🚀 go-pugleaf

To: ihnp4!cbatt!gatech!lll-lcc!mordor!sri-spam!jeffr
Subject: Re: What is inside Dr. Watt ?
Newsgroups: sci.electronics
In-Reply-To: <9861@sri-spam.istc.sri.com>
Organization: ATT Bell Labs
Cc: 
Bcc: 


I believe that the technical name for this device is the Nola Converter,
named after Frank Nola of NASA, who invented this technique.
My  understanding is that these devices measure the phase shift
(read power factor) between V and I, and then clip a part of the cycle
to reduce the time that the current is actually supplied to the motor,
thus reducing the total energy cost.

I seem to remember that this works because common (most?) AC motors draw
the same amount of current regardless of the mechanical load they are
asked to rotate.  They do, however, exhibit a different power factor
(read phase difference between V and I), when subjected to different loads.

In some cases, if the motors are oversized, their power consumption
can actually be in excess of the amount actually needed to do the work.
I believe that the device senses this, presumably (only a guess) by looking
at both the E and I waveforms, determining the phase shift, and clipping
some part of the AC cycle.  I would imagine that relatively little of this
is happening when the motor is starting and needs all the current it can get,
but there may be room once things reach a fixed velocity.  It would stand to
reason that in some applications motors would be sized to be able to
accelerate the load within a specified time, rather than for steady-state
operation, hence there should be ample room for power savings in such cases.
Also, since motors come in certain sizes, they are freqently matched to their
application by choosing the next-larger-size technique, with the same result.
Hope that this helps.

Now for a QUESTION: What is the simplest circuit that you can think of
that would measure both the voltage across an AC line, as well as the
current passing through it and do the A/D conversion into 7 or 8 bits,
to be read by a microprocessor?  The conversion should take no more than
1/2 of a 60HZ cycle, or about 8ms.  I envision some sort of inductive
pickup followed by a cheap A/D ultimately winding up at a 8-bit input
port of a micro, the whole conversion triggered at some know angle of
the cycle.  Any ideas? 

Please respond directly, since I don't read this newsgroup often.
I will post responses. Thanks in advance.


					Mark Horbal
					@ Bell Labs
					...ihnp4!ihnp3!mth