Analog dynamic range, accuracy and number of bits

Right, so we're in violent agreement again--they're all RC.

Cheers

Phil Hobbs

Interesting, thanks. Distributed amps are a case that I hadn't considered, since my chipmaking pals are all doing logic.

Reading up on this a littlet, I discover that you can make good slow-wave structures on GaAs very easily:

That looks like what the TriQuint device is doing, with the device capacitance and the high-Z links forming a slow wave line.

Fun stuff.

Now, back to lawyering.

Cheers

Phil Hobbs

If it's conveyed as speech or text, PF is usually expresed as "leading" or "lagging." If it's in a 16-bit register, or an ascii numeric string in a Telnet packet, the lead/lag is usually expressed as a sign. But it matters, and the statement in Wiki that it's in the range 0 to 1 isn't enough for a utility to know which way to kick in a PF correction. You couldn't sell them a PF acquisition box that only reports an unsigned value from 0 to 1.

The sign convention is that leading is negative.

Then you have to hang "leading" or "lagging" somewhere off to the side. That's the same as a sign. Some people and systems literally use the sign. My VME power modules don't hang an ASCII string off in 7 bytes of registers to spell out "leading" or "lagging"; we use a 2's complement (signed fractional) integer.

How is a utility going to do PF correction if their instruments don't return the leading/lagging sign?

Explain what "leading" and "lagging" mean, and why they don't matter to you.

How many electronic power meters have you designed, and how many have you sold, to how many customers? I've done 10 or so, sold to about a hundred utilities, universities, aerospace companies, apartment owners. All but the latter cared about PF and, specifically, about its sign.

Do you have any power meters? Do they report PF? Do they hide the leading/lagging thing?

The fast samplers use a NLTL (shock line) to speed up the sampling pulse edges.

John Larkin a écrit :

I don't know how power meters report, but...

Nothing new there. According to the power/VA definition, indeed you can have PF ranging from -1 to +1. No pb with that.

But now, for one value of PF (positive or negative) you have two possible phase values, one positive and one negative and if you wanted to unambiguously qualify the current vs voltage relation you'd need to tell both signed PF and leading/lagging.

Their customers are _required_ to do so. Larkin's just throwing his usual sand in the air, trying to pretend expertise.

AFAIK the utilities can only correct for motor loads, and that requirement is now, by statute, passed on to the customer, as well as added charges for uncorrected loads. ...Jim Thompson

Wrong. Utilities have both capacitor banks and rotating PF correctors in various places in their distribution systems. Some of their bigger customers may, too.

Whoever does it, they have to know the sign, so they can close a loop towards unity PF.

Idiot, as usual.

"Power factor correction may be applied by an electrical power transmission utility to improve the stability and efficiency of the transmission network. Individual electrical customers who are charged by their utility for low power factor may install correction equipment to reduce those costs."

You don't know much about this stuff.

How's that startup supply coming along? I found a fabulous, bulletproof, super-efficient, dirt cheap way to do it, from 90 volts to 264. Maybe I'll encrypt it and post it here, or maybe I'll only reveal it to people who show up on my doorstep. You don't seem to want to talk about that little problem.

the

choose

Only outside the range of interest.

?-)

I don't know where you get your ideas from, but we are not required to power correct our lines.

The only charge you receive is that of the extra energy that is used if you don't power correct. And we do have 2 of them in use. THere is no charge just for not using them.

Jamie

the

choose

So you are redefining it as signum(sin (theta)) * cos (theta).

?-)

That's a basic electrician knowledge base.

Jamie

the

Works fine for me.

That may be done, but it is not per the definition. Alternately you can partition a bit-field and achieve the same goal more transparently. The difference being lead/lag and magnitude will then not be 2's complement.

?-)

the

think

interest).

make

is

They have their usual meanings in my book. They are NOT part of the definition of PF.

@Bob K7IQ a negative value for PF would physically mean that real power is flowing in the direction opposite to expected. This happens when a motor has a change in load or supply such that the power angle crosses 0 degrees form normal operation. It does happen in real power distribution systems, it is a necessary part of fault current calculations for protective relaying. There are ANSI C37.2 relay functions that specific for this situation.

?-)