Measuring RMS current / voltage 2

[heaterguy]

We are using a solid state relay to control current / voltage to a resistive load. 480V single phase.

Using a Fluke meter, is there a way to determine the RMS current / voltage?

The meter bounces around. For instance, it will show 0.95 amps to 1.11 amps. Is the mid-point the rms current? The meter will also give a peak amps. If we divide the peak amps by 1.414, is it equal to the rms current?

TYIA,

Heaterguy

 

[itsmoked]

Thanks that V1Bll!

 

[jraef]

As itsmoked already concluded, it sounds as though you have a Zero Cross controller, which works by turning the SCRs fully on, but for only a percentage of each group of AC sine wave cycles, i.e. 12 cycles on 40 cycles off would be 30% average voltage (your example should have been 9 and 30 pulses). This makes measuring current a tricky affair as everyone else has pointed out. The easiest way is to use an old fashioned iron vane analog meter. It will pulsate a little, but just read between the lines so to speak.

If you need to digitize the signal, use a CT with a burden resistor to convert to a voltage, rectify it and make an RC filter for it with a cap with a resistor in series and another one in parallel. The RC filter will give you a good average of the current value as a voltage level, which you can feed into a meter or an A/D converter.

"Our virtues and our failings are inseparable, like force and matter. When they separate, man is no more." Nikola Tesla
Read the Eng-Tips Site Policies at faq731-376
Member, [blue]P³[/blue]

 

[Warpspeed]

It sounds to me like a conventional solid state relay being driven by some sort of PID on/off temperature controller.

It can turn on instantly at any time, stay on for some indeterminate number of cycles, then commutate itself off at the end of the last zero crossing. The on/off cyclic rate could be short or long ?? I am not surprised your Fluke meter is jumping around.

The crude way to measure may be with a rotating disc Kwh meter, or perhaps one of the newer all electronic Kwh meters that can display voltage and current independently. That should have true Rms software that can respond to sudden step changes, or do you need true power input integrated over time?

 

[VEBill]

The reason that I mentioned using a 'scope is that we've probably all seen cases where a DVOM (even a nice Fluke) is fooled by a waveform that doesn't match the inherent assumptions in reducing a time-varying signal to a simple numerical reading (even with a cute bar graph).

The varying reading was actually a lucky break, a lesser DVOM might have just provided the wrong answer with all the self-confidence of a D^4 (dumb digital display device).

The 'scope will show you what the signal actually looks like over any time scale you'd like to dial in. I wasn't suggesting that the 'scope be used to measure the final numbers, just to help figure out what's going on. Looks like it worked...

A power meter sounds ideal for, ah, er measuring power.

 

[heaterguy]

You guys are good!

We found a power controller that also measures harmonics, power factors, etc. We are hooking that up to see if we can get a good read on the power.

The downstream heater has constant resistance, so with a power meter we can figure the voltage and current. Actually, it's the power that we want to know anyway.

It's good that we tried the oscilloscope, because we might have been fooled into thinking that each cycle was idential. This kind of information is valuable to us.

Jraef, we initially thought we could average between the peaks and valleys. If this power meter works today, we can see the accuracy of that method.

Yes, the SSR's are zero-cross.

 

[itsmoked]

Confirm you are using a "POWER CONTROLLER" or a POWER METER?

You say "controller".

What the heck, I'll ask.. What kind are you considering? We might be able to approve or disapprove its suitability for your needs.

 

[heaterguy]

We just finished trying a FineTek PF-3050 and guess what...it's bouncing around just like the Fluke did. We are going to dig into it a little further. Perhaps there's an averaging feature.

 

[itsmoked]

Blurb

http://www.fine-tek.com/series5.html

Details

http://www.fine-tek.com/pdf/PF.pdf

{takes about 60seconds}

Huh doesn't seem to indicate the ability to average over X minutes or seconds...

I think you need a long average like 10+ seconds. Or an energy totalizer that you can divide by any time.

 

[ScottyUK]

Web link:

http://www.fine-tek.com/Product-2004118142750.html

Looks like a low-end instantaneous meter designed for panel mounting. Probably not bad on a load which isn't using integral cycle control, but I can see why it would struggle with this type of load. You need something with an averaging or integrating feature, not just instantaeous readings. This instrument doesn't seem to have one.




----------------------------------

One day my ship will come in.
But with my luck, I'll be at the airport!

 

[Barry1961]

A friend on mine got one of those water heaters in his house. He turned off all the power in the house then ran hot water for 10 minutes and read the watt meter outside. This is probably the best test since it is what you will be billed by.

I guess it would change as the incoming water temperature changed but at least it would be a start. The colder the water the more “on” cycles.

I went over to fix it for him about 6 years later. The heating element had burned out. If I remember it was about $10(USD). He loved it and was sure it saved him a lot of money.

Barry1961

 

[Warpspeed]

Another totally different approach would be to measure load voltage (or current), feed this into a true Rms to dc analog converter chip such as an AD536, and use the Rms corrected output to drive a reasonably high frequency voltage controlled oscillator.

What happens is that as the voltage (or current) goes up and down each half cycle, the frequency of the VCO follows it instantaneously. It could be for example 1Hz per Rms milliamp or 1Hz per Rms volt.

So the number of cycles out of the oscillator correspond to the number of Rms milliamps (or volts) in the load.

All you need then is a long digital divider to turn this into suitable units that can then operate an electromechanical (or electronic) totaliser.

For instance if each oscillator cycle equals 1mA
Divide by 1,000 to get 1 Amp per pulse.
Divide by 60,000 for one pulse per Amp/minute
Divide by 3,600,000 for one click of the counter each amp hour, or whatever.

The output could just as easily be scaled directly in power, provided the load resistance is known and constant. Watt seconds, or Kwh maybe ?

The whole thing can be calibrated by adjustment of the voltage controlled oscillator, so really odd division ratios are not required. Powers of two in the divider chain are always more convenient, but that may require an odd oscillator frequency.

It is just as easy to calibrate something to run at say 4,096 Hz as at exactly 4,000 Hz for example, so that no problem.

I have used a system like this for measuring amp hours of battery discharge, but with an Rms to dc converter it should work equally well with ac.

 

[heaterguy]

WOW Warpspeed, you really are traveling fast. We will probably use an averaging power meter or a true RMS amp / volt meter.

 

[itsmoked]

Repeating for us all... A "true RMS amp / volt meter". Is NOT what you need.[flush]

 

[Warpspeed]

It is if it hooked up to a TOTALISER that accumulates instantaneous power readings over time, hehehe.

 

[itsmoked]

grrrrrrrr.

 

[fangas]

Found around most factories:

A counter (high speed)
A 460 X 150 transformer
A timer
needed; 115 X 12 doorbell x'former

Parallel the 480 X 115 transformer w/ the load
Feed the secondary to doorbell transformer
Feed 2nd'er of doorbell thru diode to counter, through timer, to counter
Set timer to 1 Hr

come back later, read counter
counter value / 216000 (cycles per hour)
Answer will provide the "On" time percentage over 1 hr
Do the math.


Granted might need to scale counter, but still pretty cheep & dirty

Ed

 

[heaterguy]

itsmoked, Looking at your suggestion for the watts up power meter, it only takes 120V input. We are using 480V.

 

[LionelHutz]

We have used a Fluke 43B power meter. It will give you instantaneous and average power measurements. You can let it record for a period of time and it will tell you the average power. You don't even need to run for the whole time selected since it gives a running average power measurement.

It is a single phase/balanced 3-phase power meter. By this, I mean it uses 1CT and 2 volt leads to do single phase measurements but it will also calculate 3-phase measurements based on the 3-phases being balanced.

The meter will also do the same measurements with harmonics, voltages, currents, power factor and also works as a simple scope.

The downside is it's not a real cheap meter.

 

[RudyL54]

Normally when you use a multi-meter to measure voltages or currents, the measured values are in RMS. I hope you set it up to the right AC voltage range.

If the meter reading contantly varies, it is probably the variations of the load that created that.

Are you sure, your meter is not defective?

The best approach is get the average reading.

 

[heaterguy]

What do you think about this device?

http://www.pwrm.com/products/ION6200/