[HH] instrumenting a breaker box

Jack Coats jack at coats.org
Thu Oct 23 15:56:57 EDT 2014


The kill-a-watt works well, and the bluetooth hack that LadaAda did is great.

Years ago my wife and I moved into a 'new to us' house, and it was
much bigger than we had but newer.  We had no clue of what to expect.

We did a base line analysis.  We read the meter after turning off all
'optional' items in the house, including the HVAC, and left for 4
hours of shopping.  Before going in the house we read the meter.  This
was our baseline KWH/hr number.  All automatic stuff, phantom loads,
clocks, radios, TVs, were left plugged in.  (And in the day,
ground-fault-outlets used about 3W of power with nothing plugged into
them.)  Then we did a reading of 'bedtime till dawn' usage.
Eventually we kept adding things and recording the delta for each
device or set of devices.  We figured out doing a load of laundry cost
$0.25 and roasting a chicken was about $0.10.  Yes, we got anal about
it.

We got it down to knowing what running a load of laundry cost in the
washer, and the dryer.

Then by regularly reading your power meter (down to an anal-retentive
level) you can generate a profile of YOUR power usage.

Even back then I wanted to put a couple of loops of wire around every
power lead coming out of my breaker panel to measure them.  I never
did.  The only way I could find to do it better would be to have
something like the NEST system (or just about anything that works)
that could read each device separately and log their
on/consumption/off/change in consumption data for analysis.

Back to some version of reality, determine you big hitters, monitor
them with kill-a-watt or similar.  There are 110V and 220V versions.
But a large version is needed for monitoring HVAC's, or electric
dryers, or workshops (power equipment, welders, motors, etc).
Locally, the power company reads meters hourly and the logs are
available to customers online for their account.  If you have a
workshop and put it on a separate meter, it can be checked separately.

All this depends on the granularity of what you need or want to do, as
to how fine your data needs to be.



More information about the Hardwarehacking mailing list