Two meters show different amperage

Merry Christmas

Fred B

Senior Member
Location
Upstate, NY
Occupation
Electrician
I think you can have 2 meters by the mfg and same model and end up with different readings as they all have some level of acceptable deviation percentages +/- in the mfg process. Thus if the meters have a +/- rating of.05% your total variance could end up being .1% between any 2 meters of same model. And if you are using multiple meters different models or even worse result be different mfg the results will be, as already discussed by others, be very off if you need any level of accuracy.
Not sure that there would be any truly 0% +/- meter out there. Thus if a diagnostics requires looking at comparing values between simultaneous loads a single meter that allows for multiple inputs would be needed. These are very costly and for most general purpose testing not practical for a general service technician or electrician, and usually only ones having such are doing so for very specific reasons, like large motor repair or were power quality extremely critical. And costs says you need to be doing alot of that.

But those would be something similar to this
 

ptonsparky

Senior Member
Location
NE (9.06 miles @5.9 Degrees from Winged Horses)
Occupation
Electrical Contractor

gar

Senior Member
211031-1216 EDT

Continuation of my 87 meter experiment. Using AC line voltage, about 124 V, as my input and pulsing it on for 120 mS to the 87 meter the results were 100 V in MAX mode and 174.4 in Pulse mode.

Thus, the meter is reading an RMS value in MAX mode, and the AC peak in the Pulse mode. This is too be expected. And the RMS value is lower than the actual RMS value because of the long time constant in the RMS mode.

On circuit analysis and meter measurements there are two broad areas of interest. These are "steady state", statistically an invariant signal with time. This is also statistically called a "stationary random process". This does not mean the signal is not varying with time, but rather that each statistical parameter is constant with time.

If you have several different meters set to read the same kind of signal, then, if the signal is stationary, then each meter should provide the same reading within its stated accuracy. In the early part of this thread that was indicated as being approximately correct.

When dealing with transient signals, if the meters are not designed to the same criteria, then we should not expect the same result for each meter.

Under steady state conditions one generally knows what to expect. If we have a sine wave input, use an average reading meter, Fluke 27 calibrated in RMS for a sine wave, compared to a Fluke 87 measuring RMS and calibrated to RMS, then we expect both meters to read the same.

But change the waveform to a square wave of the same peak value as the RMS sine waveform, then neither meter will read the same as when reading the sine wave, and further the two meters will not read the same value.

.
 

junkhound

Senior Member
Location
Renton, WA
Occupation
EE, power electronics specialty
To summarize:
When in doubt, put a scope on it <G>

Having done a lot of 380Hz to 800 Hz current measurements, feedback balanced hall effect CT with a scope is the only reliable method to get an accurate transient current measurement.
 

Joey94

Member
Location
Illinois
Occupation
Apprentice
I have a scope but no current probe. Can you recommend any good current probes that are not like really expensive ?


Sent from my iPhone using Tapatalk
 

junkhound

Senior Member
Location
Renton, WA
Occupation
EE, power electronics specialty
I have a scope but no current probe. Can you recommend any good current probes that are not like really expensive ?


Sent from my iPhone using Tapatalk
You can get a decent ac CT (will not measure dc) where you need to string wire thru a hole delivered directly from China within a few weeks for a couple of pictures of Abe off ebay. Just search for current levels and hole size for your needs. Clamp on are higher priced.
 

gar

Senior Member
211104-1547 EDT

Continuing a study of my son's Fluke 87.

Using the PEAK MIN MAX mode, and a DC pulse I got the following results: I gave this meter to my son possibly 20 years ago, and I am using it in DC mode because for short pulses an AC signal does not make much sense, especially from a 60 Hz source. The DC source is a 12 V lead-acid battery, and the switch is a 2N4400 transistor driven into saturation.

This 87 is listed as 1 mS in peak mode. From my test I get:

0.06 mS reading is 8.56 V.
0.20 mS reading is 11.8 V, and
0.80 mS reading is 12.0 V.

The 0.707 point is about 12 times shorter than my 0.8 mS reading. So in some respects the peak reading capability is much better than implied by the 1 mS spec. Thus, pulses longer than 1 mS can be considered to be very close to their actual value.

I need to test this on transformer inrush current.

.
 

gar

Senior Member
211106-1720 EST

Further study on the Fluke 87,

This time I am using a Tektronix function generator as the signal source. Each different signal is a steady state repetitive waveform, either a sine wave or a square wave. Both have a long term average value of 0 ( no DC component ), and are symmetrical in time. I have taken the actual measured values and rounded them to their theoretical values so you can easily see the differences.

The test frequencies were 0.2 Hz, 20 Hz, and 200 Hz.

Peak voltage for all waveform was +/- 5 V.

The Fluke 87 meter settings were
DC 1 mS peak, 100 mS, and
AC 1 mS peak, 100 mS.

................... DC 1 mS peak ......... DC 100 mS ...................... AC 1 mS peak ,,,,,,,,,,, AC 100 mS
0.2 HZ
Square ...... - 5 V .....+ 5 V .......... - 5 V ....... + 5 V ................. - 10 V ...... + 10 V ........ 0 V ...... 2.08 V
Sine .......... - 5 V ..... + 5 V ......... - 5 V ....... +5 V .................. -0.8 V ...... + 0.8 V ........ 0 V ...... 0.16 V

20 HZ
Square ...... - 5 V .....+ 5 V .......... - 0.1 V ....... +0.5 V ................. - 5.5 V ...... + 5.5 V ........ 5.0 V ...... 5.0 V
Sine .......... - 5 V ..... + 5 V ......... - 0.3 V ....... +0.5 V .................. -5.0 V ...... + 5.0 V ........ 3.6 V ...... 3.6 V

200 Hz
Square ...... - 5 V .....+ 5 V .......... - 0.2 V ....... +0.2 V ................. - 5 V ...... + 5 V ........ 5.0 V ...... 5.0 V
Sine .......... - 5 V ..... + 5 V ......... - 0.0 V ....... +0.0 V .................. -5 V ...... + 5 V ........ 3.6 V ...... 3.6 V

If you saw my raw data you might disagree with my rounding, but I think it adequately describes what the meter is doing.

The results for "0.2 Hz, AC 1 mS peak, Sine wave" look out of place and uncorrelated, but I rechecked and basically saw the same results.

I haven't done the transformer inrush test yet.

.
 

gar

Senior Member
211107-1540 EST

One month from today on a Sunday in 1941 was a very important time in my memory.

Back to my previous post. The 0.2 Hz AC measurement seemed out of place, but that is because I did not really correctly think about it.
In AC mode it is clear that most electronic meters have a capacitor in series with the input to strip out the DC component. This means that the meter has an input high pass filter, but the cutoff frequency is quite low relative to audio frequencies. This is well below 50 60 Hz because we want little attenuation at these frequencies,

A true sine wave has no harmonic content. However, a square wave is very rich in harmonics of the fundamental.

So the data we are seeing is that the input filter cutoff frequency is well above 0.2 Hz and well below 50 Hz, and thus, the sine wave getting thru is small. Whereas the 0.2 Hz square has a lot of harmonic content up in the 50 Hz and above range.

Now the readings make sense.

..
 

gar

Senior Member
211109-2353 EST

Continuing my Fluke 87 meter measurements.

The load is a 75 W 120 V tungsten filament bulb.

I don't have my phase controlled control circuit setup at present. So it is necessary to try a number of times to turn on near the AC peak voltage.

The peak load current is very short in duration at 120 V RMS AC line. It somewhat approximates an exponential decay. The current probe is my Fluke Hall device. The probe output feeds both my scope and the Fluke 87 in pulse ( 1 mS ) mode. The current pulse drops quite a bit in 1 mS.
Scope read a peak of 7 A and the 87 6.4 A. At 1 mS on the scope the current was down to about 4.8 A.

So one can get an approximation of the peak tungsten current with a Fluke 87 in pulse mode. But obviously the scope is better.

.
 

Flicker Index

Senior Member
Location
Pac NW
Occupation
Lights
I think it boils down to the difference integration time. If you measure the max water flow from a toilet with a 1 second integration time, you can think of it like there being continuous sets of buckets that each only accept water for 1 second. This value multiplied by 60 gives you the max gpm seen during any 1 second period observed. If it takes 3 seconds to dump 1.6 gallons, then you'd see 32gpm max. If the integration time was 60 seconds, the it would be the maximum amount gathered in any single bucket that's computed as maximum gpm.
 

WasGSOHM

Senior Member
Location
Montgomery County MD
Occupation
EE
On the AC range measure the same voltage with both meters in parallel at the same time.
If the readings don't match within the meter's tolerance range then one or both of your meters has a problem.
 

gar

Senior Member
211110-1555 EST

This thread was started by Joey94. After seeing several of his posts, #1, #6, #11, and #20, it seemed clear that he understood that each of the different meters read about the same in Normal mode, but that in MAX mode there seemed to be some substantial differences between the different meters on the same signal. Thus, it seems that his question was --- why do not the meters provide comparable results in MAX mode?

From reading thru various posts I believe that many responders did not understand Joey94's question.

WasGSOHM in your post #73 your comment seems to imply you have not read all the posts of this thread, and/or did not understand what Joey94 was asking. You list yourself as an EE and as such I would expect you would have a better understanding of what Joey94 was asking.

Also I do not think most responders to this thread have much understanding on how various instruments work.

Steve66 in post #41 you seem to think that Normal mode is real time. I would disagree. A scope I would call real time, or at least a delayed look at real time. An average reading meter, or an RMS meter, or various other types are all averaging meters to some extent. What does average mean for a moving coil meter. Any meter with a mechanical movement will be an averaging device with the averaging time usually defined by the mechanical part of the meter movement. An RMS meter has to be an averaging device based on the definition of RMS. Ideally this would be over an integral number of full cycles, but if the averaging time is long compared to the period of a cycle, then a non-integral number of cycles still provides an adequate reading.

The word MAX reading for meters like a Fluke means the RMS value is 0.707 times the peak of a sine wave of some moderate duration, like 100 mS ( actually probably longer ), derived from some algorithm processing the input signal. I suspect for a Fluke 87 an RMS value is derived from an RMS calculation from the input, but for a Fluke 27 it is a full wave rectified signal scaled to RMS of a sine wave with a sine wave input. Therefore, a Fluke 27 will read the RMS value of a sine wave, something else on a non-sinusoidal waveform. Furthermore most RMS electronic meters do not read the RMS value of a waveform with a DC component.

What does RMS mean? It means the instantaneous input is squared, then that squared instantaneous signal is averaged over some time period, The square root of that average value is then calculated and displayed in some fashion.

.
 

gar

Senior Member
211112-0907 EST

Continuing with my Fluke 87 experiments.

The load for this experiment is a 120 V 60 Hz 1/3 HP Montgomery Ward motor with a run and start coil, and centrifugal switch that disconnects the start coil near full speed.

I am using the previous Fluke 87, the Hall device current probe, a 10x scope probe for voltage, and a Rigol digital scope.

The Fluke is in an AC fixed range mode, and with the MAX mode selected. This is indicated as a 100 mS sample or averaging mode of some sort. The meter is doing an RMS measurement, and the displayed voltage is RMS. My scope measurements are actual instantaneous values. Scope sync is single shot synced from the current channel. I can scroll up and down a horizontal line to make voltage measurements from the waveform.

The Fluke 87 in MAX mode measured:

124.0 V MAX, and 113.6 V minimum, or a 10.4 V drop during starting. From my main panel to my bench there is a fair amount of wire. But with the starting current there is going to be 3 to 4 V drop at the main panel. Whatever that drop at the main panel is it is part of the 10.4 V drop.

I had one scope channel reading voltage at the motor, and the second channel looking at motor current.

The scope maximum peak voltage before motor start was 175 V or 123.7 V RMS. Good correlation with Fluke. Minimum peak voltage during startup was 160 V or 113.1 V RMS, or an RMS drop of 10.6 V. Again a good correlation with the Fluke.

The peak current varied from 42 A at 0 RPM to 33 A just before the centrifugal switch opened.

This amount of voltage drop will cause both a 75 W tungsten incandescent, or a CREE 9.5 W LED to slightly flicker.

.
 

gar

Senior Member
211112-1739 EST

My post numbered #75 is somewhat incomplete because I did not provide the meter reading when in AC Peak mode.

The peak reading in the previous post was from the scope. With the meter looking at the Fluke current in Fluke peak mode the the current was 28 A. This is clearly not the peak value, but is 0.707 times the actual peak. This does not correlate with my results in post #69. Why? I do not presently know.

.
 

gar

Senior Member
211120-0936 EST

It is clear that the Fluke 87 has an internal high noise level that created some problems with previous lower voltage readings. Thus, the experiment for the 1 mS response time has been changed. For the present with this Fluke 87 meter I am using a square pulse from a 12 V battery, and generating the pulse from a 555 timer chip. The results follow:

Source voltage is from 12.2 V storage battery. There is a transistor switch drop in the 555 timer.

Pulse ........ Fluke 87 reading

0.02 mS ...... 3.9 V
0.04 mS ...... 6.2 V
0.10 mS ...... 10.1 V
0.20 ms ...... 11.0 V
0.50 mS ...... 11.3 V
0.90 mS ....... 11.4 V
2.50 mS ....... 11.4 V

.
 

gar

Senior Member
211120-1223 EST

I would like to hear back from Joey94 since he started this thread.

Possibly we can cook up some controlled experiments that will help him understand the differences between his instruments. The kind of parts available to him will determine how one might approach such experiments.

.
 

Joey94

Member
Location
Illinois
Occupation
Apprentice
It’s a lot to process and try to understand. I think my knowledge is still really low and there is a lot to learn for me, but as I’m studying and experimenting more and more I get to understand things better. Right now I tried to compare the three meters with a dc load to see how similar are the readings.
9e0b257899184c70ff1fed4ee5dab422.jpg

Anyway I appreciate all the replies. I will definitely read it all again and again when I gather more knowledge about how it all works.



Sent from my iPhone using Tapatalk
 

gar

Senior Member
211121-2041 EST

Joey94:

I can not tell very much from your last post. I am not familiar with those meters. I don't know what Klien means by an AC/DC meter. So you will need to tell me what AC/DC means.

I decided to stop and go look at the Kline literature. It does not measure a composite AC plus DC signal. Rather you have to select whether you are measuring AC or DC. In AC mode it measures the RMS value of the waveform with any DC component removed. In DC mode it measures the average DC value of a waveform. This is fundamentally what most VOM ( Volt, Ohm, Milliampere ) meters do.


So AC/DC to me has to mean one of several things. It turns out the Klien is not what a true AC/DC meter does.

1. If the meter is really an AC/DC meter, then the meter does really measure the voltage or current of a combined AC and DC signal. This is what an electrodynamometer meter does.

So for a square wave that ranges from 0 to 100 V the average DC value is 50 V, the peak value is 100 V, the minimum value is 0 V, and the RMS value is 70.7 V. This is because the square wave only produces 1/2 the power that a full solid 100 V signal would produce. Doing RMS you square the voltage ( instantaneous values ) add them up over your cycle period, divide by time ( gets the average value of he squared values ), then last you take the square root of that average value. By using a square pulse it is much easier to do the calculations. For a 50% duty cycle square wave this give an RMS value of 0.707 times the peak value. When you square 0.707 you get 0.5 .

If the DC component is removed the most positive value is +50 V, and most negative value is -50 V, the average DC or AC voltage is 0 V, and the RMS value is 50 V. ( Full wave rectified you get a solid 50 V ).

More later. I am trying to figure out how you can get some simply generated voltages. At the moment I do not see value in trying to compare meters until some more controlled signal source can be generated.

.
 
Top