kill a watt meter measuring discrepancy

Options
balee123
balee123 Solar Expert Posts: 86 ✭✭
in General Solar Power Topics #1
Set up:

Large capacity 12 volt battery system
12V Prosine 1800 inverter, feeding a home electrical panel.
14.8V is what this battery setup will provide. It involve's use of the Chevy Volt's 12V accessory battery which is fed by the cars high voltage traction battery.

The 12V battery is connected to a 12V Prosine 1800 inverter through 4/0 gauge cable.
A short 1 ft 12 gauge extention cord comes out of the Prosine into Kill A Watt meter.
The Prosine 1800 feeds the electrical panel through 12 AWG wire.

The Prosine's display is indicating 14.8V and 22 DC amps (325W) with the load I am running off the inverter. If the inverter is 90% efficient, that would be ~290W.

When I measure the current going into the panel with a clamp on meter (this is post Kill A Watt meter) I get 2.4A, or ~288W............This make total sense.

The problem is that when I look at the readings on the Kill A Watt meter, I get:

147 VA
123 W
1.23 A
118.7V
0.87 PF
60 Hz

I have "reset" the Kill A Watt meter and get the same measurements.

Why aren't my reading closer to the amounts reaching the panel? 1.23 A at Kill A Watt meter versus 2.4 A at the panel? Is my meter defective, or am I missing something? Have I hooked up the Kill A Watt meter incorrectly?
· Share on Twitter

Comments

  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    #2
    Options
    Re: kill a watt meter measuring discrepancy
    balee123 wrote: »
    Why aren't my reading closer to the amounts reaching the panel? 1.23 A at Kill A Watt meter versus 2.4 A at the panel? Is my meter defective, or am I missing something? Have I hooked up the Kill A Watt meter incorrectly?

    Have you tried checking the Kill-a-Watt readings by using a known power load (like a 100 watt light bulb, which should be within 10% of 100 watts if the voltages are close to its rated voltage)?

    Are you sure that the Prosine is working properly to produce a true sine wave? Any severe waveform distortion (approaching MSW) can cause different meters to give wildly different readings, depending on whether they measure true RMS or just throw in a scale factor from peak voltage or current.
    SMA SB 3000, old BP panels.
    · Share on Twitter
  • balee123
    balee123 Solar Expert Posts: 86 ✭✭
    #3
    Options
    Re: kill a watt meter measuring discrepancy
    inetdog wrote: »
    Have you tried checking the Kill-a-Watt readings by using a known power load (like a 100 watt light bulb, which should be within 10% of 100 watts if the voltages are close to its rated voltage)?

    [QUOTE/]Are you sure that the Prosine is working properly to produce a true sine wave? Any severe waveform distortion (approaching MSW) can cause different meters to give wildly different readings, depending on whether they measure true RMS or just throw in a scale factor from peak voltage or current.

    Yes. A 60W bulb, showed 60W when directly plugged into the Kill-a-Watt meter.

    Not sure. Haven't tested with O-scope, however the specs on the inverter say:
    Output waveform (resistive load) Sine Wave(<3% THD, 1% TYP.)

    Are there other ways to test for production of a true sine wave?
    · Share on Twitter
  • BB.
    BB. Super Moderators, Administrators Posts: 33,447 admin
    #4
    Options
    Re: kill a watt meter measuring discrepancy

    The farther away the load's current is from sine wave, the more the various meters will start to disagree... What was the load you were driving? Even if the inverter is sine wave, the load does not have to draw current in a sine wave pattern (from the PF/VA/Watts readings, it appears the load is "non-linear").

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
    · Share on Twitter
  • balee123
    balee123 Solar Expert Posts: 86 ✭✭
    #5
    Options
    Re: kill a watt meter measuring discrepancy

    Hi Bill,

    The loads were a mixture:

    TV/entertainment center
    lights
    Frig/Freezer combo (1.4A draw)
    very small Wine Fridge (1.2A draw)
    Chest Freezer (??)

    Could the Fridges and Freezers lead to this problem since most of the load was probably these?

    The interesting thing is that when I added a 60W light bulb, adding this load was not linear. I think it only increased 16W. Could the fridges and freezers affect how the 60W from the light bulb would show on the meter?
    · Share on Twitter
  • BB.
    BB. Super Moderators, Administrators Posts: 33,447 admin
    #6
    Options
    Re: kill a watt meter measuring discrepancy

    The electronic loads have "non-linear" current draw (typically). Usually drawing spikes of current near the size wave voltage peaks.

    Motors tend to still be sine waves, just the current lags the voltage by XX degrees. So the assumptions of measurements made by the Kill-a-Watt meter are more accurate.

    Otherwise, you have to start spending some serious cash on RMS (root mean square) reading meters and power meters.

    Look at the DC side--Did you see the DC current go up by the addition of the light bulb?

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
    · Share on Twitter
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    #7
    Options
    Re: kill a watt meter measuring discrepancy
    balee123 wrote: »
    Hi Bill,

    The loads were a mixture:

    TV/entertainment center
    lights
    Frig/Freezer combo (1.4A draw)
    very small Wine Fridge (1.2A draw)
    Chest Freezer (??)

    Could the Fridges and Freezers lead to this problem since most of the load was probably these?

    The interesting thing is that when I added a 60W light bulb, adding this load was not linear. I think it only increased 16W. Could the fridges and freezers affect how the 60W from the light bulb would show on the meter?


    The fridge and freezer are going to be fairly linear loads, but with an inductive power factor. The entertainment center, like compter power supplies and most modern electronics, will be a very non-linear load which draws high current spikes near the peak of the voltage waveform. A load like this will seriously confuse many watt and VA meters. Neither the watts nor the VA will be equal to the product of the RMS voltage and the RMS current.

    If you have only linear low power factor loads, you would indeed expect, from trigonometry, that the VA of the circuit would not be increased by the full 60 watts, because the current vector it adds is not parallel to the initial load vector.
    But an accurate measurement of the watts before and after would have to differ by exactly 60 watts. Since it does not, it is a strong suggestion that your initial load was non-linear enough that all of your watt meters were giving incorrect readings. The one measurement you can have some confidence in is the DC power consumed at the input of the inverter, since the voltage is nearly constant.
    But, just to make things less certain, the efficiency of an inverter can vary strongly with the power factor of the load.

    I would try leaving out the TV/entertainment center to see if the figures agree better with it removed from the picture.
    SMA SB 3000, old BP panels.
    · Share on Twitter
  • balee123
    balee123 Solar Expert Posts: 86 ✭✭
    #8
    Options
    Re: kill a watt meter measuring discrepancy
    BB. wrote: »
    The electronic loads have "non-linear" current draw (typically). Usually drawing spikes of current near the size wave voltage peaks.

    Motors tend to still be sine waves, just the current lags the voltage by XX degrees. So the assumptions of measurements made by the Kill-a-Watt meter are more accurate.

    Otherwise, you have to start spending some serious cash on RMS (root mean square) reading meters and power meters.

    Look at the DC side--Did you see the DC current go up by the addition of the light bulb?

    -Bill

    yes, the DC side (measure on the inverter's incorporated LED display) went up by exactly the addition of the light bulb.
    FYI, I saw the non linear power even before I turned on the entertainment center. I wonder if I should try a different short extention cord from the inverter to the Kill-a-Watt meter. I can't plug the Kill-a-Watt in directly to the inverter since the outlet is recessed too far.
    · Share on Twitter
  • inetdog
    inetdog Solar Expert Posts: 3,123 ✭✭✭✭
    #9
    Options
    Re: kill a watt meter measuring discrepancy
    balee123 wrote: »
    yes, the DC side (measure on the inverter's incorporated LED display) went up by exactly the addition of the light bulb.
    FYI, I saw the non linear power even before I turned on the entertainment center. I wonder if I should try a different short extention cord from the inverter to the Kill-a-Watt meter. I can't plug the Kill-a-Watt in directly to the inverter since the outlet is recessed too far.

    The extension cord will have such a low resistance that it will not affect the situation much at all. And any effect it has will be purely resistive, so no effect on the watts versus VA situtation. Did you try unplugging the entertainment center, since it may still be drawing as much as 20 watts (non-linearly) even when it is "off"? That holds true for just about anything with a remote on/off control. Also look out for cell phone chargers and other loads you may habitually leave plugged into the wall when not in use.
    SMA SB 3000, old BP panels.
    · Share on Twitter
  • H2SO4_guy
    H2SO4_guy Solar Expert Posts: 213 ✭✭✭
    #10
    Options
    Re: kill a watt meter measuring discrepancy

    I have a Trimetric meter that can be set to read in watts and it is neat to compare the readout on the Trimetric with the Kill-A-Watt meter which will not show the inverter draw.

    Skip
    12K asst panels charging through Midnite Classic 150's, powering Exeltechs and Outback VFX-3648 inverter at 12 and 48 volts.  2080 AH @ 48 VDC of Panasonic Stationary batteries (2 strings of 1040 AH each) purchased for slightly over scrap, installed August 2013.  Outback PSX-240X for 220 volt duties.  No genny usage since 2014. 
    · Share on Twitter
  • BB.
    BB. Super Moderators, Administrators Posts: 33,447 admin
    #11
    Options
    Re: kill a watt meter measuring discrepancy

    It is unfortunate--But a $20 to $30 kill-a-watt meter is just not as accurate as a $300 to $10,000 watt meter/power analyzer.

    If anyone has some experience with a good quality, reasonably priced AC or DC power meter--please let us know. A "cheap" $330 Meter from Extech still only goes to 10 watt resolution (40 kW peak scale--bit larger than we need here). And I don't think it does Watt*Hours / kWH anyway -- Which is what we really need.

    But it is, usually, a whole lot more convenient.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
    · Share on Twitter

Categories