BraveSirRobbin
Moderator
But you will have to do this various times of the day, during different loading periods.
Brultech Electrical Power Monitoring for your Home
- Thread starter BraveSirRobbin
- Start date
Well... I borrowed a voltmeter from work and used my own to see what was going on. I first connected both voltmeters to the same phase to see if they would get the same value and one read about .2 volts lower than the other. Then I connected each voltmeter to a different phase, and taking into account the difference in the voltmeters I disocvered that one phase was about .2 volts lower than the other. This seemed to be pretty consistant and even as the load changed (I started and stoped my heat pumps and the water heater and turned some lights on and off and such). The voltage never varied too much (probably less than 1 volt total) and both phases went up and down at the same time. The one phase stayed about .2V lower than the other no matter what the reading was.
In the real world, this probably isn't a big enough difference to be concerned about. With both heat pumps and my wather heater and a lot of other stuff running I was reading around 40 amps on each phase. 40 amps x 123.6 volts is 4944VA. 40 amps x 123.8 volts is 4952VA, so it's only off by 8VA which is about .1%. And remember that watts will actually be slightly lower than that depending on the power factor, so the error will be a little less.
I'm not sure if the results in my house will apply everywhere, but certainly for me it doesn't look like sampling the voltage from only one phase is going to be a problem.
If anyone else wants to do the test I'd be interested in seeing what they come up with.
I'm not sure if that will make a difference or not, but I'll try again later tonight and maybe a few times during the day tomorrow too.
Brett
In the real world, this probably isn't a big enough difference to be concerned about. With both heat pumps and my wather heater and a lot of other stuff running I was reading around 40 amps on each phase. 40 amps x 123.6 volts is 4944VA. 40 amps x 123.8 volts is 4952VA, so it's only off by 8VA which is about .1%. And remember that watts will actually be slightly lower than that depending on the power factor, so the error will be a little less.
I'm not sure if the results in my house will apply everywhere, but certainly for me it doesn't look like sampling the voltage from only one phase is going to be a problem.
If anyone else wants to do the test I'd be interested in seeing what they come up with.
I'm not sure if that will make a difference or not, but I'll try again later tonight and maybe a few times during the day tomorrow too.
Brett
Just wanted to follow up on my other post. I checked the voltages on both of my phases several more times yesterday and today and I discovered that while my line voltage does seem to fluctuate some (probably a difference of 4 or 5 volts from the lowest reading to the highest reading) the voltage on the two phases seems to stay very close together.
There was a little fluctuation between the phases, but it wasn't off by more than .2 volts. Sometimes one phase was .2V higher than the other, and other times the first phase was .2V lower than the other. Sometimes they were exactly the same or just off by .1V one way or the other way.
In any event, it seems that my two phases are definitely pretty close and I don't think there would be too much inaccuracy to only measure the voltage of a single phase.
I'd still be interested to see what other people find if they try this experiment too.
Brett
There was a little fluctuation between the phases, but it wasn't off by more than .2 volts. Sometimes one phase was .2V higher than the other, and other times the first phase was .2V lower than the other. Sometimes they were exactly the same or just off by .1V one way or the other way.
In any event, it seems that my two phases are definitely pretty close and I don't think there would be too much inaccuracy to only measure the voltage of a single phase.
I'd still be interested to see what other people find if they try this experiment too.
Brett
felixrosbergen
Senior Member
THat would be awesome...does CQC have an ability to store historic values in a database of somekind? I'm not very familiar with SQL, but if i need to run SQL server to store this data i suppose i would find other uses for it as well....(e.g. historic data on how often the water pump runs, how often AC turns on, etc).
I will be on a well....i know it's a bit off topic, but do these well pumps put out a flow consitent enough to approximate water usage by recording how long the pump is on?
beelzerob
Senior Member
Depends on your system. Our well uses a constant pressure pump, so it varies the flowrate of the pump to maintain pressure, no matter how much or little water you're using. Your typical on/off pump MIGHT give you a ballpark.
Actually, this is not correct. Both these units have probes/sensors on each of the feeds. The transformer is just for power for the unit. So both this unit and the TED monitor total usage on both legs.
Chris
looking at the diagram for the Brultech device, the wall wart would supply both power for the device and measure the line voltage as there is no other electrical connection to each leg, just current transformers. This is a smart way of doing it as the interface itself stays as a low voltage device and the wall wart provides a percentage of the line voltage.
BTW, I measured the difference between the two phases as .7 volts but this is probably still not enough difference to cause any significant error. I don't know how accurate these devices would really be and I didn't see any specs anywhere.
Related to this, I came to this page http://www.brultech.com/HomeEnergy/homePowerFactorError.html, anyone know or assume which are the "false or inaccurate" power meters?
Like Brightan said, both devices have sensors to read the current draw on each leg, but in addition to the current draw you need to measure voltage to be able to figure watts or VA. Voltage can't be measured with the coil, but rather needs a direct electrical connection. The transformer not only provides power to the units, but also allows the units to read the line voltage. If you look at the wattnode device mentioned earlier you'll see that it actually has two transformers... one for each leg.
In my opinion, from what I've gathered it doesn't sound like it affects accuracy too much to not read the voltage on both legs.
Brett
Yeah... .7 volts doesn't seem like too much either... I don't think it would be enough that I would worry about it. For comparason, the TED device claims to be accurate within 2% and the wattnode (which does measure voltage on both legs) claims to be accurate within .5%. I searched and couldn't find an accuracy claim for the brultech device.
I don't know if it explains the entire difference, but some of the difference between the .5% of the wattnode and the 2% of the TED could be because the TED only reads the voltage of one phase.
Brett
I suspect the major difference would be in the quality (accuracy) of the current sensor. The other issue would be that they would have to measure the phase angle between voltage and current for the power factor and depending on the A-D accuracy and the number of samples per second would affect how fine a difference they can measure. Although, a zero crossing detector on the voltage and current waveforms would be the easier way to detect PF...hmmm..sounds like a project...
Ron I can email you the answer to that question.... easy way to tell if they provide innacurate VA instead of power W, is if they have current sensors only and the measuring unit can run on batteries. With these, the consumption is based on current only. These are generally low cost. I can tell you that the Brultech and the TED do provide true power.
The Brultech monitors will provide the current (amps) for each connected current sensor. The
line voltage data is also available. The voltage is monitored via the wall
transformer's AC secondary which is also used to derrive the true power.
The ECM-1220.H provides both options: log data internaly (non-volatile memory) and
download it at a later date or collect data in real-time. The real-time data is updated
every second. Both options are capable of running at the same time, of course, the
real-time data will be suspended when downloading the logger data.
