Mains usage monitor

[paulca]

I looked into creating a mains usage monitor, but left a bit disheartened when it required a 9V AC wall wart to sense the frequency.

I want to make a remote, battery powered watt meter.  It has to be remote and battery powered as the only place I have access to the single mains feed wire is in the external (to my apartment) meter box.  I don't have the ability to plug a 9V wall wart in there though.

I was wondering if the mains monitor I was looking required the 9V wall wart as it also calculated power factor etc.

Would a basic transformer current clamp on the live conductor sensed via some electronics and sent over WiFi using an ESP8266 be doable?

As an aside question... how insane is it to consider running the ESP8266 off a coil wrapped around the mains conductor?  Creating a 1 turn to N turn transformer? To get 5V and rectify it.  (Legally I think it would be fine as long as I tap it after my meter so it measures the consumption, which is doable).

My other option is trying to hack the existing energy monitor I have which is proprietary and a 433Mhz sender.  I don't think I have the tools to do that.  A final option is to buy a better monitor with USB and hack the USB protocol.  I'd prefer to make my own.

[cravenhaven]

To calculate the power you need both the current and the voltage. You should also have the phase difference so that real and apparent power can be calculated.
The 9V AC wallwart is used to measure the voltage and frequency.
Depending on the accuracy required you could plug the AC wallwart into a remote power point, but line losses would affect the accuracy of the voltage measurement.

[paulca]

So, given that I only want indicative watt readings, my mains voltage and frequency is stable enough to consider variation negligible and I don't need apparent/real power or power factor, I would only need the current clamp transformer?

That only leaves the trouble of powering the monitor in the meter box where there is no socket.  A Wifi enabled MCU uses about 50mA, so it won't run longer than a few days on small batteries.  Even if I go for low power, sleep mode and only sample and send the data once a minute it's still going to need it's batteries changed every few weeks.

The existing proprietary meter lasts 3 months on 2 AAA cells and even that's annoying.

I can't put the meter in the apartment without hacking into the consumer unit, as the cabling is all concealed behind a wooden panel.... I rent.  I wlil double check there is no way to access the live feed line.

[fourtytwo42]

The existing proprietary meter lasts 3 months on 2 AAA cells and even that's annoying.

Why not make them bigger and rechargeable ?
Ohh and dont bother with an MCU or Wifi, just send the current signal using a simple licence free radio channel periodically, very economical.

[fourtytwo42]

I dont know if you are still interested in this but the method I use is a photo-transistor on the meters flashing led that flashes 800 times per Kw/hour. Being essentially a digital signal this would be much more suitable for radio transmission than an analogue current transformer signal and the accuracy is as good as the meter The photo-transistor is just stuck to the meter face with blue-tack in such a way the flash is till visable through the translucent transistor case, the meter reader has never complained and it's been in place about 8 years. Food for thought ?

[paulca]

I dont know if you are still interested in this but the method I use is a photo-transistor on the meters flashing led that flashes 800 times per Kw/hour. Being essentially a digital signal this would be much more suitable for radio transmission than an analogue current transformer signal and the accuracy is as good as the meter The photo-transistor is just stuck to the meter face with blue-tack in such a way the flash is till visable through the translucent transistor case, the meter reader has never complained and it's been in place about 8 years. Food for thought ?

Bush man style!  I'm not sure if mine flashes, I have a mechanical meter.  I'd be worried that anything directly attached to the meter will look like a tamper device.  However the meter reading guy didn't bat an eye lid to the current clamp I do have.

Maybe I should just take that thing apart and find out how it works.  If I could hack it's signal on the reciever and get it into an MCU it would be perfect.... bypassing the all the RF gubbins too!

[fourtytwo42]

Well I got stuff stuck on my water meter too  This time a reed switch, again with bluetack and arranged not to obscure the digits. Of course they could tell me to remove it but I guess they realise it's harmless, actually looking down the hole perhaps they think its a slug!! If it's a really old mechanical meter without a nice flashing led maybe you could persuade them to change it (even in rented) by sugesting it might be inacurate lols
Actually if it's the style of mechanical meter with a rotating disk I seem to remember that has a black marker on it that goes by once a rev, possible optical detection ? especially if it lives in a dark cupboard

[lordvader88]

To calculate the power you need both the current and the voltage. You should also have the phase difference so that real and apparent power can be calculated.
The 9V AC wallwart is used to measure the voltage and frequency.
Depending on the accuracy required you could plug the AC wallwart into a remote power point, but line losses would affect the accuracy of the voltage measurement.

Op do u mean a mains to 9VAC wall transformer? I have 2 or 3 little transformers, I should add this to the project list.