what is a virtual ground?

[peterthenovice]

I was watching Dave's op-amp videos. It got me thinking about his ucurrent.  In his circuit it has a virtual ground. Dave hasn't done a video on the topic. I wonder if it would be a good fundamentals Friday  as a edition to the new op-amp series of videos. What is  a virtual ground? do you need a ref ic like the one below? what are the benefits.

http://www.ti.com/lit/ds/symlink/tl431a.pdf
http://en.wikipedia.org/wiki/Virtual_ground

[c4757p]

I agree, it's a good topic - or even just "ground" in general.

Why would you need a TL431 for a virtual ground? That actually seems rather undesirable to me - you want it to track the midpoint between the rails. A voltage divider like Dave used ought to do just fine.

If you don't care about power consumption you don't need an IC at all - a voltage divider with a decoupling cap will do for many applications. The divider has to be rather stiff, though. The op amp, configured as a buffer, allowed him to use huge resistor values to conserve power.

[peterthenovice]

I agree, it's a good topic - or even just "ground" in general.

Why would you need a TL431 for a virtual ground? That actually seems rather undesirable to me - you want it to track the midpoint between the rails. A voltage divider like Dave used ought to do just fine.

If you don't care about power consumption you don't need an IC at all - a voltage divider with a decoupling cap will do for many applications. The divider has to be rather stiff, though. The op amp, configured as a buffer, allowed him to use huge resistor values to conserve power.

the tl431 was in a search of virtual grounds/refs on digikey

[c4757p]

DigiKey combines the two into one category? Hmm.... It's just a voltage reference.

[ve7xen]

TI actually has a dedicated 1/2 Vcc virtual ground chip with the divider and buffer integrated: http://www.ti.com/product/tle2426

But usually I guess you'd just use a resistive divider and opamp.

[peterthenovice]

TI actually has a dedicated 1/2 Vcc virtual ground chip with the divider and buffer integrated: http://www.ti.com/product/tle2426

But usually I guess you'd just use a resistive divider and opamp.

and of course you need more than a soldiering iron to properly install such a device.

[c4757p]

and of course you need more than a soldiering iron to properly install such a device.

If you set it on the PCB and precisely, sharply apply force with a BFH, it should make contact, just like repairing a thermocouple. And who doesn't have a BFH?

Yield directly proportional to precision.

[peterthenovice]

and of course you need more than a soldiering iron to properly install such a device.

If you set it on the PCB and precisely, sharply apply force with a BFH, it should make contact, just like repairing a thermocouple. And who doesn't have a BFH?

Yield directly proportional to precision.

so you need  a battle field hero to do such a job

[c4757p]

Big fuckin' hammer.

[Rufus]

I was watching Dave's op-amp videos. It got me thinking about his ucurrent.  In his circuit it has a virtual ground.

The ucurrent doesn't have a virtual ground (I would prefer to call it virtual earth, having used that term for like 3 decades).

It has a stabilized mid supply rail which we can choose to call ground but there is nothing virtual about it.

The most common virtual earth in electronics is the -ve input of an inverting operational amplifier circuit. With the +ve input connected to true earth the negative feedback forces the -ve input to be at the same potential. The -ve input has the quality (voltage) of true earth without actually being true earth, the virtual description is appropriate.

[c4757p]

The "stabilized mid supply rail" in this circuit is what I've heard referred to as a virtual ground for quite some time. It's virtual because it's not meant to sink or source current.

Edit: OK, it sinks current from the voltage divider. I meant power supply currents.

[peterthenovice]

what effect does a virtual ground have on a split rail op-amp?

[c4757p]

Effect? None. But by moving the signals you're dealing with away from the rails, you avoid any input and output swing limits.

[Rufus]

The "stabilized mid supply rail" in this circuit is what I've heard referred to as a virtual ground for quite some time.

Probably because people have heard the expression without understanding what it is and used it to describe what it isn't.

The wikipedia article on it is a joke, demanding that a supply rail must be divided by exactly 2 to be called a virtual ground. Then carry on to mix up an active divider (at least they mention rail-splitter) with a real virtual earth at the -ve input of an inverting op-amp.

I found an LTC appnote incorrectly using the term, at least a tutorial from Maxim had the decency to say "creating an artificial virtual ground as a reference voltage".

Virtual earth has had a defined meaning since before the days all op-amps ran on +/-15v why would you start using the same term to describe something completely different?

[c4757p]

The wikipedia article on it is a joke

Agreed.

Virtual earth has had a defined meaning since before the days all op-amps ran on +/-15v why would you start using the same term to describe something completely different?

Because it's used that way now.

[Rufus]

Virtual earth has had a defined meaning since before the days all op-amps ran on +/-15v why would you start using the same term to describe something completely different?

Because it's used that way now.

Doesn't mean you shouldn't stop it

[peterthenovice]

well if its a joke then change it why do think that it is a publicly editable document

[c4757p]

Have you ever tried? Typos can be fixed. Basic facts in an article that took some doofus time to write? They'll fight you to the death over that! 

[peterthenovice]

what is your opinion on virtual grounds dave

[peterthenovice]

This is the first Wikipedia article revision

http://en.wikipedia.org/w/index.php?title=Virtual_ground&oldid=75117885

[David_AVD]

TI actually has a dedicated 1/2 Vcc virtual ground chip with the divider and buffer integrated: http://www.ti.com/product/tle2426

But usually I guess you'd just use a resistive divider and opamp.

and of course you need more than a soldiering iron to properly install such a device.

Are you referring to the package type?  They are in TO-92, DIP8 and SOIC8 packages too.  Even in stock at Digikey.

[free_electron]

Ground is where one plants potato's ...
Earth is a bog planet spinning around Sol.

There is no ground or earth.
What is being talked about is simply a reference point. It is arbitrary. Sometimes this reference point may be electrically connected to the ground potential, but this is not a necessity. Public Phone systems use a negative supply for example ( to combat the dissolving of copper wire through electrolysis... A positive copper wire will dissolve copper into moist earth should the isolation fail. A negative charged one would only grow in thickness by absorbing copper out of the earth. ( if present ))

Now. The reason for creating a 1/2 supply reference point is that you can do away with coupling capacitors for signals, and you can use the full input and output swing of, for example, an opamp.

This artificial reference plane can be connected to chassis or earth or ground.
Simply see it as an aide to make signal manipulation easier.

[GK]

Oh dear........

By the mechanism of negative feedback an op-amp always strives the best it can to maintain its two input terminals at the same potential. If the non-inverting input is tied to a reference potential called either "ground" or "earth", then the inverting input becomes what is referred to as either "virtual ground" or "virtual earth".

That's all.

[David_AVD]

By the mechanism of negative feedback an op-amp always strives the best it can to maintain its two input terminals at the same potential. If the non-inverting input is tied to a reference potential called either "ground" or "earth", then the inverting input becomes what is referred to as either "virtual ground" or "virtual earth".

Yes, I remember the "virtual earth point" term from 20 years ago, in audio mixers where the signals are summed at the non-inverting input of an opamp.

EDIT: Yes, I meant the inverting input !

[AlfBaz]

Seems to me that the virtual ground in the micro current is just that, a virtual ground. It may be drawn differently but we are taking off from the inverting feedback point. The only sticking point is we are referencing a mid point between the rails created by a resistive divider. We may be a little less fussy about it if we connected 2 batteries in series and referenced to the series connection and called it ground. Either way the unit is housed in a plastic box and powered by batteries so no matter how we create the so-called ground/earth its going to be floating with respect to terra firma

[GK]

Yes, I remember the "virtual earth point" term from 20 years ago, in audio mixers where the signals are summed at the non-inverting input of an opamp.

You mean the inverting input.

[MacAttak]

Based on the subsequent back and forth in this thread, I'm agreeing with the OP that this would make a good short topic for FF.