Basic question about monitoring 240V circuits and amperage

[sub3marathonman]

I am fairly sure that I'm understanding, but hopefully I can get it confirmed.

With monitoring the 240V circuit, there are two 120V wires carrying current, and possibly a neutral wire. To monitor that 240V circuit, you essentially monitor two 120V circuits, which are going to the same place. As such, when this 240V circuit is monitored by the GEM (or ECM-1240) on a single channel, you then combine the readings for watts, and get total watts.

Now, if you want to also monitor the amperage of this 240V circuit, on the single channel, you have from each leg the amperage from that leg, for example 20 amps. However, since amperage is combined, it would read 40 amps. But in fact it is 20 amps on each 120V portion, which is 20 amps @ 240V. So you would thus divide apparent amps by 2 in order to get the true amperage of that 240V circuit.

One other aspect of looking at amperage is that it just never seems to be correct. I understand that power factor is involved in the calculation, but I thought the power factor stayed relatively close to 1 most of the time, unless it was in a circuit with a motor drawing power. But I'm finding that Volts * Amps is not getting close enough to what is reported as Watts used. From research here, I see that amps is an instantaneous reading, whereas watts is over a time frame, so there is an inaccuracy introduced there, but I would expect over time that inaccuracy would "level out" and doing multiple calculations would lead to the formula V*A = Power (watts).

[Teken]

I am fairly sure that I'm understanding, but hopefully I can get it confirmed.

With monitoring the 240V circuit, there are two 120V wires carrying current, and possibly a neutral wire. To monitor that 240V circuit, you essentially monitor two 120V circuits, which are going to the same place. As such, when this 240V circuit is monitored by the GEM (or ECM-1240) on a single channel, you then combine the readings for watts, and get total watts.

Now, if you want to also monitor the amperage of this 240V circuit, on the single channel, you have from each leg the amperage from that leg, for example 20 amps. However, since amperage is combined, it would read 40 amps. But in fact it is 20 amps on each 120V portion, which is 20 amps @ 240V. So you would thus divide apparent amps by 2 in order to get the true amperage of that 240V circuit.

One other aspect of looking at amperage is that it just never seems to be correct. I understand that power factor is involved in the calculation, but I thought the power factor stayed relatively close to 1 most of the time, unless it was in a circuit with a motor drawing power. But I'm finding that Volts * Amps is not getting close enough to what is reported as Watts used. From research here, I see that amps is an instantaneous reading, whereas watts is over a time frame, so there is an inaccuracy introduced there, but I would expect over time that inaccuracy would "level out" and doing multiple calculations would lead to the formula V*A = Power (watts).

A picture of your systems circuit would probably offer more insight. I also don't recall do you have a Dash Box (DB) or are you using something else to graph and chart the readings?

[mwareman]

My understanding is if there is a neutral on an appliance, then its best to monitor with two channels. If no neutral, then a single channel (and select in the GEM that it's a 220/240v circuit)

[sub3marathonman]

A picture of your systems circuit would probably offer more insight. I also don't recall do you have a Dash Box (DB) or are you using something else to graph and chart the readings?

It is with btmon, but it is exactly what is reported by the GEM Network Utility in Setup Mode, Live Data.

It is one 240V circuit, with a 30 amp (240V) breaker, two CT40s, one on each leg, and combined in one channel (#11) at the GEM. It is for the GE GeoSpring Water Heater, not that it matters though.

So with two CTs, wired correctly at the GEM, the wattage is added together for the correct value. If wired improperly the values would cancel out. The amperage had never been looked at until recently, as I've now with help from awtivy and wci68 and of course mwall at this thread added the ability to monitor amperage: http://www.brultech.com/community/viewt ... =11&t=1363

But it must be that amperage is also additive, as it is showing as 40 amps for 4500 watts. Which it must be 20 amps on each 120V leg. But as I have it set up, and without really thinking about it, it initially appeared as if it was 40 amps on this 240V circuit.

In other words, monitoring the amperage works fine on a single 120V circuit. If you combine circuits, or monitor a 240V circuit, you need to be mindful that the amperage is needing to be corrected, in the case of a 240V circuit, divided by two. As I hadn't considered that when I at first set up my amperage monitoring chart and 40 amps showed up on a 30 amp circuit.

I'm still not getting the exact correlation I'd like though between amperage and voltage and watts. V * A = P (watts) I know that power factor is involved, but I thought most of the time it was a non-issue unless it is on a circuit with an electric motor. (Which in the case of the GeoSpring Hybrid Water Heater it is.) So at some point it may be possible to have the power factor for the circuit displayed / calculated too, but I don't know if there is anything to be done to correct a poor power factor.

[ben]

My understanding is if there is a neutral on an appliance, then its best to monitor with two channels. If no neutral, then a single channel (and select in the GEM that it's a 220/240v circuit)

With monitoring the 240V circuit, there are two 120V wires carrying current, and possibly a neutral wire. To monitor that 240V circuit, you essentially monitor two 120V circuits, which are going to the same place. As such, when this 240V circuit is monitored by the GEM (or ECM-1240) on a single channel, you then combine the readings for watts, and get total watts.

Now, if you want to also monitor the amperage of this 240V circuit, on the single channel, you have from each leg the amperage from that leg, for example 20 amps. However, since amperage is combined, it would read 40 amps. But in fact it is 20 amps on each 120V portion, which is 20 amps @ 240V. So you would thus divide apparent amps by 2 in order to get the true amperage of that 240V circuit.

If there's a neutral you have to monitor both lines as one leg can pull more amperage then the other leg. So the dividing by 2 won't work in every situation.

One other aspect of looking at amperage is that it just never seems to be correct. I understand that power factor is involved in the calculation, but I thought the power factor stayed relatively close to 1 most of the time, unless it was in a circuit with a motor drawing power. But I'm finding that Volts * Amps is not getting close enough to what is reported as Watts used. From research here, I see that amps is an instantaneous reading, whereas watts is over a time frame, so there is an inaccuracy introduced there, but I would expect over time that inaccuracy would "level out" and doing multiple calculations would lead to the formula V*A = Power (watts).

Voltage reported is also instantaneous, for these to match you'd have to have a relatively steady load. I'm reading about 95% PF from a relatively steady load with no neutral through our DashBox.

[Teken]

Ben,

Don't you mean (IF) there IS a neutral both legs need to be monitored? Ideally a 240 balanced load without a neutral should be even.

But as you noted this isn't always true case in point my electric HWT. One element draws slightly more than the bottom one.

I would have never considered that being the case. Thank God for the GEM / DB in letting me know!


=========================

The highest calling in life is to serve ones country faithfully - Teach others what can be. Do what is right and not what is popular.

[ben]

Ben,

Don't you mean (IF) there IS a neutral both legs need to be monitored? Ideally a 240 balanced load without a neutral should be even.

But as you noted this isn't always true case in point my electric HWT. One element draws slightly more than the bottom one.

I would have never considered that being the case. Thank God for the GEM / DB in letting me know!

Correct, edited.

How much more does one leg pull from another?

[sub3marathonman]

Ben,

Don't you mean (IF) there IS a neutral both legs need to be monitored? Ideally a 240 balanced load without a neutral should be even.

But as you noted this isn't always true case in point my electric HWT. One element draws slightly more than the bottom one.

I would have never considered that being the case. Thank God for the GEM / DB in letting me know!

Correct, edited.

How much more does one leg pull from another?

Now, if we had the fancy 48 channel GEM like Teken asked for ...
OK, yes, we'd all be asking for the even fancier 60 channel GEM.

But yes, on the single 240V circuit, unless each leg is monitored on a separate channel, we can't know definitely how much is on each leg. But there was no way I could devote two channels to one 240V circuit with the 32 channel GEM. I've combined several circuits on one channel, but it still works out fine.

With Teken's water heater, I must not understand how you can know when the top element is on and when the bottom element is on. I didn't think both could be on at the same time. Maybe after a shower the top one is definitely on, and then after awhile you can see the change and know the bottom one is on?

[Teken]

48 channel GEM As for the HWT I don't pretend to know when or why the two elements kick in. But, can affirm (IF) you were to monitor both 120 feeds depending upon manufacturer.

You may see a difference as I do . . .

Keeping in mind the difference is so slight its not something a person would really notice. Since I was doing this only once I simply opted to do it right the first time. Its not like I am going to spend anymore money going forward the CT's are in place and if I really wanted to.

I could simply remove one CT and use the GEM's 240 load doubling feature which works just great! My reply about a (balanced) 240 load with out a neutral was more to offer insight that its very common to see a variance no matter how slight.

I saw a difference on a 240 VAC portable heater for the garage in the past also. As an aside looking forward to new and more updates in your projects thread!

[wci68]

I think Ben and Teken have covered this pretty well though I'll add a little too.

With monitoring the 240V circuit, there are two 120V wires carrying current, and possibly a neutral wire. To monitor that 240V circuit, you essentially monitor two 120V circuits, which are going to the same place. As such, when this 240V circuit is monitored by the GEM (or ECM-1240) on a single channel, you then combine the readings for watts, and get total watts.

In the US a typical residential service is a "split-phase" - 240V between the two legs and a neutral in the middle to create pairs of 120V circuits. To monitor both legs using individual CTs on separate GEM channels it is easy. To monitor both legs using individual CTs on the same GEM channel gets a little more complicated but the concept is the same: Keep the voltage at 120V since that is the potential of each leg to neutral. If you set the voltage at double (240V) you'll end up with both current and power at twice the proper values, and energy will accumulate at twice the rate.

But I'm finding that Volts * Amps is not getting close enough to what is reported as Watts used.

Energy (watt-seconds) is actually reported by the GEM as a counter. Combined with the time, which is also actually reported by the GEM as a counter, both average power and energy are reported. The interval for the average is arbitrary - this would be whatever the difference is between the two seconds counter values that correspond to the two energy counter values.

Ben had confirmed that the voltage and current values are "instantaneous" so there will be some differences. I haven't put together a graph yet for power factor and I currently have the GEM sending a packet every 10 seconds. I started another thread on this and haven't had a chance to get back to it yet, but reducing this to 1 second may bring the VA (instantaneous) and W (average) values slightly closer in alignment for a more accurate power factor calculation.