Grid frequency sensor

[stephend]

I'm on the lookout for a sensor that can detect frequency variations in the electrical grid, accuracy to the nearest Hz would be ok.  It's for an automation project, so a sensor that outputs 0-5V DC would be ideal.

[guydewdney]

Im sure Ive seen them on fleabay.... moving coil type panel meters? Take one apart - it must have a frequency to voltage converter inside it...

[johnrae]

If you don't mind me asking  -  Why ?

The national grid runs at 50Hz and is guaranteed to vary by less that a gnats hair

No generator attached to the grid can run at any frequency other than grid frequency - ie all grid connected generators run at EXACTLY the same frequency

In power generation, individual generator load is adjusted by either "increasing" or "decreasing" prime mover speed.  Quotes are used since this is what the controls state but all the operator is doing is either increasing or decreasing the fuel/steam feed rate.

Generator frequency can only vary from grid frequency when it is disconnected from the grid.

Trust this helps - and saves you a few bob

jack




If you are generating your

[guydewdney]

No it doesnt...

http://www.dynamicdemand.co.uk/grid.htm

this is the basic idea I suspect the OP is wanting to emulate.

[wyleu]

Counting clock pulse between resets induced by a zero crossing one way edge detector is one approach that's often used.

[stephend]

It's for an off-grid system based on the SMA sunny island and sunny boy inverters.  The sunny island communicates with the sunny boy by adjusting the frequency slightly (+ or - 4Hz max) causing it to increase or decrease its production.  I want to be able to detect this to turn loads on and off. 
So the real aim is to detect when there is more solar power available than is being delivered to the island-grid by the sunny boy.  There might be more ways to skin this cat...

Checking power consumption on the AC is not a problem - but how would one check the available power from the PV if the inverter is purposefully reducing it?

[guydewdney]

Whats the comms protocol?

[stephend]

Changing the frequency IS the comms protocol!  Well between the island and boy anyway.  It is possible to talk RS485 to both inverters, but I can't find documentation on the specific setting to query - so as a last resort I was hoping to do the detection out of band.

[David]

The national grid runs at 50Hz

Most of the time. However, measuring the frequency is precisely the means of activating the power station in a mountain at Dinorwig when it is needed to provide backup in a hurry. Generally four turbines are kept synchronised with the grid and a reduction in grid frequency automatically opens the water valves for them to start generating. There are other ways of starting them up too, but measuring grid frequency means this can be done in an emergency long before humans would have reacted.

and is guaranteed to vary by less that a gnats hair

The number of complete cycles in a day is still I believe the major standard. Only two things in life are guaranteed and neither of them are to do with the electricity system. The frequency in parts of England was zero when Sizewell B (a "highly reliable" nuclear plant) failed, followed very shortly afterwards by the working units at Longannet (I can't remember if it was one or two units of the installed four which were working, one was certainly being fitted with flue gas desulphurisation equipment and at least another one was off for another reason (perhaps maintenance)).

Generator frequency can only vary from grid frequency when it is disconnected from the grid.

Or if the generator is not directly connected to the system.

[johnrae]

"The number of complete cycles in a day is still I believe the major standard."

In my days as an apprentice (nuclear power) the grid frequency used to be adjusted at midnight to correct the loss or gain of daily hertz.

Perhaps in these days of sub-standard standards the powers that be no longer care about maintaining a decent power system

Jack

Jack

[oliver90owner]

The number of complete cycles in a day is still I believe the major standard

Yep, that is what keeps all your synchronous motor clocks accurate.

Regards, RAB

[wookey]

stephend - I was reading about this last night. The sunny-boys also send power-line modem data at a few hundred baud (so far as I can tell from the spec). I then failed to find a price for such an item to see if it was less expensive than the GBP 79 for their RS485 add-on. I naively assumed that the RS485 data would be in an obvious format - do you have reason to believe it is obfuscated and secret?

[stephend]

Probably not secret as such, but I couldn't find the "channel" to query in the sunny boy manual.  Talking directly to the sunny boys should be straight forward using their YASDI library.  Looking at the manual there are many channels you can query about all sorts of internal state - but whether the SB is limiting it's power doesn't appear to be one of them.
Have posted the same question twice to the sma-america forum... but no response.
Using the SB in an off-grid setup seems to be a second class citizen to it's grid tie role.  So they may not have included support for querying this particular parameter in the SBs....    all speculative of course, I guess I should stop mucking about on forums and just give tech support a call!   

[KenB]

List,

The grid frequency wobbles around all over the shop. 

You really need to measure it to the nearest 100th of a Hz for anything meaningful.

I made up a simple circuit with a PIC and an ac power adaptor which would measure every 2 seconds and send out to an RS232 port at 9600 baud - allowing the frequency to be captured using a PC with Hyperterminal and dumped to XL spreadsheet.

I can dig this out if anyone's interested - should cost under £10 to make.



Ken

[wdh]

The grid frequency wobbles around all over the shop. 
...

Really?
Doesn't it spend (near as dammit) all its time in the range 49.9 to 50.1 hz