Confused about how to calculate effective power/voltage/current of a pulse wave

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

So you are saying that the oscilloscope waveform doesn't show the real module current consumption. Then we shouldn't discuss this waveform.

Once again, if a device is supplied with constant voltage, average power consumption isn't but average current multiply voltage.
Pavg = Vconst * Iavg
So the problem is about measuring average supply current.

I believe the only way to calculate the effective power consumption having such a waveform is to Fourier transform the waveform and choose n to some number of accuracy and find out the rms for the different part of the Fourier series and sum it .. this is the only way to get a correct effective power consumption.

Click to expand...

You are thinking too complicated.

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

why not discussing this waveform if you can tell me how much volt it is, the problem will be solved. use any one of them..
I said simulation, as I want to solve the problem using a known voltage, known waveform, and changeable parameters. this is why I used a function generator and tried to measure the voltage using these different devices.

Still all just write down what the equations are .. No real calculation is made... I know how to multiply the numbers .. But I want to know how much is the current .. You have the real waveform and the other one which is smiler, if you can tell me how much volt is the equivalent the problem will be solved.
Why each device read the voltage in a different way? even the oscilloscope tell us a different voltage.?

The question was simple but the answer is complicated. How much volts we get as an equivalent voltage so I can divide by the resistor to get the current ?

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

Mariwan,

The O'scope gives you a graphic picture of the wave. From that you can calculate anything. I already showed you how to calculate the RMS, and it agees with what the scope says it is. The calculations for the AVE do not. So the graphic calculations you make are the baseline for determining which instrument is correct. If the instruments do not give you correct readings, change the period of the pulse up to full period or DC if necessary, and see if they are correct at that setting. Otherwise get your instruments calibrated or repaired. We responders can only do so much from afar without physical contact.

Ratch

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

I thank you all in replying to this question.. I understand that it could be difficult to get my point of view on the matter.
I by myself did calculate everything, and I showed it to my teacher. The problem is that he doesn't like rms values.

I believe that, only rms values are correct.
I may be wrong .. but still I didn't get the exact answer for my question..

It is not about fault in the instruments, all this about a none clear view about the rms and average measurements. Multimeter producer - advertise that they have true-rms multimeter, and they say "no value measured by normal multimeter will be correct when you are measuring a non sin/cos waveform, you have to use true-rms multimeter". Knowing that almost all multimeter calculate average (if it is not written it has the ability to measure true rms) .. This makes me very uncertain about the results I get.

When the loads are digital components, it will not be easy to guess anything. Rise up of the current consumption in that way may cause battery discharging quicker than normal ..

Any way, I will appreciate any one who can find out the correct answer and put it here ..

I was unlucky to choose this project to my thesis .. XBee is equal to problems.

Again many thanks for all of you

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

The O'scope gives you a graphic picture of the wave. From that you can calculate anything. I already showed you how to calculate the RMS, and it agees with what the scope says it is.

Click to expand...

Yes, but the present problem doesn't ask for a RMS quantity. It's simply about average power consumption of a device drawing a pulsed current from a constant supply voltage.

Of course you can calculate a RMS value for the current. But it has no particular relevance and as far as I understand isn't actually asked.

I by myself did calculate everything, and I showed it to my teacher. The problem is that he doesn't like rms values.

Click to expand...

He is right for this problem.

When the loads are digital components, it will not be easy to guess anything. Rise up of the current consumption in that way may cause battery discharging quicker than normal ..

Click to expand...

The battery charge balance still has to be calculated from average current. A battery can have a slightly current dependent capacity, or may dislike pulsed currents in general. In this case you want to add a large filter capacitor, which is e.g. done in mobile phones to filter the pulsed GSM PA load. But calculation of average current and respective average power is valid anyway.

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

FvM,

Yes, but the present problem doesn't ask for a RMS quantity. It's simply about average power consumption of a device drawing a pulsed current from a constant supply voltage.

Click to expand...

Look at post #1 if this thread. The OP is asking how to calculate RMS and AVE of a voltage or current.

Of course you can calculate a RMS value for the current. But it has no particular relevance and as far as I understand isn't actually asked.

Click to expand...

The RMS is used to determine the power dissipation of the circuit. It is the correct way to calculate power dissipation. As I said before, using the AVE value fails and leads to incorrect results. No one has shown me a case where the AVE voltage or current calculates the average power correctly. The OP's teacher is wrong about using the AVE V or I. The relevance in this case was the inconsistency of results the OP was getting using different measuring instruments. I suggested he calculate the RMS and AVE manually from the O'scope trace and compare it with what his instruments give as a value. Then he can continue from there.

He is right for this problem.

Click to expand...

Prove it!

The battery charge balance still has to be calculated from average current. A battery can have a slightly current dependent capacity, or may dislike pulsed currents in general. In this case you want to add a large filter capacitor, which is e.g. done in mobile phones to filter the pulsed GSM PA load. But calculation of average current and respective average power is valid anyway.

Click to expand...

Batteries don't store charge, they store energy. Batteries are not charged, they are energized. But why is a battery involved in this discussion anyway? I don't agree that average current has any place calculating average power.

Ratch

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

XBee module uses Zigbee protocol. Everything is about "Low power - network modules" and the consumption of the power is a critical issue. The master thesis I have is about proving that .. Often batteries will be involved in powering the XBEE.
I really know how to calculate both mathematically, and my question was only which one is correct. and Why these devices show so different values, which value I should use if I use instrument.
I hope that I clarified the problem well.

As I said, if we convert the f(x) to a Fourier series equivalent .. No average power will be correct, this is why I am not so confidence with the average measurements.

We can NEVER use this equation I = V/R if the waveform is not sin/cos.. In such situation we have to use the Fourier series.. to use I=V/R..
And when you have a Fourier function .. average is not possible as the average will be equal to zero. So it should be rms equivalent.

Am I wrong??

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

Mariwan,

XBee module uses Zigbee protocol. Everything is about "Low power - network modules" and the consumption of the power is a critical issue. The master thesis I have is about proving that ..
I really konw how to calculate both mathematically, and my question was only which one is correct. and Why these devices show so different values, which value I should use if I use instrument.
I hope that I clarified the problem well.

Click to expand...

I have no idea what XBee or Zigbee is, but it don't matter. All I see is a pulsed waveform. It is easy to calculate the AVE and RMS of a pulsed waveform, and you should learn to do so.

We can NEVER use this equation I = V/R if the waveform is not sin/cos..

Click to expand...

Completely false statement. That equation if the basic definition of resistance, and is always correct for instantaneous current or voltage. If you need RMS or AVE values of current or voltage, then you can integrate the instantaneous values over a period.

Am I wrong??

Click to expand...

You are misguided.

Ratch

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

Look at post #1 if this thread. The OP is asking how to calculate RMS and AVE of a voltage or current.

Click to expand...

Yes, the post speaks about RMS quantities, but in my view problem hasn't to do with RMS.

It happens that I'm involved in the design of battery powered radio systems since several years. So believe it or not, but I know quite a little about the problem.

Batteries don't store charge, they store energy. Batteries are not charged, they are energized.

Click to expand...

I didn't discuss about battery storing charges or energy. Did anyone else? Nevertheless, battery capacity is specified in charge units (mAh, Ah) and battery charge is usually balanced to monitor battery state. Most likely your notebook computer has dedicated charge balancing ICs.

I don't agree that average current has any place calculating average power.

Click to expand...

I make a final attempt to clarify things. As general case, involving both time variant current and voltage, wie get

Pavg = average (Pinst) = 1/T*∫Pinst dt = 1/T*∫I*V dt (Pinst means instantaneous power)

Reviewing the previous posts, it seems necessary to mention

Pavg ≦ Irms*Vrms

If the analysed problem is however about constant voltage (which isn't different from a switched voltage being constant during I ≠ 0),
V can be pulled out of the integral

Pavg = 1/T*∫I*Vdt = Vconst*1/T*∫I dt

which isn't but

Pavg = Vconst * Iavg

You can see that average current can play a role in average power calculation. It's the case for constant voltage.

P.S.: The previous mentioned inequation should be changed into
|Pavg| ≦ Irms*Vrms
because Pavg is a signed quantity

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

OK,
Lets talk about Fourier series. It should be possible to get the same results using Fourier series representation of the waveform

The integral of a sin wave is equal to zero..
so if you use the Fourier series to measure the average, you get an average of ZERO.


Do you believe that it is wrong to use Fourier series? if your answer is "no it is ok to use it", so you can not calculate an average of a sin waveform .. it is equal to zero as the positive and the negative part is equal.
If your answer is YES, and fourier series is not possible to use, and it is wrong to calculate the effective voltage using the Fourier .. It means that I have been in the college one year studding garbage course .. incorrect topic



About terminology "charge of the battery": I really don't understand what the problem is with what I said. Didn't you hear about these words
Battery charger, Battery discharger, rechargeable battery, chargeable battery ..etc?


About the equation I=V/R .. I am not talking about instance current .. Instance means DC at that point .. so again it is DC you are talking about .. I am talking about timeless equation .. I=V/R is a timeless equation .. When time comes in the picture, it is not the same equation we should use..

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

Fvm,

I didn't discuss about battery storing charges or energy. Did anyone else?

Click to expand...

No, but you mentioned in post #25 of this thread, "The battery charge balance still has to be calculated from average current. A battery can have a slightly current dependent capacity, or may dislike pulsed currents in general."

I was just answering that statement.

With regard to that subject, if a battery is rated for 10 amp-hours, that does not mean that is has 36,000 coulombs stored within it. It means that in theory, it has the chemical energy to pump 36,000 coulombs around in some circuit. That is why I believe it is wrong to say a battery is charged, when if fact, it is energized.

If the analysed problem is however about constant voltage (which isn't different from a switched voltage being constant during I ≠ 0),
V can be pulled out of the integral

Pavg = 1/T*∫I*Vdt = Vconst*1/T*∫I dt

which isn't but

Pavg = Vconst * Iavg

You can see that average current can play a role in average power calculation. It's the case for constant voltage.

Click to expand...

I don't agree with your premise that a voltage pulsating at a constant amplitude is the same as a constant voltage. A constant voltage means constantly on. Therefore, is a mathematical mistake to factor V out of the integral.

To prove your formula, which only works for this particular waveform, Pave = V*I*k, where k is the fractional part of the period the pulse is on. Then Pave = V*Iave, where Iave = I*k. So no integral is involved. The general formula Pave=Vrms*Irms will work for any pulse at any DC offset.

Ratch

---------- Post added at 13:54 ---------- Previous post was at 13:33 ----------

Mariwan,

OK,
Lets talk about Fourier series. It should be possible to get the same results using Fourier series representation of the waveform

Click to expand...

Let me get this straight. Instead of dealing with one periodic pulse, you want to work with a fundamental and a bunch of harmonics? Unless you are doing filter design, I would not go there.

The integral of a sin wave is equal to zero..
so if you use the Fourier series to measure the average, you get an average of ZERO.

Click to expand...

The integral depends on what the limits of integration are. If the interval if from 0 to ½ the period, the value of the integral is not zero.

Do you believe that it is wrong to use Fourier series? if your answer is "no it is ok to use it", so you can not calculate an average of a sin waveform .. it is equal to zero as the positive and the negative part is equal.
If your answer is YES, and fourier series is not possible to use, and it is wrong to calculate the effective voltage using the Fourier .. It means that I have been in the college one year studding garbage course .. incorrect topic

Click to expand...

Fourier analysis is good, but not for this problem.

About terminology "charge of the battery": I really don't understand what the problem is with what I said. Didn't you hear about these words
Battery charger, Battery discharger, rechargeable battery, chargeable battery ..etc?

Click to expand...

I was not addressing you when I wrote about the charge of a battery. Review the past posts and see.

About the equation I=V/R .. I am not talking about instance current .. Instance means DC at that point .. so again it is DC you are talking about .. I am talking about timeless equation .. I=V/R is a timeless equation .. When time comes in the picture, it is not the same equation we should use..

Click to expand...

Instant means a value at that time, not DC. I=V/R means a constant I for as long as V holds. I(t)=V(t)/R now has time as a parameter.

Ratch

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

I fear, you are quite good in misunderstanding rather clear statements.

I don't agree with your premise that a voltage pulsating at a constant amplitude is the same as a constant voltage. A constant voltage means constantly on. Therefore, is a mathematical mistake to factor V out of the integral.

Click to expand...

It's not a premise. The calculation is for constant voltage. I made a side note about a special case of pulsed voltage. Please ignore it if you have difficulties to see a relation to the discussed case. Let's assume a constant voltage and time variable current for the time being.

The general formula Pave=Vrms*Irms will work for any pulse at any DC offset.

Click to expand...

Your so-called general formula is generally wrong. Please consider as simple case
constant voltage V=5 V, pulsed (square wave) current with Ipeak= 1A, duty cyle 0.25

We get:
Vconst = Vrms = 5 V
Iavg= 0.25 A, Irms = 0.5 A
Pavg = 1.25 W, Irms*Vrms = 2.5 W

To prove your formula, which only works for this particular waveform, Pave = V*I*k, where k is the fractional part of the period the pulse is on. Then Pave = V*Iave, where Iave = I*k. So no integral is involved.

Click to expand...

You are discussing the more special case of square wave current waveform. Vconst*1/T*∫I dt is valid for arbitrary current waveforms, not only I*k.

The question is however, if the Vconst assumption is right for the original problem. If not, you need to refer to the more general Pavg = 1/T*∫I*V dt.

Pavg = Irms*Vrms is wrong for both cases. It only applies to a few special cases: pure DC, sine voltage with resitive load and a few more.

Regarding fouriers series. They can be used to represent e.g. a periodical square wave current, with or without DC offset. So you get a spectral representation of the waveform. But what should be the purpose related to your original problem of average power calculation? I don't see any.

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

Would you please look at this video ???

https://www.youtube.com/watch?v=XUdRW0XgYQs

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

FvM,

I fear, you are quite good in misunderstanding rather clear statements.

Click to expand...

I think I have a good understanding of the basics.

It's not a premise. The calculation is for constant voltage. I made a side note about a special case of pulsed voltage. Please ignore it if you have difficulties to see a relation to the discussed case. Let's assume a constant voltage and time variable current for the time being.

Click to expand...

Your cannot pull out V from the integral unless it is a constant. If V is not on for the full period, then it is not constant. If you insist that a pulsating voltage is a constant, then that is your premise.

Your so-called general formula is generally wrong. Please consider as simple case
constant voltage V=5 V, pulsed (square wave) current with Ipeak= 1A, duty cyle 0.25

We get:
Vconst = Vrms = 5 V
Iavg= 0.25 A, Irms = 0.5 A
Pavg = 1.25 W, Irms*Vrms = 2.5 W

Click to expand...

The formula I gave is complete right, as long as the voltage and current are in phase.

Vrms =√(25/4) = 2.5
Irms = √(1/4) = 0.5
Pave = Vrms*Irms = 2.5 * 0.5 = 1.25

Your way: works for positive pulsating waveform
K = fractional part of period = 0.25
V = 5 ; I = 1
Pave = V*I*K = 5*1*0.25 = 1.25
Pave = V*Iave = 5*0.25 = 1.25
Pave = I*Vave = 1*1.25 = 1.25

You are discussing the more special case of square wave current waveform. Vconst*1/T*∫I dt is valid for arbitrary current waveforms, not only I*k.

Click to expand...

No, I don't believe that formula is valid for arbitrary waveforms for the reasons given above.

The question is however, if the Vconst assumption is right for the original problem. If not, you need to refer to the more general Pavg = 1/T*∫I*V dt.

Click to expand...

The assumption is not right for any problem. The integral formula is correct. You got the correct derivation for the wrong reason, and it only applies to positive pulses.

Pavg = Irms*Vrms is wrong for both cases. It only applies to a few special cases: pure DC, sine voltage with resitive load and a few more.

Click to expand...

No, it is correct for all cases.

Regarding fouriers series. They can be used to represent e.g. a periodical square wave current, with or without DC offset. So you get a spectral representation of the waveform. But what should be the purpose related to your original problem of average power calculation? I don't see any. .

Click to expand...

I assume you are not addressing me about this.

Ratch

---------- Post added at 15:49 ---------- Previous post was at 15:41 ----------

Mariwan,

Would you please look at this video ???

Click to expand...

OK, I did. He is preaching to the choir here. I agree with everything he says about true RMS, and I already knew everything he said.

Ratch

Re: Confiused about how to calculate effective power/voltage/current of a pulse wave

More confusions .. hahahah :shock:
Oh my *** what should we do ? :roll:

Re: Confiused abou how to calculate effective power/voltage/current of a pulse wave

FvM said:

Yes. But although you can apply RMS calculation (square root of averaged square value) to power quantities, it makes no technical sense.

Click to expand...

Yes RMS power is useless in most contexts, but not all. For example, the power of sound generated by thermal expansion of an electronic load is proportional to the RMS power (or more specifically the RMS of the AC component of the power)
through that load, not the average power in the load.

Re: Confiused about how to calculate effective power/voltage/current of a pulse wave

The formula I gave is complete right, as long as the voltage and current are in phase.

Click to expand...

In other words, only for pure resistive loads.

If you review the thread, you'll notice that the term "phase" is appearing the first time in your latest post. Mariwan didn't say anything about current and voltage in phase. He characterized the measurement problem this way:

VCC=3.3 this is a DC voltage goes to the XBee module
but the current is as the waveform shown in one of the figures

Click to expand...

I agree, that the oscilloscope waveform suggests a pulsed voltage with resistive load and a respective same shape current waveform. Only in this special case, Vavg equals Irms*Vrms. As soon as current and voltage are not in phase, e.g. because the load contains a capacitive component or is generally time variant, the power calculation fails.

If you also consider the possibility of a constant supply voltage and time variable load, you'll understand the purpose of Pavg = Vconst * Iavg.

As a convenient point, the measurement of average current is rather easy.

P.S.:

Yes RMS power is useless in most contexts, but not all. For example, the power of sound generated by thermal expansion of an electronic load is proportional to the RMS power (or more specifically the RMS of the AC component of the power)
through that load, not the average power in the load.

Click to expand...

My doubt is about applying the RMS specific calculation to power quantities, squaring instantaneous power values. I don't see a technical relevance for it, because power already represents squared magnitude (current, voltage, sound pressure, whatsoever) values.

Re: Confiused about how to calculate effective power/voltage/current of a pulse wave

If you review the thread, you'll notice that the term "phase" is appearing the first time in your latest post.

Click to expand...

Actually, I mentioned it in post #3 in the last sentence, where I was referring to I and V being sinusoidal.

Mariwan didn't say anything about current and voltage in phase.

Click to expand...

No, but we all assumed it was.

I agree, that the oscilloscope waveform suggests a pulsed voltage with resistive load and a respective same shape current waveform. Only in this special case, Vavg equals Irms*Vrms. As soon as current and voltage are not in phase, e.g. because the load contains a capacitive component or is generally time variant, the power calculation fails.

Click to expand...

As stated in post #3, Pave = Irms*Vrms*cos(θ) if the waveforms are sinuoidal. If the waveforms are arbitrary, then Pave = (1/T)*∫I(t)*V(t)*dt from 0 to T. That formula will work no matter what the phase is or what the components are.

If you also consider the possibility of a constant supply voltage and time variable load, you'll understand the purpose of Pavg = Vconst * Iavg.

Click to expand...

I don't know, I think I will stick to Pave = (1/T)*∫I(t)*V(t)*dt from 0 to T. I know that works.

As a convenient point, the measurement of average current is rather easy.

Click to expand...

It all depends on how complicated the current function is.

If you are answering someone else in the same post, how about inserting the persons name so we know who you are answering.

Ratch

Ratch, I admit that I needed some time to understand the nature of misunderstandings confusing the discussion.

It's true, that you mentioned sine waveforms and cos(phi) in an early post. But I don't think that it helps much for the present problem.

According to the problem description given by Mariwan, I would suggest the below simplified model. You may want to place an arbitrary current waveform instead of the switched load, or you can discuss a pure resistive load as special case.



We already agreed about Pavg = 1/T*∫I*V dt for the general case.

But isn't it obvious that the integral expression reduces to Pavg = V * Iavg for constant V?

Generally, this case is not very special. Actually many wireless devices acts like this or even other devices running on batteries, and it is to save battery time. Yes it is complicated, the problem is not how to calculate the power consumed, the problem is which one is the correct one and correct way in calculating.
In the other hand, which one of the instruments or configuration of the instruments measure correctly?
may be none of them?
I sent the case even to another teacher .. I hope I get a response some time later