taking Home Automation to the next level...

[totallymaxed]

There's a lot of municipalities that won't let you connect to their meters..  I would bet that most won't.  If you can, all the power to you! (pun intended)

The units that do that here do not physically connect to the supply. They have a clamp on attachment that attaches to the cable without breaking or damaging the insulation and uses inductive sensing to measure the load.

[bulek]

There's a lot of municipalities that won't let you connect to their meters..  I would bet that most won't.  If you can, all the power to you! (pun intended)

Hi,

I also had problems with mine electricity provider (since they offer similar service for a lot of money - for instance web access to only average consumption per 15 mins interval is something like 9 EUR/month). But I contacted our national supervisor agency and they couldn't find any legal restriction that would allow providers to prevent users from accessing devices they ought to buy. Currently this area is a bit gray in legal meaning. Maybe this will change in future, when providers will put their pressure on this service...

By plain logic, we must buy our meters on our expences and there is awkward if someone else is limiting use of such device by no apparent safety or other "legal" reason - beside plain preference of selling own services for the same thing... Pulse output is galvanically separated from other parts of meter and is technically meant for such use (majority of "big" electricity consumers is already using this feature....).

Maybe there is a way to avoid your electricity provider's unlogical restrictions in a perfectly legal manner...

HTH,

regards,

Bulek.

[ddamron]

I agree..  I can only think of two ways to do so:

1.  Use a Field Effect Transducer to sense the spinning magnet inside the meter
2.  Passive Current sensing after the meter, but before the panel.

The first would be the most accurate.. but then again, Utility companies don't like to see 'electronics' wrapped around the meters... (something about making them spin backwards...)

The second is the type of device I was aiming for.. more or less..  If cheap enough, we could implement it on a per circuit basis too.. (that would show WHAT is using all your power!)

There is a third way, Active Current Sensing inside the panel. (I doubt this is viable)

Regards,

Dan

[Matthew]

I agree - a lot of various features seem to be important for each of us. But we have one strong common point - we need to implement generic support for possibility of showing state of any device on floorplans and that would solve all those problems. This is basic framework - on top of that, alot of new features can be added - but we need to solve basic problems (in my view "generic state changed" event or something similar can solve this)... Also a lot of stuff is currently hardcoded in plugins, so in majority it's not enough if you add certain device to certain category that it will be shown on floorplan... Behaviour is currently not consistent.

I have talked to Chris and Eugene some time ago, and we agreed that such support will be experimentally added for lighting and climate plugin, but currently I'm not sure if this has happened.... Will check that with Chris tomorrow.

I agree that truly generic device/floorplan support is essential as the base on which to add any other function, including the energy properties of existing devices, and a new device type that monitors energy at the device or supply points. But that's just system support for an application that could use such data for management. Like an app that uses a floorplan to reduce lighting to the minimum redundancy. Or running some appliances, like washer/dryer, only in offpeak grid times, or against queries of Internet parameters for overall energy efficiency (eg. when power price rates are lower). Or other automation for conservation. There are probably already existing apps that have existing interfaces that consume power usage data.

Regarding energy consumption :
1. I agree, energy metering is important, but when you'll see numbers with your core and MDs running + other amplifiers, devices etc... :-(   ...
I currently have this already implemented in home automation level, and I can see current consumption. And I'm a bit sad, we need to do a lot of work to bring such consumption down - one of the most important features of smart home platforms is also lowering energy consumption, but with system like LMCE it currently rises too much....

It looks like LMCE now has a feature to suspend/restore MDs to/from Flash drives in a few seconds per cycle. Just adding a timeout GUI query for MD shutdown could reduce their energy from "always on" to "on for a couple hours a day". If an installation has more than one MD, such operation could reduce power consumption by over 50%, perhaps even closer to 99% (like a single core and 20 MDs in a large installation).

LMCE already has Bluetooth support. It could build on that BT support to include "presence" logic. Real consumption could track which room has a user's BT device (like a cheap BT earpiece), and activate devices (including lights) only in that room. We might even see "Bluetooth slippers". There might eventually be some way to heat only rooms with presence data, like when lots of usage data shows that some zones can be shut down for some times of day.

There are other directly relevant apps. Like scenarios that shut down appliances except essentials during blackouts, while the whole house runs off a large UPS, and calculates allowances for discretionary use in the duration. If an installation has a generator (stored fuel, wind/solar, etc) and can sell back to the grid, that complexity will be even more popular to actually exploit if something like LMCE can manage it.

But the key is standardized interfaces. There are other projects for managing homes which might be better platforms for power management than is the Linux Media Center Edition. LMCE might be the system to collect power use info from sensors, and control the state of other devices. It might be the UI for creating scenarios that include power management factors. But perhaps other systems that aren't really "media center" systems are the better scope within which to handle the rest. It's LMCE's job to make sure it can interop with those. And if there's no existing outside tech, but people add it to LMCE, then that subsystem should be modular and factorable.

[colinjones]

Dan - By passive current sensing do you mean as in a current loop around the conductor? I think that is what totallymaxed is saying that the plogg system does anyway. Plus interestingly, it explicitly states that it measures the power factor too - it actually talks about the voltage/current phase angle, but I assume that this means that their software actually includes the power factor in its calculation.

1) I think that most power meters use a fixed, assumed angle of 90 degrees for domestic power. If so, measuring the actual would give a further opportunity for savings as you could not only limit total power, but also low power factor devices (motors, A/C, etc) which read on your meter as consuming more power than they actually do (rip off, I know!) because their power factor is lower than the assumed one at the meter, and so save more cash - getting into imaginary/complex number mathematics, here, but theoretically it should also save of CO2 emissions as well as the power stations would not have to generate or absorb VARS (imaginary power ... I know, but that is the reality of power generation, most powerstations can absorb the many hundreds of mega VARS that capacitive loads "generate") as well as generate Watts.

2) If there is no Linux version of their software (doesn't appear to be), then there is a bit of a job ahead in the maths area to replicate some of the algorithms so that LMCE could report and manage as well as the normal software....

[Matthew]

There's a lot of municipalities that won't let you connect to their meters..  I would bet that most won't.  If you can, all the power to you! (pun intended)

I also had problems with mine electricity provider (since they offer similar service for a lot of money - for instance web access to only average consumption per 15 mins interval is something like 9 EUR/month). But I contacted our national supervisor agency and they couldn't find any legal restriction that would allow providers to prevent users from accessing devices they ought to buy. Currently this area is a bit gray in legal meaning. Maybe this will change in future, when providers will put their pressure on this service...

By plain logic, we must buy our meters on our expences and there is awkward if someone else is limiting use of such device by no apparent safety or other "legal" reason - beside plain preference of selling own services for the same thing... Pulse output is galvanically separated from other parts of meter and is technically meant for such use (majority of "big" electricity consumers is already using this feature....).

Maybe there is a way to avoid your electricity provider's unlogical restrictions in a perfectly legal manner...

If you can get the consumption data, even in a marked-up webpage, from your power supplier live, then such data could be delivered into the local database via HTTP, too. Maybe something like device templates could be made by semi-skilled users to scrape their pages for the data. Since the power suppliers probably aren't writing their own publishing SW, but rather buying it from a SW supplier, there's probably plenty of redundancy in the interfaces presented around the world. I don't know if it's worth E9:mo, but if it can save that much (or close to it), then it might find wide adoption.

[Matthew]

Dan - By passive current sensing do you mean as in a current loop around the conductor? I think that is what totallymaxed is saying that the plogg system does anyway. Plus interestingly, it explicitly states that it measures the power factor too - it actually talks about the voltage/current phase angle, but I assume that this means that their software actually includes the power factor in its calculation.

If that's how those plogg sensors work, then maybe they can work on any current, even those running different parameters than UK/etc current. Though the plogg specs do say "50Hz", which might not work on, say, North American 60Hz current. The sensor device's own 220V power requirement might be satisfied with a simple adapter to 110V.

1) I think that most power meters use a fixed, assumed angle of 90 degrees for domestic power. If so, measuring the actual would give a further opportunity for savings as you could not only limit total power, but also low power factor devices (motors, A/C, etc) which read on your meter as consuming more power than they actually do (rip off, I know!) because their power factor is lower than the assumed one at the meter, and so save more cash - getting into imaginary/complex number mathematics, here, but theoretically it should also save of CO2 emissions as well as the power stations would not have to generate or absorb VARS (imaginary power ... I know, but that is the reality of power generation, most powerstations can absorb the many hundreds of mega VARS that capacitive loads "generate") as well as generate Watts.

I would like to research that power wasting (and overbilling) phenomenon more. I advise the NYC legislature on tech issues, and they've been interested in electric conservation policies. Which includes some oversight of our huge Con Ed power monopoly. If NYC were to set guidelines, they would probably be adopted elsewhere (the typical response).

[colinjones]

hmmm... I can answer general questions for you on the subject, but the detail is very vague - I'm in IT now, it has been years since I was doing power engineering, and that was at the power generation/transmission stage, rather than reticulation and consumption.

In short - power factor is not relevant to purely resistive loads, but across capacitive or inductive loads (in all households to some extent, motors and other coils of conductor are inductive, things like fluoro tubes are capacitive), the current cycle gets phase shifted relative to the voltage cycle - same thing happens in speakers which causes frequency response changes, hence the speakers are rated as "impedance" rather than "resistance".

As power is the multiplication of voltage and current, if both do not peak at the same time, conventional power (in Watts) isn't as efficient as the 2 highest possible numbers for voltage and current never get multiplied together - that sentence was crap! Does it make sense?

Anyway, this is expressed as a second type of power - imaginary power (in vars), which combined with conventional power (in watts) gives you complex power (in VA - Volt-Amps, which you sometimes see written on the specs for electrical equipment, a sure sign that it is an inductive/capacitive load)

The power grid then needs to compensate for inductive loads absorbing vars, and capacitive loads generating them, otherwise the system voltage will fall or rise, respectively. This in the same way the power grid needs to compensate for increases or decreases in straight power consumption (watts) otherwise the system frequency will slow or speed up, respectively.

Industrial loads tend to have power meters that calculate the power factor (90 degs, or perfectly in phase is a power factor of 1, then as they go out of phase it goes down as a Cosine(?) function of the difference in angle between voltage and current) - this allows them to be more fairly metered for their usage as they tend to be highly inductive loads. This even means that some plants install their own capacitor/shunt farms to correct their own power factor and save money!

Wikipedia would surely have heaps on this stuff... will look to see what I can find...

Edit:BTW, imaginary power isn't as flippant a term as it sounds, it just reflects the mathematics - imaginary numbers + real numbers create complex numbers, a vector, meaning a magnitude plus an angle on an X-Y 2d graph...

Edit2: Here is something, not sure if it helps or hinders! http://en.wikipedia.org/wiki/Complex_power#Power_factor

Edit3: Ahhh, this bit talks a little bit about what I was trying to say! http://en.wikipedia.org/wiki/Complex_power#Reactive_power_flow

[hari]

This even means that some plants install their own capacitor/shunt farms to correct their own power factor and save money!

customer of mine did that in Austria and had enormous savings (big machines).

best regards,
Hari

[Matthew]

hmmm... I can answer general questions for you on the subject, but the detail is very vague - I'm in IT now, it has been years since I was doing power engineering, and that was at the power generation/transmission stage, rather than reticulation and consumption.

In short - power factor is not relevant to purely resistive loads, but across capacitive or inductive loads (in all households to some extent, motors and other coils of conductor are inductive, things like fluoro tubes are capacitive), the current cycle gets phase shifted relative to the voltage cycle - same thing happens in speakers which causes frequency response changes, hence the speakers are rated as "impedance" rather than "resistance".

As power is the multiplication of voltage and current, if both do not peak at the same time, conventional power (in Watts) isn't as efficient as the 2 highest possible numbers for voltage and current never get multiplied together - that sentence was crap! Does it make sense?

Anyway, this is expressed as a second type of power - imaginary power (in vars), which combined with conventional power (in watts) gives you complex power (in VA - Volt-Amps, which you sometimes see written on the specs for electrical equipment, a sure sign that it is an inductive/capacitive load)

The power grid then needs to compensate for inductive loads absorbing vars, and capacitive loads generating them, otherwise the system voltage will fall or rise, respectively. This in the same way the power grid needs to compensate for increases or decreases in straight power consumption (watts) otherwise the system frequency will slow or speed up, respectively.

Industrial loads tend to have power meters that calculate the power factor (90 degs, or perfectly in phase is a power factor of 1, then as they go out of phase it goes down as a Cosine(?) function of the difference in angle between voltage and current) - this allows them to be more fairly metered for their usage as they tend to be highly inductive loads. This even means that some plants install their own capacitor/shunt farms to correct their own power factor and save money!

Wikipedia would surely have heaps on this stuff... will look to see what I can find...

Edit:BTW, imaginary power isn't as flippant a term as it sounds, it just reflects the mathematics - imaginary numbers + real numbers create complex numbers, a vector, meaning a magnitude plus an angle on an X-Y 2d graph...

Edit2: Here is something, not sure if it helps or hinders! http://en.wikipedia.org/wiki/Complex_power#Power_factor

Edit3: Ahhh, this bit talks a little bit about what I was trying to say! http://en.wikipedia.org/wiki/Complex_power#Reactive_power_flow

Well, I understand that the inductive behavior of equipment load can consume power in a way where nonlinear (self-reflexive, thus plottable on an imaginary-space graph)  interaction between the current and voltage makes the demand curve less purely sinusoidal than the supply curve. But I thought that the device itself, or maybe the device's site, would have to compensate for that with capacitors and inductors, not the supply source. But it's interesting to see how the power industry actually delivers the compensation.

What I'd like to research more is the waste that comes with that architecture. Either actual waste by overgeneration (or inefficiencies), or just accounting waste, where the accounting model overcharges for actually lesser generated power (if that's what happens). And more to the point, what can onsite power management do to directly compensate for that real/accounting waste (other than just turn off when unused). Especially in residential sites where LMCE fits.

[Matthew]

"Watt's Up" devices are single power socket meters with ethernet to an embedded httpd or USB (unknown protocol) for North American residential power.

And the PowerMeterStore.com has a catalog of them.

[1audio]

Power meters vary around the world. in a sense the US is very behind since a large majority of the meters here are electromechanical, most based on a design that was last changed in the late 1930's. It does measure true power and that is also true for meters in the rest of the world. Where money is involved things can get remarkably sophisticated quickly and this happened around 1900. I linked two products that do interact pretty accurately with the meter. One literally watches the spinning dial and reports back to the monitor. The other comes with two current clamps and a voltage sensor that communicates over the powerline to the readout module. The readout module has all the calculations of the complex metering tariffs (5 levels here in California for residential service).

The rest of the world is moving to electronic meters fast to save money on reading meters (Its starting here in the US but Utilities are notoriously bad about implementing change, PG&E in California started the project 10 years ago and have not yet installed any that I know of.) The electronic metering opportunity created the market for smart metering chips, which spawned the devices I mentioned and the Kill_a_Watt http://www.p3international.com/products/special/P4400/P4400-CE.html which sells for $25 on eBay.

Italy standardized on the Echelon meter and Lonworks for reporting back to the system. And they are reputed to use it to manage loads in the home. I was told that the infrastructure is limited so the appliances talk to each other to manage the maximum load on the circuit. I.E. the washer turns off while the electric range is running.

Getting LMCE to work with this stuff would make it very powerful. As for energy management I see the biggest opportunity would be to replace the MD with a media adapter like one of these http://popcornhour.com/ that draws 15 Watts when running and starts pretty fast. And then to actively manage energy in a home. After the big things are improved it will take something like LMCE to reduce usage of things like low power lighting, HVAC and all of the other small loads.

[Matthew]

I think that if there's existing Linux SW with standard interfaces/protocols for managing power other than just computer equipment (and UPS'es etc), then that would be the place for LMCE to start. Unless it doesn't manage enough actual devices, and it would be less work to produce something new than to expand the existing SW.

But on a higher level than that, it would probably be best if the LMCE project actually included people with experience writing such power management network SW. Maybe it does, in which case the approach which offers LMCE the most reuse of the most generic but effective code to include the most popular monitoring sensors should be a question for them.

Failing all that, we'll just see who starts working it successfully first, like any open source / open project .

[1audio]

The place to look for power management is commercial building management software. I think Crestron, AMX, Siemens etc. all have stuff for this. We could at least understand the problem and lift specs and requirements for it from their datasheets. However this is still new stuff and there are lots of new things to figure out.

[Matthew]

The place to look for power management is commercial building management software. I think Crestron, AMX, Siemens etc. all have stuff for this. We could at least understand the problem and lift specs and requirements for it from their datasheets. However this is still new stuff and there are lots of new things to figure out.

Well, a cumulative hour or so Googling for various permutations of (linux "power management") showed nothing existing except managing power inside the computers themselves. Maybe if we knew at least the industry term for this facility we could find something to incorporate. Which is why I think the best approach would be to start with someone who already knew what they're talking about .