Re: power measure

"Jasen Betts" <jasen-b@xxxxxxxxxxxxxxxxxx> wrote in message
news:j1ef03-9ul.ln1@xxxxxxxxxxxxxxxxxxxxxx
> On 2005-09-19, John Fields <jfields@xxxxxxxxxxxxxxxxxxxxx> wrote:
>> On Tue, 20 Sep 2005 00:08:55 +1200, Jasen Betts
>><jasen-b@xxxxxxxxxxxxxxxxxx> wrote:
>>
>>>On 2005-09-14, cgrahl@xxxxxxxxxxxxxxxxxxxx <cgrahl@xxxxxxxxx> wrote:
>>>> Hi!
>>>>
>>>> I need to measure the power of some devices. I'm thinking to made a
>>>> digital ampermeter, since I already know the voltage (220 V), I only
>>>> need to measure the current to get the power (according to the formula
>>>> P = I x V), right?
>>>
>>>only if it's 220V DC if it's AC power can flow the opposite direction
>>>during
>>>some of the cycle meaning that VxA is more than the watts used.
>>
>> ---
>> If it's AC, the voltage across the load alternates as does the
>> charge flowing through it. Power though, continues to be dissipated
>> in only one direction. For example, an incandescent lamp doesn't
>> light up for one half-cycle and then dim the next.
>
> I was thinking of an reactive load... I'm surprised you didn't catch that
> with a reactive load for some period of the cycle the current oposes the
> voltage, during that period the instantaneous wattage of the device is
> negative. - energy is flowing from the device into the supply.
>
>> No. You need to measure the voltage and current and, if necessary,
>> convert the values read to RMS, then measure the phase angle
>> between the voltage and current, get the cosine of that angle and,
>> finally, multiply the voltage, the current, and the cosine of the
>> phase angle together to get power.
>
> how do you compute phase diference when there's no gaurantee that the
> current looks anything like a sine wave it seems to me that that, it would
> be harder to compute than calculating the average of a number of
> voltage.current products.
>
>>>for
>>>this you're going to want a microcintroller that knows how to multiply.
>>>(since frequent ADC readings give better accuracy, and software multiply
>>>is
>>>typically 5-20 times slower than hardware multiply.
>>>
>>>not being familiar with the PIC line I can't reccomend a particular
>>>device.
>>
>> Well, then, how about recommending one you _are_ familiar with?
>
> Atmel ATmega48 - 20 MHZ risc microcontroller with plenty of 10-bit A-D
> channels.
>
> Bye.
> Jasen

I would recommend that the OP investigate four quadrant analog multiplier
chips, such as the Analog Devices AD633 and AD532, and the OnSemi MC1495,
etc. There have been several power meters designed and available on the
web using these and similar devices.
Start by looking at the design at
http://www.maxim-ic.com/appnotes.cfm/appnote_number/1160
You should also look at the data*** for the MC1495. Lots of circuit
ideas in it. The MC1495 is obsolete, but the circuit design ideas can be
used with other devices as well... Or, if this is a one-off design for you,
the MC1495 is still available from some surplus outlets ( www.jameco.com ,
www.cietron.com , etc.)
The basic multiplier circuit has been designed for you. All that's left for
you to do is to control the amplitudes of the voltage and current waveforms
into the multiplier circuit.

Cheers!!!!
--
Dave M
MasonDG44 at comcast dot net (Just subsitute the appropriate characters in
the address)

Never take a laxative and a sleeping pill at the same time!!


.