Power Grid Freq Variations To Be Allowed

[Gz]

There are, and I have built devices with rudimentary oscillators that sync with the line, but are not so precise, like 555 timers.

Greg

[Jim Yanik]

"Pete C." wrote in news:4e0613a5$0$31685$ [e-mail]:

many of those "flashing 12:00" clocks DO need line frequency to count time. Not all of them have xtal or ceramic resonators providing the 1 sec pulses needed for the time count. Otherwise,a simple lithium coin cell or Supercap backup would retain the timekeeping for a long time,and short power interruptions would not send the clock back to flashing 12:00.

Actually,using line freq is usually more accurate than the xtal or ceramic resonator timebases,over a long time frame. Unless an xtal timebase is precisely tuned and temp compensated,they are not accurate over long times.

I used to do such calibrations.

[Jim Yanik]

"Charlie" wrote in news:iu5a91$pbd$ [e-mail]:

When you have a phase mismatch,you get partial wave cancellations and additions (and standing waves)between the different out-of-phase sources,resulting in power losses,and for the gigawatts electrical grid,that is meaningful.

[George]

Often they are simply counting zero crossings as a timebase so if the line frequency drifts the device does too.

[George]

>

>

Sounds like a he must have turned off his brain when he started working for Obama considering the statement he made.

[Don Kelly]

--------snip-------- Seems like the variation allowed, the more variation syncing is problematic. The whole grid is always loosing power to all the stations producing slight variation pulling on each other.

------- Not really:

a)synchronizing is matching an incoming generator to the grid frequency, voltage and phase. Normal variations in these are not a problem- if they are, then there are more serious problems occuring.

b)There is power transfer between machines- the resultant forces act to pull all machines to the same frequency. You can't run one machine at 60Hz and another at 60.1 Hz. However, the system transmission losses, for short periods of time, may increase or decrease slightly. It averages out.

c) Not correcting time as often will not adversely affect operation of the system - all it means is that there may be (not will be) some more time error before correction. This is really not a big issue.

Don Kelly cross out to reply

[Salmon Egg]

I must have lucked out. For about $30 I bought a Pulsar wrist watch at least 15 years ago. I estimate its stability as about one or two parts in 10^7 as long as the temperature does not have extreme deviations. I only reset it when I need to change for daylight savings time or for a new cell. It is about four seconds fast this now. That may be because of the cell having been in the watch for over two years now and it is probably going to fail soon.

[dpb]

...

Well, my turntable does... :(

[JIMMIE]

The 60 Hz has never been the "rock" you think it is. While its long term accuracy is pretty good its short term accuracy isnt that great and very rarely exactly 60hz. As a matter of fact my freq counter says it is 60.07xxx Hz, the xxx means that these digits are constantly chnging. Georgia Power used to give a tour of their Plant Hatch facility and one of the topics mentioned was frequency accuracy. I would like a couple of examples of any products that use the power line frequency as a CRITICAL reference as I know of none.

Jimmie

[Dean Hoffman]

I wonder how many people under 30 know what a turntable is. Spinning records? Etc. I was wandering around a museum a few years ago. Some teenagers were there on a school tour. One asked me what time it was. I said "It's about a quarter past eleven." That didn't register with her. She apparently had no idea what that meant. 11:15 made sense to her though.

[cjt]

20 minutes in a year isn't exactly "meander[ing] all over the place." And frequency stability is different from long-term time accuracy.

[daestrom]

An interesting story, some of the early crystal watches had their accuracy tied to maintaining the crystal temperature. By accounting for the watch being on a human's wrist, the designers were able to design for a specific temperature at the back of the watch.

In some old power plants, we had ovens to maintain the crystal temperature of our frequency base. But curiously, the frequency base was not used for controlling the generator (base-load plant), only the oscillograph for transient plots.

daestrom

[Mark Lloyd]

[snip]

A digital clock is a counter. A plug-in clock counts the cycles of the AC coming in. A battery clock uses a crystal oscillator.

A plug in clock with battery backup probably does NOT have a crystal, but a cheap RC circuit. That's not very precise. Most seem to be fast during power failures. If power fails from 6:00 to 7:00, the clock may show 7:34.

BTW, I once worked on a fancy clock that runs a chart recorder. It used a PLL to convert the 60Hz to something higher (6.4KHZ IIRC), which would then be divided down to 64Hz to drive a stepper motor. This extra complexity had to do with being able to get the exact same speeds while using 50Hz. b

[Pete C.]

I take it you haven't looked at a computer since the '70s. There hasn't been a computer power supply that cared about line frequency in decades, most these days don't care about line voltage either.

I doubt that there are any synchronous motor based traffic light controls left in use anywhere in the US either. The only places rural enough to have such old controllers are also too rural to have traffic lights.

I don't recall ever seeing a security system that cared about line frequency either, all they care about is if there is line power present.

[Pete C.]

Even the newer / better turntables don't rely on line frequency, they have crystal references and servo loops to maintain accurate speed even with varying loading.

[Steve Barker]

he was probably referring to the old direct drive models that had a neon light and hash marks on the edge that you used to set the proper speed.

[Pete C.]

Nope, that was the very very early days of digital clocks with discrete logic. Any device built in the last few decades that "flashes 12:00" is microprocessor based and will not be using line frequency for reference.

Adding such a backup adds cost to the device, which is why many do not have such backup. Most actual clocks have a backup battery compartment for use with regular batteries.

Sure, but the crystal is accurate enough for consumer use. Many of the consumer devices don't even have AC power available to them to monitor the line frequency of since they use wall wart or line lump DC power supplies.

I've soldered PTC thermister heaters onto crystals to provide temperature stabilization. Most applications, including consumer timing don't need that level of accuracy, particularly given that most such devices have their time reset at least twice a year.

[Pete C.]

Only if the device in question is about 30+ years old and using discrete logic. Anything "flashing 12:00" built in the last few decades is microprocessor controlled and using a crystal timebase.

[Jim Yanik]

Dean Hoffman wrote in news:iu79g2$oa7$ [e-mail]:

didja ever see the Two and a Half Men episode where Jake is spotting hot chicks and calling out the time on his digital watch,not understanding the concept of a clockface as a compass direction indicator.

"hot chick at 12:05"..... when the girl is at their 9:00. 8-)

[Jim Yanik]

daestrom wrote in news: [e-mail]:

all crystal timebases rely on constant temp for accuracy. if the crystal temp varies,the crystal freq varies.That's physics. Temp-compensated xtal oscillators(TXCO's) rely on opposing R and C temp coefficients to compensate somewhat for temp variations. KEYWORD;"somewhat". it's not 100% effective.

These days,they use atomic clocks for precise freq reference. like Rubidium oscillators,cesium,or quantum methods.