3 phase not measuring correctly.

Hello
I have just set up my Flukso and it seem to be working fine on small loads but when I turn on my A/C (big 3 phase unit) the power measurment seem way to high.
Will I have to set the line voltage to a lower value?

I live in Australia and we have 240 volts and our 3 phase is 415 volts. Has the fact when my A/C is on it is using all phases at 415 volt divided by 3. Will that cause it to show a high reading?

Thanks Brad.

michi's picture

It's not just the minute readings. I also see the oscillations in the hour and day views. It's most noticeable when consumption is steady, and the Flukso records these sinusoidal up-and-down patterns that you can see in the image I linked to earlier.

But what is accurate is the consumption, that is, if the Flukso, at the end of the day, reports that I've used X kWh, that figure is accurate and matches what is reported by both my own meter and by the utility company's meter.

Michi.

adamcowin's picture

Well i had my mate who's a sparky here today. We've run cables and connected up everything except for the 3 phase din meter - i'm holding off as there is no point putting it in if it wont work

I'm going to have a ring around some electrical shops tomorrow and see if I can find another 3phase din meter with pulse output - but one that the flukso can read correctly.

I'm not quite sure how to calculate the impulse/kWh to constant number to know what ones would be ok to get?

I know that the single phase din meter I have has a 1000 imp / kWh. Would any one else be able to suggest what other impulse rates per kWh would be ok with the Flukso? - so I can take the pulse rates list with me to the electrical supplier shops tomorrow?

Cheers
Adam

Fluc's picture

The most common is 1000 imp / kWh, but 500 imp and 2000 would be fine also.
Have you think about to put 3 pieces of the single phase din powermeter ? I just don't know if this works to see the global 3 f consumption correctly in comparison with 3 current clamps.

adamcowin's picture

Thanks Fluc. I'll try and see if I can find a 2000/1000/500imp 3 phase meter tomorrow - still have to convince the sparkie to come back again :(

I thought about running 3 x single phase units in place of a 3 phase unit, but i wouldnt have enough inputs for it - this is how it will be input wise, once its finished:

port 1 - 3 phase consumption
port 2 - solar pv generation
port 3 - single phase consumption - yet to be decided which circuit i'll monitor yet. possibly power points or lighting circuit (or possibly spare)
port 4 - gas consumption
port 5 - water consumption
port 6 - not sure what i can actually use this one for in the real world

Hence really needing either the flukso bug to get resolved, or a 3 phase meter that will work with the flukso

Cheers
Adam

bazzle's picture

Might work out cheaper to get a 2nd Flukso and 3 single phase units?
Just an idea ;)

I think if I read it right what Mici said earlier was that the 3 phase unit is ok on daily just not on instantaneous reading with the pulse number?

ygeffens's picture

Would it be possible to get 3 single phase DIN kwh meters and connect them all to one and the same input on the fluxo, just to save on ports, since I have a lot to track and I only want 1 flukso.

I already have the flukso with 3 clamps, but I want to take them out and replace them by a (3) kwh meter(s).

Wizzopa's picture

Or http://www.amazon.de/Eltako-Drehstromzähler-DSZ12DE-3x65A/dp/B0020RLNAC

gebhardm's picture

@ygeffens - the issue with more than one pulser at one flm port is that you will not be able to detect separate pulses if they overlap. That may seem " no issue" as the pulses are just 90ms, but the devil is a squirrel...

ygeffens's picture

@Gebhardm: I see, with just a bit of bad luck they overlap and are only counted as one pulse... Thanks for the warning

@Wizzopa: I'm using exactly that one for my heatpump and it's working great.
I was also thinking of installing one on the main power coming in from the street.
That way I could sent both measurements to PVOutput, subtract the one from the HP from the main one, and I know what the rest of the house is consuming.

Only, and for that I don't know enough about electricity, the HP is 3 phase, and for the rest of the house the 3 phase's are equally devided over all the consumers.
Won't that give a wrong calculation in some way? That was the main reason for maybe using 3 single-phase meters instead of 1 3-phase meter.
I that doesn't make any difference I surely will go for another Eltako DSZ12DE-3x65A.

So, good or bad idea ?

Thanks
Yves

gebhardm's picture

Option: http://swissnox.de/ - I have good experiences with a single-phase one. Prices are different if you use a calibrated one for accounting purposes or a "non-calibrated" out of own curiosity - you have to be on "your side of the installation" anyway...
"Three phase equal distribution" does not make a difference; otherwise an old Ferraris counter would not work...

adamcowin's picture

So to sum up:

* Run 3 x single phase din kWh meters - not enough inputs when also using water and gas metering

* Run 3 x single phase din kWh meters all connected to single flukso port - Wont work, wont be able to determine which pulse is from which meter and may overlap / pulse be missed

* Purchase another flukso + another 3 single din kWh meters - would work, but an expensive option. especially when I already have a flukso (with a spare port) + a 3 phase din meter ready to go (Just wont get read properly by the flukso) and limited din spaces left in the meter box

* Purchase a different 3 phase din kWh meter that has 2000/1000/500 imp per kWh instead - meaning $200 down the drain for the existing 3 phase din meter I already bought (does 400imp per kWh)

* Hope for the bug in flukso 232 firmware/constants that are not an integer fraction gets fixed - Unsure if this will happen or not, but please?

* Install and use the 3 phase din meter, in knowledge that

So realistically it looks like the only option is to purchase yet another 3 phase din meter that's pulse's convert to a integer fraction that's compatible with the flukso.

@bazzle - Yeah Michi was saying that the end of the day total is correct, but instant / minute / day views are all over the place. This then affects the graphs on the flukso website and pvoutput website.

Whilst its nice to have the end of the day total, one of the main reasons for my flukso purchase was to see the instantaneous consumption to be able to track down power hungry devices and keep an eye on how much we're using

Cheers
Adam

bazzle's picture

Did you notify scnapp to return the meter as unsuitable due to 400 pulse?

michi's picture

Except for a software fix, none of these options are terribly attractive, especially considering the cost. I've been stuck in the same situation for the past year.

Unfortunately, I have not been able to get any kind of statement from Bart as to whether or not a fix will be released. Weeks ago, I sent a private message weeks ago asking again about this, but didn't receive.

It really would be nice to know whether a firmware update for this will or will not be made available. That would be better than the eternal silence.

Cheers,

Michi.

adamcowin's picture

@bazzle - No unfortunately. I did think about doing it straight away, but I held off to see if there was a way around the issue (firmware update from flukso/adjusting the pulse rate on the din meter - which wasnt possible) etc.

Now it's been almost 2 weeks since I received the unit, so I doubt i'd be able to return it. Even though its still in the original box and unused.

@ Michi

I agree that apart from a software fix, none of the other options are attractive. To be honest and i'm quite sad to say this, but i'm really starting to regret buying the flukso...

adamcowin's picture

So far I've got the following up and monitoring - kind of..

- Solar PV generation. Via pulse input from the inverters din meter. Latronics PVe2500 with din kWh meter @ 1000 pulse/kWh
- Water usage. Via pulse input from the flukso water probe on a grenade style (elster V100) water meter @ 5L/Pulse
- Gas usage. Via pulse input from the flukso gas probe on an alinta 750 meter @ 10L/Pulse

The water usage was working, although the flukso dash was telling me that I went from using 0L to almost 80k Lperday under the hour tab (for almost 30 mins continuous) - https://www.flukso.net/sites/default/files/waterreading.jpg + separate post: https://www.flukso.net/content/water-probe-80k-liters-35minutes

The water usage was reporting and working up to 3am yesterday and has since stopped reporting completely - I'm away for work at the moment and wont be able to SSH to the flukso to see if its recieving pulses from the water probe.

The gas probe was working, but has also since stopped working around 2 am this morning

I've only had the solar generation connected to the flukso and monitoring for the last 3 days. With 2 of those 3 being full days of monitoring.

I am still running my old effergy energy monitor on the solar circuit via clamp (Old house only had 1 phase, so i only had 1 clamp for it) which I was using to monitor the solar circuit whilst the flukso was in transit.

Yesterday being the first full day of monitoring, the totals from the flukso and efergy unit were quite different:

Efergy total PV generation 28/1/14 = 17.12 kWh
Flukso total PV generation 28/1/14 = 19.445 kWh (Via pvoutput - http://pvoutput.org/intraday.jsp?id=28541&sid=26142&dt=20140128)

I know that the clamp style monitors are not the most accurate but a 2.3kWh~ difference..? I got the above generation figures from the LCD display on the efergy unit and from the pvoutput total's for the flukso as the flukso dash is giving me the sh!ts.

The din meter on the inverter has a LCD display with the units lifetime total kWh on it. So I recorded the numbers last night and will record the numbers again tonight to see how much the inverter has generated today. Then I'll compare that figure to both what the efergy unit reports and the flukso unit reports to see which one is telling me lies

I still dont have the 3 phase din meter hooked up, so I can't monitor our consumption yet. I told the sparkie to hold off installing it incase I had to replace it with one that will work correctly with the flukso.

So far i've spent a fair bit of coin on getting this monitoring setup (along with lots of time and sweat trenching the water probe out to the meter) and i'm still in the same position as when we didnt have any monitoring gear - that is guessing the usage and consumption.

To semi quote the flukso ssh banner: I may as well just grab our cables and guess.

/end frustrated rant

@ Michi

I'm also going to send Bart an email / pm and see if I can get any response as to if this flukso bug will get fixed at all / any time soon. It would be great to know if its going to fixed or if i'll have to find another solution myself..

icarus75's picture

@michi The 'aliasing bug' is caused by the counter fraction not being passed on between the AVR on the sensor board and the MIPS on the main board. I am currently evaluating a new internal communication bus for the FLM. Once this is in place the messaging protocol between sensor and main board will be updated as well. And I'll make sure to get that aliasing bug fixed during that update. So the next firmware release note will mention the fixing of the aliasing bug. I'm sorry that this is taking some time. But I want to do this properly.

Wizzopa's picture

@adamcowin,

What kind of KWh meter does the power company uses

Is there by any change a blinking led on this meter.

Wizzopa's picture

typo chance

michi's picture

@Bart: Thanks for the reply! Look, I understand that it might be awkward and not easy to fix this. That's perfectly OK. Now that you have made a statement that (they way I read it, correct me if I misunderstood) this will be fixed in the firmware update after the next one, that's totally cool with me. I'm prepared to wait, no problem, no that the silence has ended.

Cheers,

Michi.

PS: I *do* hope that I won't have to upgrade to a Fluxo 2b though for the fix to work!

icarus75's picture

@michi The aliasing bug will be fixed for both the A and B versions in the next firmware release.

michi's picture

@Bart: Bloody awesome, thank you for the confirmation! :-)

Cheers,

Michi.

bazzle's picture

@Adam

Looks like you can keep the 3 phase kw/h meter now :)

adamcowin's picture

@Bart

To quote Michi - Bloody awesome!

Thanks for the reply and update on the aliasing bug. Michi pretty much summed it up in his comment above, We understand that its probably quite a complicated fix and no problems waiting for it to be resolved.

Thanks again

Cheers
Adam

adamcowin's picture

@Bazzle

Indeed it does - much happier with that, then trying to return the existing din/finding a different one that would work.

I've been monitoring the inverter (has din meter with lcd screen + pulse output) and also the efergy clamp meter that i have clamped to the solar phase. Whilst i knew that the clamp style meters were accurate I was not sure if the flukso was 100% accurate as well (not the device it self, more my configuration of it)

To find out which (flukso or efergy) was giving me incorrect readings, i've taken a reading off the inverters lcd din meter each night and then compared it to the flukso / pvoutput and the efergy (lcd screen). Hows this:

28/1/14:
Inverter DIN generation: Reading not taken
Efergy reported generation: 17.12kWh
Flukso reported generation: 19.445kWh

29/1/14:
Inverter DIN generation: 17.95kWh
Effergy reported generation: 16.57
Flukso reported generation: 17.960kWh

30/1/14
Inverter DIN reported: 17.34kWh
Effergy reported generation: 15.89kWh
Flukso reported generation: 17.325

So i'll eat my words above about the flukso not working and that I wasnt able to monitor anything. The flukso is proving itself to me as being accurate and reliable. Whilst the water/gas probes are still not working, they have worked and i'll sort them out when i get home.

I cant wait until the 3 phase din meter gets hooked up and I can then monitor the consumption and also the instantaneous consumption once the aliasing bug gets worked out - i'm sure i'll prob be horrified when I find out how much power we are using, both currently and in a unit/day sense. But thats why I purchased the flukso, so we can change our usage and reduce our bills.

Also whilst i'm just rambling on..

@Michi

With the aliasing bug causing the instantaneous readings to be out, have you still been able to work out roughly what your using and, when a device is turned off roughly how much the usage has dropped?

Cheers
Adam

michi's picture

@Adam: I can sort of eyeball the consumption if I know that consumption overall is steady while I'm turning a device on and off. You basically have to read the difference between the tops of the square waves before and after turning something on or off. Note that this is not terribly accurate (even if there weren't any of these fluctuations) unless the device being turned on or off make for a sizeable percentage of the total (say, 5% or more).

I would recommend to buy one of those simple consumption meters that have a power point on one side, and a power plug on the other, with a display that shows consumption in watts, amps, etc. They work reliably and accurately, especially for things that draw only a small amount of current. (They are also accurate with phantom loads.) That works better than trying to identify consumption of individual devices with the Flukso.

Cheers,

Michi.

bazzle's picture

To see what each device uses I use the Flukso Android free app on this site ...somewhere..
I put it into an old unused HTC phone I had lying around.
Note: the Pulse meters have been proven .. as you found out... to be more accurate than current clamps.
Prob as they have the available voltage and current to determine wattage .
CC only has current flow and a voltage preset by the user.

adamcowin's picture

Ok well i've got the 3 phase din meter connected now - but im not sure what to put as the constant?

Anyone know what the constant would be for a 400imp/kWh din meter? - 0.04 ?

Cheers
Adam

Fluc's picture

1000imp/kWh = MC 1 because 1 pulse is 1Wh
2000imp/kWh = MC 0,5 =
500imp/kWh = MC 2
So 400 imp/kWh = MC 2,5
The problem i think is with 1600imp/kWh = MC 0,625

icarus75's picture

Please sign up for the r241 alpha release at this forum post.