The entity that is responsible for maintaining the power line frequency is planning to relax the standards. This means that clocks that run on the AC line will not keep good time. It also means that turntables and tape recorders that run off the AC line with hysteresis synchronous motors may have more significant speed variations. It may not be enough to matter but who knows?
More here: TEC Elimination
More here: TEC Elimination
I do not see how fluctuation of North American power line frequency is at all related to tapping some of the strategic petroleum reserve. Those two would seem no more related than line frequency is to my tire pressure.
Demian, it would seem that a well-functioning power grid would push them to tighter frequency control. Curious...
Demian, it would seem that a well-functioning power grid would push them to tighter frequency control. Curious...
The first problem is that the grid slows down when it gets loaded, just a little but it does. The question is whether to continue the practice of speeding it up when lightly loaded to get the clocks back to the correct time. The max fluctuation seems to be about .1% which may not be audible under any circumstances but who knows?
Sure, just one of those crazy leaps of faith I make sometimes.
If the grid's going to be more unstable in future (with relaxed frequency tolerances) that might be because they know of some pending fuel shortages. That's all.I worked in the electric utility industry for several years. Substations have UFLS "under-frequency load shedding" on their protective relays. If the grid frequency drops a couple hertz, breakers open and dump load to keep the grid stable.
With wind farms and solar power coming, I think they have to loosen up these settings.
With wind farms and solar power coming, I think they have to loosen up these settings.
Maybe I have misread it, but it seems they are aiming at better short-term frequency control at the expense of worse long-term frequency control. It appears that the need to keep clocks correct (on average, over 24hours) is causing them to sometimes drop the frequency too low so they are proposing removing or relaxing the clock requirement.
This should mean that the synchronous motor in your turntable is more likely to run near the correct speed, but your clocks may be less accurate. I guess it depends on whether you listen to a lot of music, or have lots of planes/trains to catch.
This should mean that the synchronous motor in your turntable is more likely to run near the correct speed, but your clocks may be less accurate. I guess it depends on whether you listen to a lot of music, or have lots of planes/trains to catch.
We had a discussion (or was it an argument) about the frequency variations on the UK 50Hz supply system.
About why they deliberately varied the frequency and how "energy" could be stored to meet a sudden increase in demand. Few if any wanted to agree on anything.
I see 0.1% and 2Hz difference in variation estimates by these two contributors.
Last edited:
Energy can be stored. For example, in the UK we have pumped storage hydro schemes such as Dinorwic. Spare electrical energy is turned into gravitational potential energy by pumping water uphill. I believe the Germans do something similar with compressed air in a big sealed cave.
The exact frequency depends on the balance between generation and load. Add extra load and the frequency reduces. Put a bit more coal on the fire and the frequency increases. In essence, it is as simple as that. There are automatic trips if the frequency varies too much from 50Hz, but this rarely happens. The 'instantaneous' frequency is kept near 50Hz by adjusting generation and, when necessary, shedding load. In addition, the 24-hr average is kept even closer to 50Hz by speeding up a little at night to compensate for any daytime reduction. I'm not sure what there is to disagree about on this.
The exact frequency depends on the balance between generation and load. Add extra load and the frequency reduces. Put a bit more coal on the fire and the frequency increases. In essence, it is as simple as that. There are automatic trips if the frequency varies too much from 50Hz, but this rarely happens. The 'instantaneous' frequency is kept near 50Hz by adjusting generation and, when necessary, shedding load. In addition, the 24-hr average is kept even closer to 50Hz by speeding up a little at night to compensate for any daytime reduction. I'm not sure what there is to disagree about on this.
It started with my comparison of mains frequency accuracy to record deck accuracy
http://www.diyaudio.com/forums/analogue-source/105452-so-hows-done.html#post1260360
http://www.diyaudio.com/forums/analogue-source/105452-so-hows-done-6.html#post1260569
http://www.diyaudio.com/forums/analogue-source/105452-so-hows-done-6.html#post1260490
And went on from there.
I think there was another thread that I contributed to, but my first and second searches have not located it.
http://www.diyaudio.com/forums/analogue-source/105452-so-hows-done.html#post1260360
http://www.diyaudio.com/forums/analogue-source/105452-so-hows-done-6.html#post1260569
http://www.diyaudio.com/forums/analogue-source/105452-so-hows-done-6.html#post1260490
And went on from there.
I think there was another thread that I contributed to, but my first and second searches have not located it.
I saw the article in the paper but have no idea exactly how much they'll let things drift. Whenever I've put a frequency counter on my line (upstate NY) it's so close to 60 I'd never worry about it. It's so close that the idea of building a quartz based strobe for a turntable is absurd. I can't remember the numbers but it was something like the mentioned 100 PPM. I assume that's not the case everywhere. Most of my clocks are synced to WWVB or run on batteries. Only the kitchen microwave clock and the old VHS tape player rely on the line. I just can't think of many things that a small error would bother. If they keep things as good as they are now, but don't bother to make up the long term drift, it wouldn't seem to matter. OTOH, how much trouble can it be to make it up?
I think the typical figure of +-0.2% quoted by AndrewT in the linked thread is probably about right for the UK. I would expect excursions beyond +-0.5% to be rare, but +-0.1% to be common. I used to work for the electricity supply industry but it was a while ago. So we are talking about 2000ppm for short term variations, versus 10-100ppm for typical quartz. I would certainly not set my frequency meter from the mains! However, I am reasonably happy with clocks running from the mains as over a 24 hour period they are very accurate - possibly better than a cheap quartz one.
Why do they let the frequency drop a bit during the day? It is because it saves a little energy: lower frequency usually means lower voltage so lower power consumption so less coal/oil/gas/uranium is used. Overnight the load is lower so it is less costly to speed up a little. It is impossible to maintain 'accurate' frequency control second by second without huge cost because it would require large amounts of 'spinning reserve' (generators synced to the grid but exporting no power) and standby generation (expensive things like diesel units or gas turbines). Control is instead carried out by running some units in regulating mode, so a drop in frequency means they automatically put more coal on the fire and vice versa. This fuel change takes a few minutes to work through in increased steam pressure/temperature so control cannot be exact in the short term but is very good in the medium term.
Whenever you measure frequency you must specify the time period over which the measurement is taken. People often forget this.
Why do they let the frequency drop a bit during the day? It is because it saves a little energy: lower frequency usually means lower voltage so lower power consumption so less coal/oil/gas/uranium is used. Overnight the load is lower so it is less costly to speed up a little. It is impossible to maintain 'accurate' frequency control second by second without huge cost because it would require large amounts of 'spinning reserve' (generators synced to the grid but exporting no power) and standby generation (expensive things like diesel units or gas turbines). Control is instead carried out by running some units in regulating mode, so a drop in frequency means they automatically put more coal on the fire and vice versa. This fuel change takes a few minutes to work through in increased steam pressure/temperature so control cannot be exact in the short term but is very good in the medium term.
Whenever you measure frequency you must specify the time period over which the measurement is taken. People often forget this.
Here in Texas it is estimated that clocks will be off by about two minutes over a one year period which translates into about four parts per million. I would imagine that this will be a boon for belt-drive turntable enthusiasts since that kind of error could never possibly be audible, but when the power plant made corrections it was claimed by some to be audible.
John
John
You have to look at both long-term and short-term frequency.
If you scope (line trigger) an xtal oscillator, you can see the line frequency move around, usually it holds for several seconds, worse during the day. Short term variations you might hear with synchronous motors (in tapes machines, older turntables, Hammond organs etc.)
In the long run, over days it averages out and clocks stay accurate. But it looks like NERC wants to eliminate this "Time Error Correction" and is doing a field trial.
I double checked and I set RTU alerts to the Control Center at 0.1Hz and larger (0.167%) swings. Load shedding varied with the region: The Eastern US has load shedding at lower than 59.7Hz, in BC it's 59.5Hz; that's -0.5% to -0.83%.
The couple hertz I mentioned is wrong, I thought it was +/-.
If you scope (line trigger) an xtal oscillator, you can see the line frequency move around, usually it holds for several seconds, worse during the day. Short term variations you might hear with synchronous motors (in tapes machines, older turntables, Hammond organs etc.)
In the long run, over days it averages out and clocks stay accurate. But it looks like NERC wants to eliminate this "Time Error Correction" and is doing a field trial.
I double checked and I set RTU alerts to the Control Center at 0.1Hz and larger (0.167%) swings. Load shedding varied with the region: The Eastern US has load shedding at lower than 59.7Hz, in BC it's 59.5Hz; that's -0.5% to -0.83%.
The couple hertz I mentioned is wrong, I thought it was +/-.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.