Phase angle controller Question

Also, is it possible to have phase angle shift in an ac circut without capacitance or inductance?

Phase angle firing is a term that usually refers to controlling the point in a cycle when a thyristor is turned on. If firing is very early, near 0 degrees, then the thyristor conducts for most of the half-cycle and current can flow to the load for most of the cycle. If firing is delayed to nearly 180 degrees, then current only flows to the load during the very late portion of the half-cycle.

In a resistive only load, the current is in phase with the voltage when the thyristor is conducting, but effectively shut off during the first part of the half-cycle before the thyristor is fired.

daestrom

Phase angle between what and what? If you mean between current and voltage, then in a purely resistive network the answer is no.

daestrom

Oh ok. that's what I am looking for. I think you answered my question. I am trying to settle a dispute. I have Phase angle fired controllers on a purely resisitve load only and they are Porprtional Zero cross triacs and they have 100% power factor acording to specs. That tells me there is no shift. I am being told my power meaurment is off becuse he thinnks the controller is changing the phase angle of current to voltage. I don't see how anything else could cause this. there are no caps or inductors in the circut. When I told him this he says hes not talking about Inductance and phase angle relationship between current voltage. So I have no clue what he means. To me that is the only thing that could cause a power reading error.

The description of what you are using isn't clear enough to me to identify what it is. On one hand, you refer to phase angle firing which is like a standard light dimmer. On the other hand you make reference to Zero cross which might be referring to zero crossing switching, which is a different technique for heating loads, where on/off switching is always done at the zero crossing point (no phase control) and the proportional control is done by varying the ratio of on half-cycles to off half-cycles. You didn't say what the resistive load was either, so I can't guess from that.

I have put together an example of a phase control dimmer waveform where you can change the level and see what happens...

it's the second example down the page, "Phase Control (Light Dimmer)".

For a phase control dimmer, there is scope to easily mis-measure the power consumed, depending on the type/quality of the power metering device used. Your main electricity meter will get it right, but some of the devices you may have bought to measure power consumption might not. If you play with the Phase Control applet on the web page above, you will see that the power factor drops below

1 when you start dimming, and the failure to take this into account with some measuring devices will lead to an over-estimation of the power consumed as you start dimming the light.

Thanks for the help.

Here is what I am using.

They say power factor is 100%.

There is indeed a mis-reporting of power at low levels but for a different reason I think. I have an uneven heating load using these controllers on both sides of a 240 volt genset to the N phase The current sensors are taking average current, not peak. If they were peak sensors that would indeed show a gross over reporting at the load decreases beacuse the current is chopped.

Here is where things take a bizzare turn. From 30% power on up the readings appear correct usning 1.111 X average correction.

At about 25% power on down going lower let's say at around 10% Things get ugly, the atcual heating load using true rms voltage measurements at the load and known resistance it appears to be 250 watts when controls are at this low setting. I belive that. I should be using RMS sensors I know this but the error using average sensors is only 1.111 X avaerage.

. Well guess what happens when I hook up true RMS current meters? It reads Higher! It says there is 4 amps on L1 and so does my average sensor using

1.111 correction. The current on L2 reads 6.5 but the average sensors says 3.4! SO the RMS reading is double on this side! In other words it agrees with one but not the other.

. This tells me that we have a total of 10.5 amps in the system. Well that's about 1250 watts. There is no way that load is on the engine. If you bypass the controllers with a known 1200 watt load you can clearly here the engine groan. So where is the other 1050 watts of power going or is not really there? I think it's either there and being stored in one side of the stator or it's apparent or reactive power becuase there is a phase imbalance.

Do you see what I see? Here is what I think is going on. As the voltage is off longer there is more energy stored and there is an imbalance of load. It appears the stored enegery is being pushed back to the side with lower load on it and the system is now has inductance or reactive power in it. I think my rms current measurments prove it. It is a resisitive load but the generator is bascilly an electric motor and this is a closed system. Once you start turning off the voltage longer things go haywire.

Keep in mind that we want the actuaual power to the shaft. If the power factor was alway's 100 then not an issue. I am no expert but it sure looks to me like we do have a big power factor problem. I think balancing the load would help. I guess my question is, just WTF is going on?

I think the best way to deal with this is to correct the power factor if that is the problem, not try to measure it.

Thanks.

Well this is 'mostly' right. When the triac is conducting, the current and voltage are in phase. If you go by the older idea of power factor, (pf=cos(theta)), then with 0 degree phase shift you would conclude that pf=100%

When you're firing the triacs very early (nearly 0 degree firing delay), this all works out pretty much okay.

But power factor is *really* defines as pf=W/VA. And measuring the current when you use this type of controller is a bit tricky. For example, when you fire the triacs at 90 degrees, there are some harmonics to the fundamental frequency current so a typical ammeter won't give accurate results.

On top of this, is the issue of actually measuring the current. A typical D'Arsonval movement ammeter (an 'analog meter') responds to

*average* current, not RMS. The meter face is re-calibrated by a factor of 1.1 so that it reads out RMS if the current happens to be sine wave. So determining the current is not so easy.

A few meters are 'true RMS' and will read a different value than the typical D'Arsonval meter when used on a waveform like the current that flows in this case (90 degrees firing delay).

He's wrong here. The triacs are not changing the phase relationship between V&I.

The current measurement can have errors in it as I mentioned above, so that can introduce error in power measurement.

If you had a meter that digitally sampled the current very fast and gave you the 'true RMS' reading, you won't have any error.

Or a power meter that digitally samples the instantaneous voltage and multiplies by the instantaneous current, that will give an accurate reading as well.

daestrom

Thought I would follow up. I installed a balanced load and it's like magic. Everything works fine even at late firing of the controllers. All measurnets are correct. One side still wants to drift a bit but I think the simple answer is a load balance capacitor. Any thoughts? I think this kinds proves my case that the controllers are making the system reactive.

Get a good power electronics book. You will find that when using a phase angle fired controller, at any power level other than 100% on (0 degrees) you get BOTH a reduction is true power factor (W / (Vrms*Irms)) and displacement power factor (angle between voltage and current). Yes, Yes, the current is in phase with the voltage when during the portion of the half cycle after the firing BUT the fact that you are not getting a full half cycle of current causes a phase shift between the voltage and current at the fundamental frequency. This is explained in many power electronics books. When you do an FFT on the resulting waveforms (say at 50% power) you will see that the fundamental component of the current is shifted (lagging).

What this means is that if you use this type of controller on a purely resistive load, at other than 100% power, your displacement power factor will be less than unity.

It is quite easy to test with a light dimmer, an incandescent bulb, and a good PQ meter like the Fluke 43B. The Fluke will show both true power factor and displacement power factor. Changing the dimmer from full on through the range to zero on you will see the displacement power factor drop.

If you had attended EPRI's Power Quality Assurance conference in New York last year you could have seen a paper presented on this very topic. It is not much of a problem when you are dimming a light bulb, BUT when you want to control a 4kV, 2 MW resistive heater....then it can have a negative impact on your bill (power factor penalties are not uncommon).

Charles Perry P.E.

Charles that was most informative. Thank you. I agree the power factor should change off 100% on but it's not untill You get below 50 or 40% load. The total load is 6600 watts of heaters. With one controler or 2 in line There is no error in Power factor untill 40 or 50 % power reduction if there is then it must be very small This could be becuase the controllers are not standard light dimer. They cost 7 bucks and these cost 100 Each. Or perhaps the genset has some built in PF correction and we don't see the error? Anyway, thanks for help. At least we know just what is going on. Just guessing but based on readings it looks the PF is dropping from to about .08 at the low end of power on. .