Archived Forum

"Maybe this corresponds to one revolution of the disc?"

I believe most such meters carry the calibration on the face: mine says "187.5 rev/kWh".   When I next get the car out, I'll take a close look at my meter (it has a transparent case) and I might be able to see the ratchet mechanism.

I really think all the accumulated knowledge on this site about electricity meters ought to be collated into one article. Does anyone share that view? Would it be useful?

I'd appreciate that Robert, I'm new to all this I have a Landis & Gyr meter (5258 I think) that was recently installed as my old disc meter was merrly going backwards, I know nothing about how these beasts work. I've just got an Eco Eye Smart PV, this is telling me that I'm importing & exporting whilst my Mk2 is heating water, now I trust that the Mk2 is working faster than the meter so I'm not actually importing and that the smart PV is not mimicking the meter but working much faster, a matter that I've e-mailed Eco Eye with, as yet no reply. It would be good if someone with the knowledge could confirm that all is in order.

While the LED on my digital supply meter remains 'on', I feel confident that I'm not being charged.  Because Mk2 maintains a zero net flow of energy at the supply point, our meter soon enters its 'anti-creep' mode whereby its registers are no longer updated until any significant amount of consumption occurs.

With a non-reversing analogue (disc) meter, the acid test is presumably whether the disc moves forward or not while Mk2 is doing its thing.  Because a disc meter doesn't have a lockdown mode, I would expect the dial to drift very slowly one way or the other when Mk2 is in operation.  Varying safetyMargin_watts should have a linear effect on the speed of drift.  With this value set to, say, 50W, I would expect the dial to slowly drift backwards until the mechanical end-stop has been reached.  With it set to -50W, I would expect the dial to drift forwards at a rate corresponding to around 50W.

If forward movement of the dial can't be halted by this means, then that would be a problem.  Such an outcome would mean that the technique of temporarily lodging surplus energy in the grid doesn't work with disc-style meters.  With no more than 100J flowing to and fro, I would be most surprised if that behaviour was unacceptable to the non-reversing mechanism.

My meter is of this type...

i had a unlimited reverse rotation disc type meter prior to PV..  which was 187.5 rev per kw...
then it was replaced with another disc meter, rated at 200 rev per kw.

it now however only goes back about half a turn..before stopping.

there are 2 stop points, one near the zero marker, one near the 4 marker..  so not exactly half a turn...
one is a bit  more, the other a bit less.

when testing my SSR + arduino i had it set to a 2 second on, 5 second off duty cycle...
and as you would expect... it went forward to 2seconds... and then backwards for 5 seconds :)

until it either hit the stop point (so indicating i'm losing energy to the grid...)
or until the next 2 second on cycle,  depending on how much sun is shining..

i had previously tested it with a kettle on an extension lead...(as the meter is outside.)
and i played around with the kettle switch cycling the power on and off...much to the amusement of the postman and the meter reader (who i'd never seen before) who both gave me very odd looks..

so for my meter.. half a rev.. is about 2.5w/h leeway i guess.

rich.

Shameless beg, does anyone know anything about my Landis & Gyr meter 5258? Mainly what the LED should be doing during import, export, and no flow and what does it mean as it's flashing. The data sheet tells sweet FA, the display reads alternately the usage reading and "red" so I presume I'm on the red phase. does anyone know how many Watts I can pull without being charged etc.

I flatly refuse to stand outside with a kettle on an extension lead, that is far too Victor Meldrew!

richd: so for my meter.. half a rev.. is about 2.5w/h leeway i guess.

Thanks for the quick reply.  2.5Wh is over twice as much flexibility as our digital meter allows.  This suggests that the Mk2 philosophy should work fine with disc meters (phew!)

richmc: Shameless beg, does anyone know anything about my Landis & Gyr meter 5258? Mainly what the LED should be doing during import, export, and no flow and what does it mean as it's flashing.

The manual doesn't help much for our Landis 5235 either.  However, this is how it appears to work:

When consuming power, the LED flashes every 0.001kWh, or 3600J.  As an aside, the flash duration is around 38mS, and can be easily detected using my cut-down 'mini' Mk2 code.  Being able to detect meter pulses is really useful for calibrating your setup.

When exporting power, the LED is constantly on.

If the net flow of power through the meter drops below around 12W for some minutes, it goes into 'anti-creep' mode which (unhelpfully for us) is indicated by the LED being constantly on.  

Now for the fun bit.  Having been charged for some power energy (pulse seen), if the meter falls back below that same level, the LED goes on.  It stays on until it's next time for you to be charged at which point it goes off.  On to off transitions are chargeable events.  If the level drops by less than 0.001kWh after the LED goes on, then that's fine - no energy is lost.  But if it drops too far, some energy is permanently lost to the grid - you've exported it for no gain and it's no longer freely available for your use.  When energy is next drawn from the grid, the LED will go off 'sooner' than it did before.  You've essentially been charged again for the same packet of energy.  Providing that the packets of energy are sufficiently small, i.e well within the 0.001kWh window, the same energy can pass to and fro endlessly without being charged for. 

You can see all this in action by uncommenting the two lines which contain energyInBucket_4led just after the "WARNING" statement in my cut-down Mk2 code.  This will display the amount of power flowing each second in either direction.  Simple experiments with a suitable load should show exactly how your meter works. 

If a kettle seems rather naff, how about an angle grinder.  That should keep the postman at bay ...

richmc

Nobody knows which phase is which (the supply company ought do, but they don't always!), you only reliably know the sequence if you have more than one!

RED means Reverse Energy Detected.  I don't know where I got my manual from now, it was floating around on the web somewhere.

calypso_rae on Tue, 18/0

Robin, does this mean that old disc type meters could have a larger energy bucket than 3600J ?

Would there be any benefit to this?  Perhaps evening out spikes in load a little?

stuart: does this mean that old disc type meters could have a larger energy bucket than 3600J ?

Yes, I suppose that it does.  My value of 1800J really shouldn't be regarded as a key parameter.  Any number will do.  It's simply a threshold above which the triac is turned on, and below which it isn't. 

The upper half of the energy bucket simply serves as an overflow.  If there was no overflow, the bucket would fill up to some huge value after the thermostat had tripped, and the algorithm would then attempt to consume all that (chargeable) energy after the water had cooled.

Having min and max limits of 0J and 3600J respectively feels right to me because this replicates my understanding of how digital meters work.  Mk2 needs a range of around 200J max to do its thing.  This is well within the zone of tolerance in all known types of meter.  The nominal capacity of the bucket is of no importance so long as it's sufficiently large to accommodate that range. 

If calibration is way out, then each 'on' cycle of the dump load will be seen to consume something other than 60W.  If the system is out by a factor of three, then Mk2's operation will nominally cover a 600J range.  With a bucket having a nominal capacity of 3600J, there is considerable tolerance of poor calibration.

None of these absolute values have the slightest effect once distribution of power has got under way.  What does matter, however, is the system's linearity.  More on this to follow on your transformerless thread ...