RMS of 210 volts ac is (using quick and dirty 0.7 for mental arithmetic) 147, RMS of 240 volts ac 168, so 10 amperes at each for an hour will be 1,470 Wh and 1,680 Wh, which is a significant different is implementing billing.
Also, mains comes with a SLA as regards to voltage, it would be nice to be able to prove brownout and demand a refund or even repair cost to synchronous equipment.
Kill-a-watt measures voltage, current and frequency... very very very useful.
Essentially for me this means flukso is *estimating* consumption over time, not actually recording it or calculating it, which is very sadly a deal breaker, unless / until flukso incorporates this functionality.
Just a voltage and frequency probe on the flukso mains power supply is all it needs.
Many thanks for the speedy reply.
John
drowe67 |
For me the main purpose of the Flukso is to help me reduce household power consumption.
I have a conventional meter in my fuse box for accurate billing.
A few percent difference on the Flukso doesn't affect this mission - for example I can still check the graph and make sure nothing has been left on, or determine where and how high my peak power loads are. Watching the graph instinctively makes me reduce the power the house is consuming.
As a manager said to me once "if you can measure it, you can manage it." He was talking about project management, but I have found this principle to be generally true. Measuring something makes you act to tweak it.
I have an accurate, power factor corrected, plug in type meter for zooming in on individual appliances. Adding voltage sensing would make installing the Flukso much more dangerous compared to a current clamp. Although voltage sensing would allow power factor correction would be nice, more accurate for small inductive type loads like appliances and laptops on standby.
- David
GuyFawkes |
I'm going to disagree with you drowe67, for the following reasons.
It displays kWh, Volts, Amps, Watts, Hz, VA (power factor) and is accurate to 0.2%
Incidentally a "clamp" is quite capable of reading voltage and frequency as well as current, just ask Fluke etc..
What the kill-a-watt does NOT do...
1/ log readings every t seconds
2/ http
3/ more than 1.5 kW loads
It wouldn't actually be that hard to add the missing functionality to flukso, just as long as A, V, VA, Hz readings are better than 1% accuracy, you can then do the following.
a/ monitor power usage, and cost, for the casual user, who appears to represent 100% of the market so far.
b/ monitor line voltage and frequency, for the purposes of comparison to SLA's etc. Every SOHO server room on the planet will buy two of these things then, one for each side of the line interactive UPS.
c/ integrate it with other systems, especially if there is a method to export the flukso data to a spreadsheet, live or static, I can imagine integrating this into a weather station with internal and external temperatures, and external ambient light levels, and not merely predict failing equipment, but make meaningful savings by simple analysis.
d/ actually VERIFY billing, you would not credit how bad domestic house meters are when it comes to accuracy, in UK Law, meters have to be to a standard of accuracy, but, NO INDIVIDUAL METERS are calibrated or checked, the house owner has to request that his meter is calibrated. I know of three who did, all three found that the meter over read by 5%...
e/ the Open Source nature of this project makes all of the above possible, nobody car argue with the results.
cheers
GuyFawkes |
I should say that for the UK, the mains electricity meter must be within 2.5% by Law, and the test fee if you want it tested is £100, you see where you could sell THOUSANDS of flukso's right away here....
I have heard of Uk meters that read as much as 37% over...
Thanks for all the input and especially the pointer to the pdf on metering accuracy. As already stated by David in a previous comment and mentioned on the main page, Flukso wants to help you reduce your resource consumption, electricity being the resource we are initially focusing on. By visualising what we are consuming and comparing our consumption to other Flukso members we can get a much better impression about our electricity usage. E.g.: If my friend's nighttime electricity usage is three times lower than mine, then I should wonder why this is the case.
We indeed use current clamps for measuring aggregate, i.e. mains electricity usage. This is a deliberate choice and is the result of a trade-off between accuracy and ease of installation/safety. One can design the 'perfect' meter, which according to the pdf 'real' meter manufacturers do not even achieve, but if no-one can install the device, it serves no purpose whatsoever. However, for people really keen on a high accuracy measurement, we can provide Fluksometers with a pulse input that can be hooked up to an external DIN-rail kWh meter, see this forum topic. We have some people using these pulse inputs for measuring their PV solar production with a Fluksometer. The latter then measures household consumption (clamps) as well as solar production (pulses from kWh meter). This will be covered in a future blog post.
Comparing a kill-a-watt to a Fluksometer is comparing apples to oranges. The kill-a-watt measures on socket level and thus has none of the installation issues the Fluksometer is confronted with. A Fluksometer monitors total household electricity consumption. A Fluksometer has to obtain its readings from the main wiring. You can't 'just' plug something in between your utility's energy meter and the fuse box. In most countries it's not even legal to tamper with the wiring in your fuse box. As mentioned by David, the clamps can be attached to the mains without any re-wiring (provided that you don't have a PV installation). They provide a full galvanic isolation, include a zener-diode preventing output voltages higher than 7.5 volts DC. And as the utility is indeed bound to an SLA regarding your mains voltage level, it is warranted to assume a value for the voltage. In short, they can be considered a safe way of obtaining fairly accurate readings of your household consumption.
Let me finish with an analogy, knowing perfectly well the limitations of any comparison: Mankind has been using sundials, mechanical and quartz watches for centuries. None of these devices are 'perfect' time measurement tools. They nevertheless provide their keepers with very useful information. Even today, I do not see a lot of people carrying cesium atomic clocks on their wrists. You do not need measurement 'perfection' for a sensor to be useful. You could apply the pareto rule (more or less): 20% of the measurement will yield you 80% of the benefit. The benefit lies in being able to act on the measurement information. Find out where those inefficient devices are. Switch them off or replace them. And let the utility meter accurately measure your reduced consumption.
Regarding your suggestions for double-checking the utility's energy meter: You could have one of those DIN-rail kWh meters installed that provide you with information on voltages, power factors, etc. If you detect a significant deviation between both, ask for a calibration.
Cheers,
Bart.
GuyFawkes |
I'm saddened to read that. I guess I will be looking / buying something else.
Many thanks anyway.
skynetbbs |
Bart,
Did you also do tests at locations where a lot of TL lamps are being used?
They are famous for their negative cosinus phi loads... ?
no the clamps only measure current
they don't even measure direction...
so you can have 230V...400V...voltages ...
there is indeed a difference between 210V-230V
I believe it was checked in this blog using a device that was able to countermeasure current & voltage
do read: http://www.flukso.net/content/testing-new-sensor-board
That is a great pity.
RMS of 210 volts ac is (using quick and dirty 0.7 for mental arithmetic) 147, RMS of 240 volts ac 168, so 10 amperes at each for an hour will be 1,470 Wh and 1,680 Wh, which is a significant different is implementing billing.
Also, mains comes with a SLA as regards to voltage, it would be nice to be able to prove brownout and demand a refund or even repair cost to synchronous equipment.
Kill-a-watt measures voltage, current and frequency... very very very useful.
Essentially for me this means flukso is *estimating* consumption over time, not actually recording it or calculating it, which is very sadly a deal breaker, unless / until flukso incorporates this functionality.
Just a voltage and frequency probe on the flukso mains power supply is all it needs.
Many thanks for the speedy reply.
John
For me the main purpose of the Flukso is to help me reduce household power consumption.
I have a conventional meter in my fuse box for accurate billing.
A few percent difference on the Flukso doesn't affect this mission - for example I can still check the graph and make sure nothing has been left on, or determine where and how high my peak power loads are. Watching the graph instinctively makes me reduce the power the house is consuming.
As a manager said to me once "if you can measure it, you can manage it." He was talking about project management, but I have found this principle to be generally true. Measuring something makes you act to tweak it.
I have an accurate, power factor corrected, plug in type meter for zooming in on individual appliances. Adding voltage sensing would make installing the Flukso much more dangerous compared to a current clamp. Although voltage sensing would allow power factor correction would be nice, more accurate for small inductive type loads like appliances and laptops on standby.
- David
I'm going to disagree with you drowe67, for the following reasons.
I bought one of these (240 VAC Uk version)
http://www.p3international.com/products/special/P4400/P4400-CE.html
some time ago, cost less than 30 euros.
It displays kWh, Volts, Amps, Watts, Hz, VA (power factor) and is accurate to 0.2%
Incidentally a "clamp" is quite capable of reading voltage and frequency as well as current, just ask Fluke etc..
What the kill-a-watt does NOT do...
1/ log readings every t seconds
2/ http
3/ more than 1.5 kW loads
It wouldn't actually be that hard to add the missing functionality to flukso, just as long as A, V, VA, Hz readings are better than 1% accuracy, you can then do the following.
a/ monitor power usage, and cost, for the casual user, who appears to represent 100% of the market so far.
b/ monitor line voltage and frequency, for the purposes of comparison to SLA's etc. Every SOHO server room on the planet will buy two of these things then, one for each side of the line interactive UPS.
c/ integrate it with other systems, especially if there is a method to export the flukso data to a spreadsheet, live or static, I can imagine integrating this into a weather station with internal and external temperatures, and external ambient light levels, and not merely predict failing equipment, but make meaningful savings by simple analysis.
d/ actually VERIFY billing, you would not credit how bad domestic house meters are when it comes to accuracy, in UK Law, meters have to be to a standard of accuracy, but, NO INDIVIDUAL METERS are calibrated or checked, the house owner has to request that his meter is calibrated. I know of three who did, all three found that the meter over read by 5%...
e/ the Open Source nature of this project makes all of the above possible, nobody car argue with the results.
cheers
I should say that for the UK, the mains electricity meter must be within 2.5% by Law, and the test fee if you want it tested is £100, you see where you could sell THOUSANDS of flukso's right away here....
I have heard of Uk meters that read as much as 37% over...
a pdf worth reading
http://www.powermetrix.com/downloads/2009-BillingvsMeteringAccuracy.pdf
Hi John,
Thanks for all the input and especially the pointer to the pdf on metering accuracy. As already stated by David in a previous comment and mentioned on the main page, Flukso wants to help you reduce your resource consumption, electricity being the resource we are initially focusing on. By visualising what we are consuming and comparing our consumption to other Flukso members we can get a much better impression about our electricity usage. E.g.: If my friend's nighttime electricity usage is three times lower than mine, then I should wonder why this is the case.
We indeed use current clamps for measuring aggregate, i.e. mains electricity usage. This is a deliberate choice and is the result of a trade-off between accuracy and ease of installation/safety. One can design the 'perfect' meter, which according to the pdf 'real' meter manufacturers do not even achieve, but if no-one can install the device, it serves no purpose whatsoever. However, for people really keen on a high accuracy measurement, we can provide Fluksometers with a pulse input that can be hooked up to an external DIN-rail kWh meter, see this forum topic. We have some people using these pulse inputs for measuring their PV solar production with a Fluksometer. The latter then measures household consumption (clamps) as well as solar production (pulses from kWh meter). This will be covered in a future blog post.
Comparing a kill-a-watt to a Fluksometer is comparing apples to oranges. The kill-a-watt measures on socket level and thus has none of the installation issues the Fluksometer is confronted with. A Fluksometer monitors total household electricity consumption. A Fluksometer has to obtain its readings from the main wiring. You can't 'just' plug something in between your utility's energy meter and the fuse box. In most countries it's not even legal to tamper with the wiring in your fuse box. As mentioned by David, the clamps can be attached to the mains without any re-wiring (provided that you don't have a PV installation). They provide a full galvanic isolation, include a zener-diode preventing output voltages higher than 7.5 volts DC. And as the utility is indeed bound to an SLA regarding your mains voltage level, it is warranted to assume a value for the voltage. In short, they can be considered a safe way of obtaining fairly accurate readings of your household consumption.
Let me finish with an analogy, knowing perfectly well the limitations of any comparison: Mankind has been using sundials, mechanical and quartz watches for centuries. None of these devices are 'perfect' time measurement tools. They nevertheless provide their keepers with very useful information. Even today, I do not see a lot of people carrying cesium atomic clocks on their wrists. You do not need measurement 'perfection' for a sensor to be useful. You could apply the pareto rule (more or less): 20% of the measurement will yield you 80% of the benefit. The benefit lies in being able to act on the measurement information. Find out where those inefficient devices are. Switch them off or replace them. And let the utility meter accurately measure your reduced consumption.
Regarding your suggestions for double-checking the utility's energy meter: You could have one of those DIN-rail kWh meters installed that provide you with information on voltages, power factors, etc. If you detect a significant deviation between both, ask for a calibration.
Cheers,
Bart.
I'm saddened to read that. I guess I will be looking / buying something else.
Many thanks anyway.
Bart,
Did you also do tests at locations where a lot of TL lamps are being used?
They are famous for their negative cosinus phi loads... ?
http://www.splatco.com/tips/pwrfact/pfarticl.htm