Why can I measure AC on a DC power supply?

[vk6zgo]

"Alternating Current" does NOT imply that is is alternating around the zero reference.  That may be the most common kind of AC waveform you might encounter, but the definition does not require BIPOLAR in any way.

https://en.wikipedia.org/wiki/Alternating_current says the definition of AC is "...the flow of electric charge periodically reverses direction"

My understanding of AC is the green line below.
 

where the electric charge in the conductor moves back and forth at some predefined frequency (E.g. 60Hz for wall plugs here in Canada)

So in my humble opinion I think:

AC = alternating current (electrons physically move back and forth in a conductor)

The "AC" setting on a multimeter simply reads a change in voltage. It does not care if the voltage is going positive or negative.

When the oscilloscope set to AC coupling it simply inserts a capacitor in series. The capacitor, when charged, prevents any more steady DC current from flowing. When the voltage does change the capacitor will discharge or charge allowing some current to flow again. The result shows up as "AC" on the oscilloscope. Just like the multimeter, "AC" simply shows a change in voltage.

If what I'm saying is correct this explains why my multimeter was showing AC connected to my DC power supply. It was simply reading the change in DC voltage. I thought it meant the voltage was reversing polarity. Silly me, I thought the A in AC stood for alternating.

Wiki looks very authoritative,but they sometimes get things wrong,& people can,& do, amend it..

A case in point was a "Ham Radio Wiki",where they showed a Bridge rectifier & labelled it a "Kratz" rectifier.
I had never heard of this name,& a search both online & in Electronics books showed no sign of him,so I amended it to read "Bridge" rectifier.

It turns out that the name is really "Graetz"-----a well known name in European Electronics history.
That said,I had never heard such a rectifier called by this name,either.

Anyway,back to topic:-

"Pulsating dc" is a fairly archaic term,which was often used to describe the output of dc generators.
(All generators are really alternators,but in the dc generator,the commutator acts as a mechanical rectifier)
The output waveform of a dc generator does return to zero at the points where the (internal) ac waveform goes thru zero prior to commencing the next cycle.

Another case where this term was used was in unfiltered rectifiers.
When the rectifier ceases to conduct,the output waveform returns to zero.

A dc voltmeter will see these "positive  going" or "negative going" half cycles as a dc voltage.
(An ac voltmeter will also give a reading as it sees them as an  ac voltage).

Early EEs did not have Oscilloscopes,but they knew how dc generators worked,so the term pulsating dc was a reasonable term.

When amplifying devices appeared the "cat got among the pigeons"!

There are very few such devices which do not work by converting the energy of a dc power supply
into amplified ac.
Their outputs looked like an ac signal riding upon a dc voltage.
 
(Users normally only needed the amplified ac,so had to separate it from the dc.
This usually entailed coupling capacitors,interstage coupling transformers ,etc.)

"Pulsating dc" was no longer a valid term to describe the combined signal.

"it looks like a duck,quacks like a duck,,,,,,," so the description of such a waveform  became that of "dc with an ac component".

In the meantime,more theoretically inclined EEs & others,devised the concept of  electrical signals being made up of a "fundamental" ac signal,plus a number of harmonics.
Using this concept,it is possible to analyse any waveform.


PS:- the previous poster has "hit the nail on the head"!
If you add 2volts (peak) ac to a +12volt dc supply,it will change between +!4 volts  (for the peak of the positive half cycle),through +12volts at the zero crossing between 1/2 cycles,& +10volts  (for the peak of the negative half cycle).

You can see (for this example),that the ac component passes through its positive peak where it adds to the dc voltage,passes through zero,where the dc voltage is that of the +12volt source alone,& then subtracts from the dc voltage at the negative peak.

With batteries,you can demonstrated that the resultant voltage of sources connected in series with reverse polarity subtract,& in the same polarity,add.

All the sources have an independent existence,regardless of their series connection,as does the ac component of a signal.

[Rick Law]

...
Why can I measure AC on a DC power supply? To show you want I mean I’ve setup the following test:

I made my DC power supply switch between 2 and 3 volts as fast as it can. It turns out it can do this around 4.5 times a second creating a square(ish) wave.  My first multimeter measures DC, the second measures AC and my oscilloscope shows the waveform.
...
...
Mark

[ bold added in quote]

Since no one yet brought this up...

It is likely not an an issue of AC verse DC.  You are changing it only 4.5 times a second, by changing, I take it you mean voltage change up or down.  So a full cycle takes two changes (high-to-low then low-to-high).  Two change cycles occurs only 2.25 times a second.  That is a very very slow wave at 2.25Hz.  Even if I misunderstood and your "changing it" mean a complete hi-to-low plus low to high, that is still a slow 4.5Hz, your DDM's  sampling window is very significant when it is that slow.

At 4.5Hz, that is a good 222/1000 seconds per cycle (ie: 222ms).  It is likely out of working range for your DMM's TRMS/RMS to work so AC reading is meaningless.   (My UT61E's AC measurement is 41Hz min to about 1KHz, 2.25hz does not even come close.)

With DC reading, 222ms is a long time.  Averaging just 10 full cycles will take 2.22 seconds and most DMM report at least once a second.  If say your DMM is updating three times a second, it would have just 1.5 cycles to average.  Hard to average out to a good number with just one full cycle and a fraction of the next.

So, you are not within the working range of your tool (DMM).  A scope will give you far better measurement.

[eilize]

you must absolutly see the difference between a periodic signal and an alternating signal/

as i said, average of an alternate signal =0
if you add a dc to an alternate signal, you have (always) a periodic signal (but not anymore alternating)where the average is the dc value

so :
when you switch to ac on your scope, you see only the alternate signal
back on dc position, you see the periodic signal (ac+dc)

on your fluke, your read the rms value of the alternating signal(it's put on alternating signal measurement)

[Mark Loukko]

Thanks for the history lesson vk6zgo :-)

"Pulsating dc" is a fairly archaic term,which was often used to describe the output of dc generators.
(All generators are really alternators,but in the dc generator,the commutator acts as a mechanical rectifier)
The output waveform of a dc generator does return to zero at the points where the (internal) ac waveform goes thru zero prior to commencing the next cycle.

For those who like visual aids (like me) this is what the output voltage of a DC generator looks like:

In the meantime,more theoretically inclined EEs & others,devised the concept of  electrical signals being made up of a "fundamental" ac signal,plus a number of harmonics.
Using this concept,it is possible to analyse any waveform.

I will look this up. More bedtime reading!

If you add 2volts (peak) ac to a +12volt dc supply,it will change between +!4 volts  (for the peak of the positive half cycle),through +12volts at the zero crossing between 1/2 cycles,& +10volts  (for the peak of the negative half cycle).

I graphed this in Excel. Please see the attachment. Did I get it right?

With batteries,you can demonstrated that the resultant voltage of sources connected in series with reverse polarity subtract,& in the same polarity,add.

All the sources have an independent existence,regardless of their series connection,as does the ac component of a signal

I'm will have to process this for a bit :-)
 

[Zero999]

Thanks for the history lesson vk6zgo :-)

"Pulsating dc" is a fairly archaic term,which was often used to describe the output of dc generators.
(All generators are really alternators,but in the dc generator,the commutator acts as a mechanical rectifier)
The output waveform of a dc generator does return to zero at the points where the (internal) ac waveform goes thru zero prior to commencing the next cycle.

For those who like visual aids (like me) this is what the output voltage of a DC generator looks like:

Actually that's not true. As long as the shaft is turning and it's working properly, a DC generator's output voltage never falls below to zero.

A DC generator will have more than two poles on its armature, normally an odd number. Suppose it has three poles on its armature, the output will look like a rectified three phase rectifier which never goes to zero.
#
EDIT:
The output of a DC generator will look something like this:


Except it will be much more noisy.

[Mark Loukko]

Since no one yet brought this up...

It is likely not an an issue of AC verse DC.  You are changing it only 4.5 times a second, by changing, I take it you mean voltage change up or down.  So a full cycle takes two changes (high-to-low then low-to-high).  Two change cycles occurs only 2.25 times a second.  That is a very very slow wave at 2.25Hz.  Even if I misunderstood and your "changing it" mean a complete hi-to-low plus low to high, that is still a slow 4.5Hz, your DDM's  sampling window is very significant when it is that slow.

At 4.5Hz, that is a good 222/1000 seconds per cycle (ie: 222ms).  It is likely out of working range for your DMM's TRMS/RMS to work so AC reading is meaningless.   (My UT61E's AC measurement is 41Hz min to about 1KHz, 2.25hz does not even come close.)

With DC reading, 222ms is a long time.  Averaging just 10 full cycles will take 2.22 seconds and most DMM report at least once a second.  If say your DMM is updating three times a second, it would have just 1.5 cycles to average.  Hard to average out to a good number with just one full cycle and a fraction of the next.

So, you are not within the working range of your tool (DMM).  A scope will give you far better measurement.

Good point about the frequency. I have the Fluke 87-5.  I just checked the manual and it looks like the min frequency is 45Hz and max is 20kHz for reading AC.
I knew the frequency was super super slow but that's as fast as my DC power supply could change the voltage up and down.
At the time I was surprised to see anything showing on AC while connected to my DC power supply.

The same signal was connected to my oscilloscope which had no problems reading the slow frequency.

[Mark Loukko]

Just an FYI...

For those interested, I found another thread on this forum asking the same sort of AC / DC question:
https://www.eevblog.com/forum/beginners/what-really-is-ac-and-dc/

On another forum I found the following... It was quite entertaining to read. After reading the second quote I believe I'm a "purist". 

http://boards.straightdope.com/sdmb/archive/index.php/t-494980.html

If you have a wire (or any circuit component, for that matter) and the current never changes direction, then by definition it is DC. The current can go up, down, even to zero. But as long as it never changes direction (i.e. goes negative), then it is by definition DC. Obviously, current flows in both directions in an AC system.

As a side note, you know the 120 VAC receptacle on your wall? It is more correct to call it an alternating voltage (AV) receptacle than an AC receptacle, since the former is always true, and the latter is usually - but not always - true.

Having said that, EEs are sloppy when it comes to the terms AC and DC. When we talk about an audio signal that is represented by a voltage, for example, we will say it's an "AC" signal, when in reality it's an AV signal. We also like to break signals up into their AC and DC components. Take a class-A transistor amplifier, for example. Strictly speaking, the overall current through the emitter is DC at all times, since its direction never changes. EEs will say, "There is a DC current through the transistor for bias purposes, and a small AC current riding on the DC that represents audio."

No, that is not how AC is defined. if it has a DC component then you could say it is AC with a DC component but it is not pure AC.
I understand what you're saying. Perhaps there are two definitions of AC, then:

1. AC is what's left over after you subtract the DC component from the overall signal.

2. When looking at the overall signal, if the current ever reverses direction (i.e. goes negative), then it is an AC signal.

You are using #1. I am using #2.

Look at the top waveform in this pic (http://zone.ni.com/cms/images/devzone/ph/3a74325328.gif). Let's assume these represent current and not voltage. Most EEs would say, "It is an AC signal with a DC component." (Definition #1.) But a purist might argue that the top waveform is really a DC signal, since the current never reverses direction. (Definition #2.)

So which definition is correct? I dunno. I would normally say "it is an AC signal with a DC component." But at the same time I would have to acknowledge that, in the very strictest sense, it is a DC signal, since it never reverses direction.

 

[vk6zgo]

Just an FYI...

For those interested, I found another thread on this forum asking the same sort of AC / DC question:
https://www.eevblog.com/forum/beginners/what-really-is-ac-and-dc/

On another forum I found the following... It was quite entertaining to read. After reading the second quote I believe I'm a "purist". 

http://boards.straightdope.com/sdmb/archive/index.php/t-494980.html

If you have a wire (or any circuit component, for that matter) and the current never changes direction, then by definition it is DC. The current can go up, down, even to zero. But as long as it never changes direction (i.e. goes negative), then it is by definition DC. Obviously, current flows in both directions in an AC system.

As a side note, you know the 120 VAC receptacle on your wall? It is more correct to call it an alternating voltage (AV) receptacle than an AC receptacle, since the former is always true, and the latter is usually - but not always - true.

Having said that, EEs are sloppy when it comes to the terms AC and DC. When we talk about an audio signal that is represented by a voltage, for example, we will say it's an "AC" signal, when in reality it's an AV signal. We also like to break signals up into their AC and DC components. Take a class-A transistor amplifier, for example. Strictly speaking, the overall current through the emitter is DC at all times, since its direction never changes. EEs will say, "There is a DC current through the transistor for bias purposes, and a small AC current riding on the DC that represents audio."

No, that is not how AC is defined. if it has a DC component then you could say it is AC with a DC component but it is not pure AC.
I understand what you're saying. Perhaps there are two definitions of AC, then:

1. AC is what's left over after you subtract the DC component from the overall signal.

2. When looking at the overall signal, if the current ever reverses direction (i.e. goes negative), then it is an AC signal.

You are using #1. I am using #2.

Look at the top waveform in this pic (http://zone.ni.com/cms/images/devzone/ph/3a74325328.gif). Let's assume these represent current and not voltage. Most EEs would say, "It is an AC signal with a DC component." (Definition #1.) But a purist might argue that the top waveform is really a DC signal, since the current never reverses direction. (Definition #2.)

So which definition is correct? I dunno. I would normally say "it is an AC signal with a DC component." But at the same time I would have to acknowledge that, in the very strictest sense, it is a DC signal, since it never reverses direction.

The terms "ac voltage",& "dc voltage" may be read,for the former case,as:-

"A voltage,which if connected to an external circuit would cause an alternating current to flow",

& in the latter case:-

"A voltage,which if connected to an external circuit would cause an direct current to flow".  


As in your quote,EEs are sloppy---but not only EEs,the general public do the same all the time.

An "American" (actually a citizen of the United States of America) may put "Gas" (Gasoline) in his "Auto"(Automobile).

From context,we know that he isn't literally putting "GAS" into it.

An Englishman may put the same stuff in his "Motor"(Automobile).

From context.we know he isn't pouring fuel into the oil filler cap of his engine,or into the guts of an Electric Motor.

Getting back to Electronics,even something that a "purist" may use to describe a current,say :-

"A dc current" is like this,as it really means literally "a direct current current"


Such "sloppy" terms,(which have been round for a century or so),& the concept of an "ac component" are useful,as they make it possible to describe circuit operation which it would be horribly unwieldy to describe if we used strict interpretations of definitions.

Another way of addressing things is to think of dc as ac of 0Hz frequency.

It we had the time,& enough capability,we could synthesize a signal which is the same in all respects as your
original on/off PSU switching,by the combination of purely ac signals.