Accuracy of 60Hhz power measurement device on 50Hz AC

Jun 10, 2006 31 Replies

--- Well, if you want to determine whether the PF meter is accurate or not you'll have to do _something_ , so let's start at the beginning, OK?

Power factor is defined as the cosine of the phase angle between current and voltage, so if we can make a circuit with a predictable phase angle we can then determine what its power factor will be and, if we plug it into the meter, and that meter doesn't affect our circuit, the PF meter should read what we predicted.

So, What circuit? A simple resistor and capacitor wired in series and plugged into the PF meter's output receptacle would be perfect.

Let's say that we choose a power factor of 0.5 just because it's halfway between 0 and 1, 0 corresponding to a perfecty reactive load and 1 corresponding to a perfectly resistive load.

Now we've gotta do some numbers. Sorry...

Since:

PF = cos(phi)

we get the angle like this:

phi = cos-1(PF) = cos-1(0.5) = 60°

Next, since we don't want to blow anything up we'd like to keep the power dissipation in our test circuit down to something reasonable.

Unfortunately, we don't know what the accuracy of the meter is when it's presented with low power loads but, just for grins, let's go for a 10 watt load.

Then, since:

E² P = --- R

we can rearrange that to get the value of a resistor that'll dissipate 10 watts when it's connected across 240V mains like this:

E² 240V² R = --- = ------ = 5760 ohms P 10W

Not bad. Keeping in mind that adjustable resistors are cheap and adjustable capacitors aren't, lets double that resistance and wattage to 10k ohms and 20 watts, so that the resistor's slider will be somewhere near the middle of the resistor (hopefully) when we figure out the capacitor we need to give us that 60° phase shift.

Now, since:

Xc tan(phi) = ---- R

and we know that phi = 60°, in order to get Xc (which will eventually get us to C) we rearrange again to solve for Xc like this:

Xc = R tan(phi) = 5760R * 1.732 ~ 9976 ohms

And then, finally, since:

1 Xc = --------- 2pi f C

we can rearrange to solve for C like this:

1 1 C = ---------- = --------------------- = 3.19E-7F 2pi f Xc 6.28 * 50Hz * 9976R

That's 0.319µF, which isn't readily available, but 0.33µF is, and that's pretty close, (since we have an adjustable resistor to play with) but what about the cap's voltage rating?

For 240VRMS mains we're talking about 340V peak which will be placed across the series RC, and the voltage will divide between the R and the C depending on the impedance of the circuit, which will be:

Z = sqrt (R² + Xc²) = sqrt (5760² + 9976²) = 11519 ohms.

That means the current in the circuit will be

E 240V I = --- = -------- ~ 0.0208 amperes. Z 11519R

and that current in the reactance of the capacitor will drop:

E = I Xc = 0.0208 A * 9976R ~ 208VRMS

across the capacitor, so a 0.33µF 350V cap would do.

Now, since the cap is 0.33µF instead of 0.319µF, we've got to go back and find what resistance we need to get a phase angle of 60°.

First though, we need to find the actual value of the cap, so you can either measure it if you have a capacitance meter or use the

50Hz mains and a 1% resistor in series with the cap to find out. Use a transformer to step the mains voltage down to something you can work with comfortably if you're scared of the mains.

Now, once you've determined the capacitance of the capacitor, (let's say it's exactly 0.33µF) you determine its reactance by solving:

1 1 Xc = ---------- = ----------------------- ~ 9651 ohms, 2pi f C 6.28 * 50Hz * 3.3E-7F

then, knowing you want a 60° phase angle, rearrange:

Xc tan(phi) = ---- R

to solve for R, like this:

Xc 9651R R = ---------- = ------- = 5572 ohms tan(60°) 1.732

And here's your circuit:

+------+ 240V50Hz>---| |------+ | | | | PF | [5572R] |METER | | | | [0.330µF] | | | 240V50Hz>---| |------+ +------+

Just to make sure we're not going to exceed the resistor's wattage rating we need to go back and figure out the current in the circuit.

Since its impedance will be:

Z = sqrt(R² + Xc²) = sqrt(5572² + 9651²) = 11144 ohms

the current through the resistor will be:

E 240V I = --- = -------- ~ 0.022A Z 11144R

and it'll be dissipating:

P = I²R = 0.022² * 5572R ~ 2.7W,

so a 10000 ohm 10 watt adjustable resistor would be fine.

Notice, however, that the original assumption of a 10 watt load won't be true because of the increase in the circuit's impedance due to the capacitive reactance. But, now that you know how to choose the R and C, you ought to be able to figure out how to get whatever load impedance you need.

-- John Fields Professional Circuit Designer

The reason I'm sticking with resistive load is largely because I think it's easier to get resistors :P

Although I'm not all too sure yet what happens if you connect 230 AC to a normal resistor meant for DC? Or does it not matter because the resistor will waste the same amount of power as in a norminal 230DC circuit because first it burns off +V * +A = +W in half an AC cycle, then -V * -A = +W in the second half?

The same goes for a transformer with nothing connected, it's a purely inductive load right? But if it's connected to nothing, then nothing will happen no? Wouldn't it become merely a coil of wire acting as a (very low) resistive load?

A Lost Angel, fallen from heaven Lost in dreams, Lost in aspirations, Lost to the world, Lost to myself

But you can\'t determine whether the power factor meter is working properly for a reactive load unless put some reactance in there!

How do I tell what voltage is the resistor rated for? I've never actually noticed any of them having a voltage rating. Shouldn't it be safe say if I use a 12K Ohm 10W resistor since that with the peak of

340V would still be under the 10W rating. Or am I missing something totally in between?

I was still thinking of it in purely DC terms when I wrote that. After doing more reading later, I then understood (perhaps wrongly) that the alternating voltage causes EMF to be generated in the iron core which then acts like resistance to prevent current from flowing. So there wouldn't be a short circuit even though the coil is directly connected from AC live to AC neutral.

However, it's still confusing because I read that I shouldn't be putting DC at the input nor have DC at the output in most cases. Does it mean putting a DC component like a resistor at the secondary winding end is something I should approach with great caution and not as simple as normal DC circuits?

A Lost Angel, fallen from heaven Lost in dreams, Lost in aspirations, Lost to the world, Lost to myself

Firstly, thank you very much for the detailed post!

It's going to take me quite a while to digest those formulas (ok, the sight of them simply gives me brain freeze). So I'm just not going to think about how the various forms are derived and just plug the exact formula and numbers into a spreadsheet.

Would this step be necessary if the objective is simply to determine the general accuracy of the meter? Since the 0.33uF cap would tend to be within 5% of its stated value no? Yes, I'm still kinda worried about plugging anything to the mains, esp since I read stories about them caps exploding real easy.

I don't think we have 10W pots here. I think 5W or 2.5W are the only ones I've seen the last time I looked. So reworking for a max of 2.5W load.

Initial target resistance is 23040 Ohms Calculated XC is 39906.45 C is 7.98E-8, effectively looking at a 0.08uF 350V capacitor right? But since 8 is not a standard number, I can use either 0.068uF or

0.1uF.

Going with 0.1uF since it sounds like a more readily available item and assuming the measured uF is 0.1 The actual XC would be 31830.99 R needed would be 18377.63 Z = 36755.26 I = 0.01A which means 0.78W, easily handled by a 50K 2.5W VR correct?

In the event I cannot obtain a 50K VR, would it be the same if I put a

12K fixed resistor in series with a 10K VR?

Once again, thanks for your patience and help!

A Lost Angel, fallen from heaven Lost in dreams, Lost in aspirations, Lost to the world, Lost to myself

It depends on the manufacturer\'s spec\'s. For instance, Panasonic\'s thru-hole 1/4 watt 5% carbon film resistors go up to 2.2 megohms, so that would imply that for that wattage and resistance 742 volts could be applied to the resistor, yet their working voltage limit is 250V.

Well, the last time I suggested this, John had an episode, but the "official" way to "determine the general accuracy of the meter" is to find someone who has one that can measure the same stuff, and compare them. You can do a lot of experiments, yes - do you have an LCR meter? But the sight of you at some lashup with 240V here and there and inductors and capacitors just lying there withe a bunch of loose clipleads just gives me the willies.

Sorry for being pedantic, but not getting zapped is a good thing, regardless.

Good Luck! Rich

The last time you suggested what?

yeah, just use a toaster, a 100W lightbulb, a cheap soldering iron. these are resistive devices and have the advantage of being readily availablem and safe to connect to the mains.

it warms up.... - nothing special happens.

no... ir becomes a (nearly) pure inductive load.

electric energy will flow into the inductor (where is it convertted to magnetism) and outof the inductor (converted from the magnetism) as the AC cycle progresses,

with a resistor electric energy only flows in and heat comes out.

Bye. Jasen

Yes, of course. If you use 10% "standards" to test a 1% device then you\'re only getting a very gross indication of whether the device is working properly or not.

well, if it gives you the willies, think about what I feel when I picture pretty much the same thing myself.

But, at some point in the past, I thought drawing up a circuit of any sort and having it work would be crazy too. Never would expect that I could make a working (more or less) replacement adapter for my LCD until I did it. So while I'm trying to beef up my knowledge of exactly what's involved, I'm also prepping myself mentally to go over another hurdle :P

I'm the sort who can't really learn something without actually doing it.

Yup, tell me about that, esp when I'm the one who might get zapped. Though I would probably be doing this with rubber gloves and rubber soled shoes :P

A Lost Angel, fallen from heaven Lost in dreams, Lost in aspirations, Lost to the world, Lost to myself

sufficient tests can probably be done by plugging a reular 4 way power strip into the metering device and then using a combination of ordinary appliances with known characteristics as loads. theres's no need for a live lashup.

Bye. Jasen

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required