onboard kilowatt-hour meters question

[madderscience]

... such as the E-meter, link-10 and a bunch of others that I can't think of right now.

question:

in calculating kilowatt hours used during driving, do they factor voltage sag of the battery pack when under load into the calculations, or do they only actively monitor current, and multiply that by a preconfigured nominal battery voltage?

So for an example, if I have a nominal 120v battery but it sags to 100V at 200A load, if I drive for 30 mintes at this rate would the meter read 10KWH used, reflecting the voltage sag, or 12KWH used, reflecting nominal voltage?  which option is considered correct?

Thanks
 

[leitmotif]

Both can be correct.  It really comes down to how much money you are willing to trade for accuracy and resolution.  The most accurate of course is actual line voltage and current.  I would assume battery line and current of course since you are tracking miles per Kw.  Depending on where you place sensors you can track overall battery load or just battery load drawn by motor.

Just as a caution - if you are using a regular shunt -
do NOT install (as I did) without double checking polarity on terminals for a current meter -
ESPECIALLY so - if meter is solid state and it uses a common (this is what screwed me up). 

The other thing to remember is even though the shunt output is only 50mV -- each "side" of 50 mV ie plus and neg is still at DC line voltage (plus or minus) depending on which side you place shunt.  I would definitely fuse the 50mV signal leads going to the meter with say 1/2 amp quick blow instrument fuses.

 
Dan Bentler 

[ev-supply]

The other thing to remember is even though the shunt output is only 50mV -- each "side" of 50 mV ie plus and neg is still at DC line voltage (plus or minus) depending on which side you place shunt.  I would definitely fuse the 50mV signal leads going to the meter with say 1/2 amp quick blow instrument fuses.

It is for this reason that shunts are traditionally placed on the 0V side of the battery pack, even though functionally, they can go at any point in the load.

[madderscience]

So if both original scenarios can be correct, what do most watt hour meters do to calculate KWH used?  sample voltage and current, or just current?

Given that a certain amount of KWH are stored in a battery, If voltage sag were primarily due to internal battery resistance, you could probably get an accurate reading of watt hours used with a preconfigured nominal battery voltage.  If voltage sag were due to the chemical reaction in the batteries only being able to progress at a certain rate (so the voltage drop is not due to wasted energy, just a limitation on how fast it can be removed from the batteries) then the more accurate reading would be to sample volts and amps.   I suppose it is in fact a combination of the above factors so the safest thing to do would be to sample voltage and current.

[ev-supply]

Maybe the usage is calculated periodically, averaged, and summed together.  In that scenario, it would basically be tracking the usage with the voltage drop.

[leitmotif]

So if both original scenarios can be correct, what do most watt hour meters do to calculate KWH used?  sample voltage and current, or just current?

I have a CD from City Light that explains how household type meters work.
The Kw hour meter on your house is actually a motor.  It has two windings the one with the small wires (higher inductance) is the voltage.  The one with only several windings is the current.  Because of the difference in inductance there is a phase shift between the two windings.  Now if there is no current the motor does not turn, speed of the motor is proportional to current.  Power is measured by revolutions of the dial.

The most accurate is constant sampling of current and voltage.  For DC work this is most commonly done with a shunt sized for the current and using line voltage.

The only way you could get zero volts on a battery is to ground it.  Since EV batteries are floated ie not grounded there is NO zero voltage.  126 cell 250 VDC submarine batteries are not grounded but believe you me touch ground and any point anywhere in that battery circuit and you will get zapped but good - the insulation was not perfect.  There was a zero voltage at a grounded cell which is how we found which cell to clean.

Dan Bentler

[BSherry]

I don't know what other folks do, but when I had kilowatt/hours in my instruments, I would sum (average) both current and voltage over a period of 1/4 second, multiply the result together to get watts, and sum those each 1/4 second. I would then divide that sum by a calibration factor, that included time, in order to get the number to display as W/H.

I admit it would be better to multiply each sample and sum, but the processor I was using then had no hardware multiply, so I had to compromise. My current units have a processor with multiply, but I haven't gotten around to writing the algorithm. I am always trying to do too many things all together, simultaneously, at once.

Bruce Sherry
Bruce Sherry Designs

[Paul]

The e-meter samples the voltage and current several times a second. Multiplying those together give watts. Watts times time (in hours) divided by 1000 gives kWh. As an example (using one sample per second):

300 amps * 110 volts = 33,000 watts, 33000 * 1/3600 * 1/1000 = .0092 kWh for second 1
100 amps * 118 volts = 11,800 watts. 11800 * 1/3600 * 1/1000 = .0033 kWh for second 2

The e-meter has to do that a whole bunch of times and add up the kWh to give you the result. I don't know how many decimal places the e-meter keeps and I don't know how many times a second it samples.

HTH