#79: Op Amp Power Supply Considerations: split, single, virtual ground, etc. - a tutorial

This video discusses the power supply considerations for op amps. It talks about split or dual power supply and single supply operation, and why the op amp often doesn't care which you use! It shows how traditional op amps designed for split supply operation can be used in single supply applications. The most important consideration generally is taking care of where the input and output voltages are with respect to the supply rails. The input voltage and output voltage range specifications are examined in a datasheet. The operation of a op amp in a single supply application is examined on an oscilloscope. This operation is compared to a modern rail-to-rail op amp in the same circuit. Notes from the video are here:
www.qsl.net/w/w2aew//youtube/o...

[@jeffbandy3873]

I love that you use the 741 here. This clears up a lot of questions about dual supply requirements for the 741 (and others)! Thanks for moving slowly and deliberately for us beginners. I also learned a few new things about viewing signals on Oscopes in this tutorial. Thanks for making this!

[@johnwhitehead8006]

Still useful information in year 2019. That's 6 years after publishing! Some things don't change.

[@beoddtodd1726]

I have been watching various opamp videos over the last week or 2. this is by far the best demonstration and delivery of Info on this subject I've found yet. You answered many questions I still had with both on paper and physical examples. Lemme see what else you've posted!

[@realchristopher4334]

This is an 'old' video. But a very important one! Many people, including yours truly, blindly use an op-amp and then wondering why it is damaged or why sellers con us, we then buy some more and same story, we buy even more and same story. Lots of wondering and blaming.

[@stevegoodjob5902]

What do you mean? Where is the problem?

[@realchristopher4334]

@@stevegoodjob5902 To put it simply, many people using an operational amplifier or transistor without understanding how it works. This is what I mean and is the problem. Problem is the human and not the operational amplifier or transistor. For instance, many people wonder why audio or signal clips or looks weird. The first thing they do is keep changing and buying different op-amp or transistor to test. They do that until a period of time they get fed up and start to blame the maker of op-amp or transistor. How is it possible that all transistors of all makers fail all together with 99.999999% rate of coincidence?!?!??!?!?! For some weird reasons they are not bothered to change other components and try different levels of power/voltage/current. They are somehow braindead. Doing the same thing expecting different results. You would be surprised the sheer amount of people having an oscilloscope still wonder why a circuit does not work as expected, and play the blame game. Any fools can own an oscilloscope these days anyway. As if having a digital clock and wondering how come it does not have the good old hands moving, that clock must be faulty, buy another digital clock and wonder why again no moving hand, must be another defective clock, buy some more, wonder some more, rinse and repeat.

[@thecombatengineer7069]

binge watching all the circuit tutorials, trying to remember to like and comment on each as well

[@guillep2k]

I've been watching your videos for years, and it just hit me that your notes are actually analog Power Points!!! Thanks for all the theory. Very useful stuff!!

[@vernbaumgardner8374]

Thank you! Finally someone who can explain the difference for connecting split and single power supplies to beginners.

[@jspencerg]

Flawless presentation and demonstration of material. Thank you.

[@mikebrady2193]

Really great tutorial. Clear and simple diagrams with a delivery style that I found very easy to follow.

[@christheother9088]

Wow, this could have save me a lot of pain back when I was figuring out the limits of a 741 the hard way.

[@FreddyBNL]

Very well explaind. Doing and showing works better then text books.

[@bergaouikais5951]

Thanks to you I finally understood why my 741 is behaving in such a weird way, liked and subscribed

[@ec11368]

I really like the way you explain. You always go the right level of the viewers. Thank you very much for all your effort and time. Good bless you

[@Pwaak]

Great information! I very much appreciate when you show the "visual" of what is actually happening within the circuit! Thank You!

[@dragonex68]

This is much better than using a powerpoint :D

[@MrWompz]

Just a man, his pen, and some graphing paper.

[@saturnsoiree]

2022 watcher. thanks for this! helps a lot when u dont have split supply

[@demonring]

Your videos are really instructive. I decide to watch 2 vids of yours every day from now on.

[@andrewferg8737]

Thanks for sharing! I review this video each time I use an op amp to refresh my memory.

[@Bhuntah1]

Great delivery and information. I appreciate your ability to share the knowledge you have in a logical, easy to understand manner. Keep up the great work. By the way, I decided to subscribe. Thank you.

[@alexmaximas384]

I have to thank you for your efforts to explain many interesting things in electronics. I have weak knowledge in this field. This video help me better understand about op amps things.

[@purza95]

2nd time watching in as many days. Great videos! Having graduated in '88 with a BEET degree. I decided to try out these new fangled single supply op amps i.e LM358 lol.... After some frustration and thinking they somehow magically used a single supply your video set me straight. Doh!

[@movax20h]

BTW. Do not go too much beyond power supply range with your input. The AD8032 does support going about 200mV below (above) negative (positive) supply safely, and still maintain linearity of output vs input (obviously you would use gain less than 1 in this case usually). Above that the linearity (and distortion characteristics) can't be guaranteed. Additionally, 500mV is stated as a absolute maximum, beyond which opamp can be permanently be damaged. It is recommended to add external clamping diodes if you expect this to be possibly happening.

[@michaelmolter8828]

Man, having the horizontal lines there makes this crystal clear for me. Always been kind of fuzzy when playing with this on my own scope. I’ll have to make sure to draw those in. Nice trick 👍

[@jonka1]

beautifully described and explained thank you

[@spagamoto]

Thanks, seeing the behavior on the scope helped solidify my intuitive knowledge.

[@txsharpshooter85]

Another excellent video. Thank you for taking the time to make these, as they are very helpful to me.

[@stompreaper]

Extremely effective explanation! Thanks!

[@whitefields5595]

Thanks for this. I am measuring AC current and feeding a 12bit adc so this was essential for my passive cct conditioning.

[@sdscotto]

Great tutorial as always, Alan. Thank you.

[@keithostertag]

Thank you Alan, you have such a great way of explaining things.

[@ryux]

Thank you very much sir... hooking up the oscilloscope really helped understanding the virtual ground!

[@ElectronicsIsFun]

Excellent video. It's great opamp demo of single supply rail to rail operation.

[@supertruckertom]

#2 view. Great work W2AEW. This is better instruction than at our local community college. My neighbor is taking electronics classes there and I shared your videos with him. He is impressed.

[@KetilDuna]

Very nice and easy to grasp presentation - thank you!

[@jrallen870]

Great explanation Alan... always enjoy your videos

[@jeromequelin]

Thank you so much or this excellent video which clear out some confusion on the rail to rail and split supplies concepts. I can't wait to see more OP-AMP tutorials (if you have more coming).

[@wrlgmail]

As always, a great video Alan- perhaps in a future installment you may touch on the areas of device types (e.g. bipolar vs MOSFET) and how that affects ability to drive near the rails, slew rate (speed) considerations (e.g. why use a bipolar op-amp vs MOSFET for speed), etc. Great little tutorial with very nice demonstrations. 73 Bill N4BKT

[@ohmslaw6856]

Man, best teacher on youtube

[@KLATUBARARA1]

Thank you for your work, and the didactic excellent Hand-draw sheets via the review of the Data-Sheets. keep on . It will help me a lot.

[@DavePKW]

Excellent presentation. Thank you.

[@geoffreykeane4072]

Thanks Alan - like many others I refer back to your tutorials often. You balance theory and practical perfectly every time. One thing to note here is that rail-to-rail op amps are expensive! Order of magnitude up on the likes of an LM358.

[@w2aew]

Keep looking. There are many inexpensive rail-rail op amps out there. The Microchip MCP6001 can be found for less than $0.25

[@geoffreykeane4072]

@@w2aew Cheers Alan! Thanks for the heads-up.

[@DAVET38]

Another informative and interesting video. Thanks Alan.

[@CrowsBrother]

Thank you for taking the time,- very enlightening.

[@raguaviva]

Even if this is stuff I already knew I have fun listening to your explanations (plus I always learn something new!) :)

[@tunicana]

as always your tutorials are great...why in schools we dont have teachers like you?

[@fursino]

Beautiful tutorial! Thanks!

[@ilyaorlov]

Thank you! This is very informative!

[@3anthos]

Very VERY good Sir! Thank you!

[@hubercats]

Very nice tutorial. Thank you!

[@rakes150]

thank you for the awesome video alan

[@GaRbAllZ]

Great Job! Clear and concise.

[@akasickform]

Great demo!

[@dhakajack]

Great demonstration and explanation; you make it very intuitive. 73 de AI4SV Jack

[@antronicx]

great stuff as always

[@marisacunha2116]

Thanks for the great information😊

[@sciencetroll3208]

Thanks. Very informative.

[@billfield8300]

Thanks for the great information. I was, however, hoping to learn a bit about how to calculate the power consumption of the op-amp. So.. if the output feeds say an ADC, how do we calculate the power loss through the gain resistors and the chip itself. Once again, great tutorial. Thanks.

[@kelbersaluna]

it's very informative..can you upload more videos about other useful op amp applications, like half/full wave rectifiers, differentiator/integrator, window comparator and etc.. tnx

[@jacka55penguin]

Great stuff! Thanks!

[@billwilliams6338]

older amplifiers used transformer less power transformers, it was called hot chassis which didn't use an AC power transformer but a series of tube filaments to step down the AC voltage. I'm not sure how tube filaments step down an AC voltage, but many you might know. On the schematics it looks like 4 to 6 filaments stepping down 120VAC to 40VAC for the rectifier tube.

[@w2aew]

The filaments are simply resistive elements - so there's a I*R voltage drop across each one in series.

[@movax20h]

Super useful. Are there some performance tradeoffs to the rail-to-rail opams? Noise, bandwidth, input voltage offset, temperature stability, drive/sink output current, etc.

[@MissionFitnessCTC]

Good video!

[@jonathanj-g-yyelle6144]

This did answer some questions! Questions that I didn't think of asking! Do you _teach_ the subject? Because you seem to know possible questions students will ask. Thank you very much for this video!

[@w2aew]

Nope - I am not a teacher - I'm just familiar with the common questions that people might ask.

[@postpoquelin]

Super useful!

[@dannlaboratory]

This video is so damn good, thank you!

[@JulioAstudilloCampos]

Thanks so much for the explanation! Now I understand why my circuit wasn't behaving as I expected :P

[@aliapoonawala5392]

you're the cutest

[@bm830810]

thanks a lot, great videos

[@electrotsmishar]

thank you. very helpful

[@DantalionNl]

when you connect the positive input of an amplifier to 5v volts wont your output be offset to a positive voltage ?

[@johnmcgiv1]

Another very good video Alan. Could it be possible to show your power supplies setup and signal generator and earthing arrangements, many thanks.

[@w2aew]

The schematic on the last page of my notes is accurate. In my case, the signal generator features a floating output (the shell of the BNC isn't grounded), so I was able to put it in series with the supply that I used to adjust the offset.

[@imabeapirate]

for a balanced signal input, where you have a floating differential signal going into the +/- terminals of the opamp, a split supply would be more necessary correct?

[@SCKentrol]

if your opamp circuit with a virtual ground connects to a another opamp that is using a single supply (ex 0-12V), in the gain loop of a non-inverting amplifier, would you connect the resistor to the 0V ground of to the virtual ground?

[@snnwstt]

12:50 "And you plug it in, ... boom ..." Boom??? Reallly??? :-) But in general, there is a (small) perturbation near their zero, though. Good, very good, video. Second point, if the op amp has its offset not defined as being the earth ground we have to adjust the + and - signal with respect to the same zero origine... which can then be THE real headhake. I really hope everything is fine on your side, it is 2021 here!

[@billwilliams6338]

yes true, I just didn't know that different tube filaments has different resistance values which is a voltage drop. I thought as tube filaments all had the same resistance value.

[@JonathanAnon]

Thanks for responding. I'll give it a shot :-)

[@JonathanAnon]

Hi there, I tried to set this up with a non-inverting configuration and a 741CN, but I'm guessing this is not possible because the feedback resistor going to virtual ground from the output will mess up split voltage source? Is it the case that the potential divider only works as a split voltage source if there are no components (across which the voltage drops) between the potential divider and the 741?

[@bm830810]

Hi, we can make an offset on the positive input simply using a resistor divider, but how can we offset the negative input which takes the input signal without using an additional power supply? thanks

[@jmu9969]

Well done! What is the most effective method to bias the op amp using single power supply? Would it be split resistive or using other means? Thank you for your great to the point presentation!

[@w2aew]

resistor divider is certainly most common, but it really depends on the application.

[@ts4124]

thank you!

[@MrWasim100100]

nice video

[@markrhine5809]

Great video, very easy to understand, I am building an audio signal generator and you can use eiher 6 volt sprit or 12 volt single, can you tell me which way to go? Thank you kindly!!

[@w2aew]

From a circuit performance standpoint, they'll be identical. So, it really comes down to details of your requirements. Do you want the output to be centered around ground and DC coupled? If so, then use a split supply.

[@metalhobo46631]

Well, from my interpretation, what you're calling a "virtual ground" is just a voltage reference. You could also call it an "ac ground" because of the very small shunt impedance due to the capacitor. The actual virtual ground exists on the "-" input, not the "+".

[@bm830810]

My question is that you used a variable voltage source to make an offset on both the positive and negative inputs (schematic 7:55), thus your input signal that you want to amplify is floating relative to ground, what if the input signal is not floating? how can i make an offset on the input signal, can i use a voltage divider, what if the input signal source has a very large impedance, should i use a buffer before? thanks

[@Skcrog86]

Hi, very instructive video. What is the role of the 2.2k resistor connected to the + input ? The one that has its value in parenthesis, it is seen in the drawing of your example circuit at around 8:00 . Sorry if that question was asked already.

[@BR.]

+w2aew I would also like to know why this 2.2k resistor was used.

[@w2aew]

A resistor is usually used in that location to help minimize offset due to the input bias current of the op amp inputs. The value is usually equal to the parallel combination of the resistors connected to the inverting input, so it *should* be approx 1.4kohms. I used 2.2k simply because it wasn't critical for the video topic, and it was laying around on the bench near the breadboard ;-)

[@BR.]

Thanks for the explanation! As far as I understand, since the gain here was less than 1.5, I suppose even this resistor wasn't necessary because the error would have been very small. Nevertheless it is important to know about this stuff. I was able to find a bit more in-depth about this watching EEVblog #479.

[@cpackard127]

Around 1:50 you mention you AC couple the bias to ground "to keep it stable" can you point me in the direction of where I can get some additional information on that? I'm going to guess you mean the capacitor will keep the voltage "less spikey" and closer to a flat line? If that is the case I'd think you'd have to carefully select the cap so it let low frequencies pass but not high ones?

[@Banzay20]

So if you are using the non inverting input as a voltage sensing/ smith trigger comparator mode is it still possible to use that input as a midpoint for split supply. Im using lm358 as a compartor to turn of a constant current load when a low voltage is sensed on the non inverting input. But the Op Amp output would not turn completly off it would stay 0.8 volts above ground, the solution may be to supply it with negative voltage to swing it negetive on the output. Il give it a shot, any input of some electro wizard appriciated

[@warrenking1815]

It should be noted that the output can be isolated through a capacitor to remove the dc portion of the waveform.

[@n2mpm]

so many wonderful toys

[@GranQuesoQuemado]

At around the 2 minute mark, I don't think that point should be called virtual ground, unless it's a term that is used pretty loosely. Virtual ground for the op-amp would be there when you have the positive terminal connected to GND and negative feedback configuration, where the feedback action is trying to make the voltage at the negative op-amp input equal to the voltage at the positive input. In the circuit shown at the 2 minute mark, it looks like an open loop.

[@waynegram8907]

I have tried measuring the AC line noise on the primary side of various equipment on an oscilloscope but not sure how do it correctly. How can you measure the AC line noise level using an oscilloscope or is it better to use a spectrum analyzer to measure AC line noise? my Fluke DMM meter can't measure the AC line noise level so what meter would do this?

[@michaelmolter8828]

What is the purpose of the 2.2k resistor at 7:30? I’ve seen these before, or even in the feedback loop of unity gain buffers, but never understood what they’re for?

[@Hadi-zw9mb]

Great video. thank you very much. What happened to output swing voltage if the output current is increased? how to use opamp to drive a low impedance load?

[@w2aew]

It depends on the specific op amp. In general, the available voltage swing is reduced into low impedance loads. Some op amps can drive low impedance loads better than others. Often times a buffer designed for low-impedance loads (like a push-pull stage) is included in the feedback loop of an op amp to improve the low-impedance drive capability.

[@Hadi-zw9mb]

@@w2aew Acutally I used push pull or single transistor at the output stage of opamp; it works fine but not efficient. With single supply, the output voltage has the dc offset.

[@OmBotCult]

How to make a duel or split supply out of a single power source say a 24dc wall wart?

[@waynegram8907]

What are the advantages of using Two Diode Full wave rectification compared to using 4 diode bridge full wave rectification? The Two Diode Full wave rectification has benefits how so

[@billwilliams6338]

But electronic engineers and technicians say that the laminations changes the frequency response of the transformer is this true or false? or do they mean the core material inductance will change value because of the wire winding lamination?

[@losaminos59]

Great op amp video! I'm wondering how did you probe the signal generator input in ch 3? Where did you connect the ground clip at @8:00? Did you connect it to the ground (-) of the DC offset at non-inverting input OR at the node after it (+)? I feel like if I would connect it to the (+) node (or just before the 2.2k resistor) I would then short out the DC offset voltage because the other two oscilloscope channels ground clips are connected to the circuit ground.

[@w2aew]

All scope probe ground clips are connected to the circuit ground.

[@losaminos59]

@@w2aew Thanks for the response sir. I have actually sent a pm about the same on eevblog :D got really anxious about this ground loops ever since i have watched Dave's video on blowing oscilloscope probes.

[@MultiSigen]

Could you please show the design of your voltage offset circuit?

[@w2aew]

+James .Mwage I just used a floating lab power supply.

[@iamAK47]

hi , At 6:42 can you use the same divider on the inverting terminal to get Vcc/2 volts and convert this into a non inverting amplifier?

[@w2aew]

Yes, I could have. But, I used a variable power supply so that I could adjust it up or down easily.

[@ElectroMac74]

hi is there a way to get pdf or jpeg of your notes. thanks

[@kraig800i]

I'm thinking of adding an Eq circuit on to my DIY 200w amp Hifi. On the lm833 the Vcm is +/-12 +/- 14 but the Absolute rating Vcc ~ Vee is 36V does this mean I can use 24v Dc power source to power two lm833 plus the 200w amp it came with or will it fry the added chips ?

[@w2aew]

The LM833 is a pretty low power part, it'll only draw what it needs from the supply, so no danger I. Frying them even though the same supply is also powering a 200W amplifier.