How to measure power consumption at home?

[rossini]

It turns out, PG&E actually breaks down my monthly bill by gas, electric, and car charging line items. Once I switched to the EV monthly plan it added that feature. I have no clue how they can tell when I'm charging my car versus when I'm running the dryer (same outlet) but it knows so that answers that question. Talk about simple!

 

[bxr140]

Or Sense

Big +1 for sense. Solid technology, solid user experience. If you like data, its got data. I was skeptical of its load sensing capability, but its pretty decent at identifying different loads, and the loads it can't ID it will just assign a generic name for you to flesh out later.

 

[L_Mont]

Did you insert an additional 240V breaker for the Sense device?
I like the concept but I do not have the space for an additional 240V breaker.

 

[bxr140]

Did you insert an additional 240V breaker for the Sense device?
I like the concept but I do not have the space for an additional 240V breaker.

You can use a tandem 240 breaker or piggyback on an existing breaker if code allows.
Otherwise yeah, you're probably SOL.

 

[Johnny Vector]

Well, if you read the website more thoroughly you'd see in the specs the following:

Signal Processing
1MHz Sampling Rate (4M data points per second) 1GHz ARM Processor

There may or may not be a point to it as far as you know, but they implemented it. I suspect it's a byproduct of the architecture they chose and Marketing decided to trumpet it as a selling point.

There is just no reason to sample that fast. What are you even going to do with all that data? (Answer: filter it down to a reasonable rate, and discard most of it.)

 

[mspohr]

https://smile.amazon.com/DROK-80-30...9Y2xpY2tSZWRpcmVjdCZkb05vdExvZ0NsaWNrPXRydWU=

 

[bxr140]

I suspect it's a byproduct of the architecture they chose and Marketing decided to trumpet it as a selling point.

Yeah, there's plenty of reasons to pick a processor beyond a baseline engineering demand.

 

[L_Mont]

There is just no reason to sample that fast. What are you even going to do with all that data? (Answer: filter it down to a reasonable rate, and discard most of it.)

The RMS voltage and current are typically calculated by sampling the waveform at higher frequencies. After processing, the data could be updated at 1 Hz rate. There is nothing wrong with oversampling and decimating, this is how thing works.

Another reason for sampling at a higher rate is to identify the load. Switching power supply (operating at up to 1 MHz) often introduce high frequency noise on the power line. For example, my true-RMS multimeter has a switch to enable a low-pass filter that remove these high frequencies.