🚀 go-pugleaf

In article <184@zorin.UUCP>, chris@zorin.UUCP (Christopher Nielsen) writes:
> I have a room, X. It has a 110V power line running to it. I want to detect
> if there is any current being used in it. I don't want to place a resistor
> inline, but rather, put a coil or something around the 12 guage conductor
> leading to it, and some kind of amp driving a meter, or LED. Any
> ideas/circuts?
> 
> I can probably run an get a clamp on type of meter, but I want this to be a
> semi-permant installation, in a wall box. Hence, what I would like to have,
> is a coil around the conductor in the wall box, 2 wires leading to an
> external detector circuit.

	No problem.  Since your current draw will probably be 15 amps or
less you want a better current ratio than one turn, so just wind about half
a dozen turns of _solid_ #20 AWG insulated wire around the "hot" wire.
Terminate your current pickup coil leads in some #22 AWG twisted pair to
run to the electronics circuit.  A good idea would be to use some silicone
rubber potting compound to keep the pickup coil and its turns in a stable
configuration.

	The above "pickup coil" MUST be terminated in a resistive load in
your electronics circuit (try 50 or 100 ohms as a starting point).  You
want to use an op amp to measure the voltage across this load; NOT the
open circuit voltage of the pickup coil.  There are a number of RMS
detector circuits available in any op amp application guide which you
can use to provide a DC voltage to feed at an A/D converter or to a meter.

 	If you want to measure watt-hours, you will have to integrate
the DC voltage from the RMS detector.  This is a trivial software task if
you are using a microprocessor having realtime capability to monitor this
circuit.  You can perform the integration in hardware using a V/F converter
and a counter circuit, but with inexpensive microprocessors this technique
is becoming a thing of the past.

	If you do not care about great accuracy, you can assume your line
voltage to be a constant.  If you do care about accuracy, you need a
separate line voltage RMS detector, and must multiply the current times
the voltage signal.  This is trvial in software using a second A/D channel,
but a lot tricker (if you are not a good analog designer) in hardware since
you will need to use a multiplier chip.  Also, if you are going to measure
the AC voltage, then use a small plug-mounted transformer for isolation;
I do NOT recommend connecting directly across the power line!

> Steve Ciarcia, of Byte fame, some months/years ago (I can dig it up) had
> an interesting article about measuring power consumption of household
> appliances with your CPU. It involved canabalizing a power strip, and
> placing a number 1 paper clip inline with the outlets, and measuring the
> power drop.

	This sounds like crap to me, and seems like an invitation to
trouble.  I would suggest staying with a safe, isolated current pickup coil.

> in Steve's article, his circuits would output a pulse to the CPU, each time
> .1 KWH, or so would be consumed. So instead of complicated parallel/data
> latch/counter arrangement, you would simply count pulses. Very interesting,
> but not quite what I'm looking for...

	You are going to have to do quite a bit in hardware, and do it
_accurately_, if you want to count pulses.  I would suggest using the A/D
approach, which you can also implement using a serial data interface by
connecting an A/D directly to a UART.  GE/Intersil, among other vendors,
has an A/D converter and UART which are compatible in this mode.

<>  Larry Lippman @ Recognition Research Corp. - Uniquex Corp. - Viatran Corp.
<>  UUCP   {allegra|boulder|decvax|rutgers|watmath}!sunybcs!kitty!larry
<>  TEL  716/688-1231 | 716/773-1700  {hplabs|utzoo|uunet}!/      \uniquex!larry
<>  FAX  716/741-9635 | 716/773-2488     "Have you hugged your cat today?"