Analog dynamic range, accuracy and number of bits

[rickman]

PF is signed. Power/(Irms*Vrms) isn't.

Why isn't Power/(Irms*Vrms) signed? What happens if the current is
flowing in the opposite direction to the voltage? The RMS value will be
negative and the above equation will be negative. That is the meaning
of a negative power factor, no?

Rick

 

[Jamie]

No, no, no! Think back about the definition of Power (which is the the
integral of V(t)*I(t) averaged over the period of interest. Then think
of apparent power as Irms*Vres (taken over the same period of interest).
Power factor is Power/(Irms*Vrms). These definitions can be used
anywhere at any time under any circumstances and are always correct.
Knowing phase is a special case of using the above on sine waves to make
it easier to calculate under those circumstances.

Oh my fucking god, you can't even get that correct...

Jamie

 

[[email protected]]

Ah - okay. For some reason, I thought power factor correction was mainly
about phase.

Just think of it as PF=W/VA It doesn't matter whether the reason is
phase offset of two sine waves or harmonic distortion.

 

[rickman]

No. Wikipedia defines PF as a dimensionless value from 0 to 1, but
hangs "leading" and "lagging" off to the side.

If voltage and current are in phase, PF is 1 and the sign is
indeterminate. If current phase leads voltage (capacitive load), by
even one degree, PF is called "leading" and if current waveform peaks
later, it's called "lagging." Capacitive (leading) loads are usually
assigned a negative PF if the word "leading" isn't used instead.
Whichever terminology is used, something often needs to be said about
whether the current leads or lags the voltage, and the simple
Power/(Irms*Vrms) equation loses that.

This is getting into the twilight zone territory. Near the top of this
post you are quoted saying, "PF is signed". Now you are saying it is
NOT signed, no?

The point is that power factor is a result of reactance in the load, or
non-linearities. When reactive, the current will be out of phase with
the voltage. Blah, blah... The power factor with a reactive load
tells you how much of the current is reactive and how much is
dissipative. But what if the load is actually shoving power back into
the line?

BTW, Wikipedia is not a primary source. In other words, I don't rely on
it for significant facts unless the references are consulted. Wikipedia
often gets facts *wrong*. Who was the reference they cited for the
power factor info?

Rick

 

[Jamie]

Just think of it as PF=W/VA It doesn't matter whether the reason is
phase offset of two sine waves or harmonic distortion.

Some one finally hits the nail.....!

Jamie

 

[Jamie]

This is getting into the twilight zone territory. Near the top of this
post you are quoted saying, "PF is signed". Now you are saying it is
NOT signed, no?

The point is that power factor is a result of reactance in the load, or
non-linearities. When reactive, the current will be out of phase with
the voltage. Blah, blah... The power factor with a reactive load
tells you how much of the current is reactive and how much is
dissipative. But what if the load is actually shoving power back into
the line?

BTW, Wikipedia is not a primary source. In other words, I don't rely on
it for significant facts unless the references are consulted. Wikipedia
often gets facts *wrong*. Who was the reference they cited for the
power factor info?

Rick

Please read this. I won't bother to explain it myself because everyone
thinks I am an idiot, they maybe correct in that insertion, I am an
idiot in trying to lead people in the correct direction.

True power (Resistive), Reactive Power(Xl,XC etc..) and Apparent, the
difference between the two = PF...

http://www.allaboutcircuits.com/vol_2/chpt_11/2.html

Jamie

 

[rickman]

I took two semisters of Electrical Machinery in college, and I've
designed a dozen or so electronic AC power meters, many of which have
been tested by utilities, universities, and test labs. I wouldn't cite
Wiki if I thought they were wrong.

There's debate about what PF really means in
non-constant/unbalanced/non-sinusoidal situations.

You miss the point about Wikipedia. If I cited my cousin on this
problem what would you say about that? He isn't wrong just because he
isn't a primary reference. But he still *isn't* a primary reference, so
I don't cite him. Wikipedia is full of errors whether this one is right
or wrong.

The point is that citing Wikipedia without citing the real source of the
info is just like citing my cousin. It lends nothing to your case.

Rick

 

[rickman]

I took two semisters of Electrical Machinery in college, and I've
designed a dozen or so electronic AC power meters, many of which have
been tested by utilities, universities, and test labs. I wouldn't cite
Wiki if I thought they were wrong.

There's debate about what PF really means in
non-constant/unbalanced/non-sinusoidal situations.

You miss the point about Wikipedia. If I cited my cousin on this
problem what would you say about that? He isn't wrong just because he
isn't a primary reference. But he still *isn't* a primary reference, so
I don't cite him. Wikipedia is full of errors whether this one is right
or wrong.

The point is that citing Wikipedia without citing the real source of the
info is just like citing my cousin. It lends nothing to your case.

Rick

 

[rickman]

Please read this. I won't bother to explain it myself because everyone
thinks I am an idiot, they maybe correct in that insertion, I am an
idiot in trying to lead people in the correct direction.

True power (Resistive), Reactive Power(Xl,XC etc..) and Apparent, the
difference between the two = PF...

http://www.allaboutcircuits.com/vol_2/chpt_11/2.html

Thank you.

Rick

 

[rickman]

Case? I'm telling you how people define and measure PF in the real
world. If you disagree on anything substantive, say so.

You don't get what I'm saying at all, do you?

Rick

 

[Martin Brown]

U and V both are and u and v for that matter after that you end up with
subscripts - though I have seen S,s & T,t used in hand based analysis.

E gets confused with electric field--having a quantity and its line
integral called by the same symbol is, *ahem*, suboptimal.

Technical vocabulary is funny, especially in EE and other engineering
fields. Normally it's sort of folksy but very descriptive, e.g.
"soakage" or "walkoff". Sometimes it's quite irrational,
though--everybody says "voltage" for "potential", but if you say
"amperage" for "current", everybody laughs at you.

The Brits used to call it "electric tension". Do you UK folks really
say "the tension at this point is fifteen millivolts?"

No, but we still talk of HT and EHT even today.

 

[tm]

E gets confused with electric field--having a quantity and its line
integral called by the same symbol is, *ahem*, suboptimal.

Technical vocabulary is funny, especially in EE and other engineering
fields. Normally it's sort of folksy but very descriptive, e.g. "soakage"
or "walkoff". Sometimes it's quite irrational, though--everybody says
"voltage" for "potential", but if you say "amperage" for "current",
everybody laughs at you.

The Brits used to call it "electric tension". Do you UK folks really say
"the tension at this point is fifteen millivolts?"

Cheers

Phil Hobbs

--

Seen on a T shirt:

E/C^2 -1^-2 PV/nR

JT should know this one.

 

[rickman]

Seen on a T shirt:

E/C^2 -1^-2 PV/nR

JT should know this one.

M 1 T?

Rick

 

[[email protected]]

M 1 T?

You fail (no surprises).

 

[Spehro Pefhany]

You fail (no surprises).

I think it should be E/C^2 (-1)^1/2 PV/nR

For mjT, of course, or miT if you're not an EE.

 

[Tim Williams]

Nah, you're all RC transmission lines, and the clock frequencies stopped
increasing years ago.

Not always. You can dig a trench in a lots-of-metal-layers process, line
it with vias and get a not-quite-as-horrible transmission liney sort of
thing, or pattern a spiral on top to make a crappy inductor. Typical Q is
a max of 10 at a few gigs, but it's an inductor alright.

IIRC, picosecond stuff (pulse gens, samplers) are built on monolithic InP
using schottky loaded transmission lines. I guess the Q of the stuff is
pretty good to do that. "Dielectric loss" of silicon shottkies can be
nasty in certain applications.

An interesting bit of work just came in--building SPICE models of an
ion-trap mass spectrometer. The current product works OK, but
apparently was designed by a chemist, so there are a few production
problems and lots of mysteries.

Hmm, that is interesting.

Tim

 

[josephkk]

While adding dithering noise, oversampling and postprocessing low pass
filtering will certainly give you some extra bits, at least if the
actual ADC is monotonous, I am a bit suspicious of the accuracy of

^^^^^^^^^^ monotonic, though that is rather boring.

Spell checkers canbestupid.

 

[josephkk]

No, you're ignoring PF. It is an amazing widget but it *is* an
integrating power meter, *not* a current meter.


Measure voltage and current and multiply. Why do you care about the
phase angle? It's meaningless if there are any uncorrected electronic
power supplies downstream.

It makes a big difference when you average, integrate, or both. As for
motor or electronic loads, just pay attention to the difference between W
and VA with or without scaling prefixes. The abrupt popularity of PFC
circuits is due to new regulations about harmonics introduction from the
served loads (see CFL lamps which must include PFC).

 

[josephkk]

PF is signed. Power/(Irms*Vrms) isn't.

Signed? Really? I don't think so. Just for fun i checked wikipedia, is
reasonably good on this point.

http://en.wikipedia.org/wiki/Power_factor

?-)

 

[josephkk]

No. Wikipedia defines PF as a dimensionless value from 0 to 1, but
hangs "leading" and "lagging" off to the side.

If voltage and current are in phase, PF is 1 and the sign is
indeterminate. If current phase leads voltage (capacitive load), by
even one degree, PF is called "leading" and if current waveform peaks
later, it's called "lagging." Capacitive (leading) loads are usually
assigned a negative PF if the word "leading" isn't used instead.
Whichever terminology is used, something often needs to be said about
whether the current leads or lags the voltage, and the simple
Power/(Irms*Vrms) equation loses that.

You are conflating phase sign conventions with the unsigned PF value.

For nonlinear and time-variant loads (like burst-mode temperature
controllers) the exact meaning of PF is debatable.

Not in the least applicable to real measurements. The definition is what
it is. Your wishing it otherwise to support your position is foolish.

?-)