Symmetrical values ensures that output can swing equally in either of the direction.
If +Vcc=+12V and -Vcc=-10V, then output has greater chances to get clipped in the negative side.
If +Vcc=+10V and -Vcc=-12V, then output has greater chances to get clipped in the positive side.
I hope, it will clear the doubt.
I am currently running some op-amps with +7.5V and -2.5V supplies as they are input buffers for a fast ADC which has a 0 to +5V input range.
Dear Susan Parker ,
For sure you are doing it by considering the common-mode input range. This requires some sufficient background and experience about OpAmps (including knowledge about the inherent structure), which avail the confidence to do it.
Nevertheless, I understand why some young researchers hesitate to apply a "less frequent" method and tend to get opinions and advices of the experienced. This is normal and inevitable since some of such usages may not be found in classical books easily.
I remember, when I was young, I was not able to understand the reasons for some of the conventional usages in circuit design and I was also not able to find them in classical books. I regret, I was a bit stubborn to try searching harder hoping to find them in some book, due to which I lost substantial amount of time in some occasions, without a result. Today, I suggest my students and young colleagues to make relevant research first, but if it does not help for a long time, to ask an experienced eventually.
Best regards...
THE OP AMP - Analog Devices:
https://www.analog.com/media/en/training-seminars/design-handbooks/Basic-Linear-Design/Chapter1.pdf
- Badges
- Science topic