Archived Forum

i have looked at Home Energy Monitor Kit Bundle 868Mhz i am just wondering if this would facilitate my needs?

The bundle will provide you with a basic monitoring facility and a live display of voltage, current and power. It won't give you the ability to log and record the data (except for example a basic daily total). For that you need to add the emoncms database. That can run locally on a Raspberry Pi (with an RF module), or you can use the online emoncms.org and you'll then need a base station (NanodeRF, OKG or Raspberry Pi with RF module) to connect to the Internet to dispatch the data.

The other concern is how accurately you need to measure your power and energy. The emonTx has only one voltage input, so if the voltages of the other two phases are significantly different to the one being measures, errors will creep in. If that is a concern, it is possible to stack 3 emonTx's, one dedicated to each phase. This will give you the most accurate results.

what i am looking todo is using a raspberry pi as a base setup one monitor to monitor the three currents at the incoming mains. and the to monitor the voltage make three other monitors to monitor the voltage as i'm not likely to find three plug sockets next to each other on different phases. so all in all i would be looking at 4 separate modules all talking to raspberry pi logging live data with a time stamp. im hoping this is possible with a Raspberry Pi and 4 emonTx's with relevant sensors?

I assume you want accurate measurements of real power? If you are content with only apparent power (VA), then what you suggest should work.

If you do need real power, then what you propose won't work because you cannot monitor the voltage and current in separate emonTx's, since there is no mechanism to synchronise the readings (you need to consider how real power is calculated to see why). Even with the stack of 3 emonTx's communicating on the serial bus, each one still measures both voltage and current on one phase. If you cannot arrange to measure the voltage and current of the three phases at the same place - presumably the incomer - then you must measure one voltage and look at the approximate method in software to synthesize the voltages for the other two phases in order to calculate the real power - with the caveat that the voltage hence the power will not be accurate but it will be the best you can do.

However, there might be a way out to measure the voltages using a capacitive pick-up. This would only require a socket for the power supply. I proposed this as a solution to an enquiry some time ago, then it turned out it wasn't needed. Here's a link. You could have three such channels on a small p.c.b external to your emonTx.

could i not use the rtc to sync the data after the fact use the difference in phase between current and voltage to calculate pf and work backwards to get real power? i am really struggling to get a code to just read voltage any how to be honest as everything seems to wont to calculate the current at the same time

Your problem is you need the real power measurement. To get that, you need to know the phase difference between the voltage and current waveforms. In the meter that your energy supplier provides, that's easy as both are in the same instrument. You seem to need to have them separate, for a reason that's not entirely clear to me.

However, pursuing your line of thinking, to be able to measure the voltage and the current synchronously, your clocks need to be super accurate. Any error will be reflected in the accuracy of the real power measurement - it's not so critical where the load is near unity power factor, but when the load power factor deteriorates, it becomes more important.

Let's say you need the phase error to be within 3.6° (just to make the sums easy). The timing on a 50 Hz system works out at 3.6°  / 360 x 20 ms, or 200 µs. So your real time clocks need to stay within 200 µs of each other for ever. Is that realistic? Possibly if the time is received by radio code or GPS, a resounding no otherwise.

If you want to calculate the power after the event, you must transmit every sample, and they come at around 3000 per second, to the central point and then match / interpolate to be able to calculate real power. (I think you're forgetting that even though the voltage wave might be close to a pure sine wave, the current wave will have significant distortion due to non-linear loads, so power factor = cos(phi) is not true.)

the reason i need to collect them from different locations is the original problem as i can not develop a non invasive way to measure the voltage at the same time from all three phases at the mains and it needs to look respectable as it will be used in different places and in other peoples properties and i cant find anything relating to capacitive pick ups on 230v i apologies for my knowledge downfall but this is why im trying to complete this project and unfortunately my lecturer seems to be none the wiser kind regards

A capacitive "Voltage Transformer" is a common piece of equipment on high voltage transmission, but I'm not at all surprised that you cannot find any information of its use at lower voltages. "Transformer" is the wrong name, it is not a transformer in the usual sense and it does not have windings and an iron core. A conventional wound transformer would be very difficult and expensive to make at the voltages where capacitive v.t's are normally found.

A voltage divider does not have to be made with resistors. What I suggested in the link a few posts up was a voltage divider made of capacitors rather than resistors. The "bottom" capacitor is a conventional component on a circuit board, the "top" capacitor is manufactured by you in situ by wrapping copper or aluminium foil around an insulated mains cable. You do not make a metallic connection to the mains cable.  The top capacitor is necessarily of very low value, therefore to make the divider accurate, it needs an amplifier with an extremely high input impedance to buffer the signal into the Arduino. If you have a place where you can clip 3 c.t's onto the cables, you have the ideal place where you can sense the voltage, provided you have access to a some form of power supply to power the electronics.

You will need to manufacture 3 channels as I proposed earlier, and feed the three voltages into three analogue inputs, and the three currents into another three. Whether you can do this with one processor or whether you need three depends on how often you need to measure the power. If you are content with short samples of say half a second every 5 seconds, then one processor should suffice. If you need continuous measurements, you will almost certainly need 3.

You don't say what your lecturer's speciality is - I suggest you talk both to someone with experience in high voltage transmission and to someone with experience in electronics. I've never had anything to do with 275 kV stuff and above, the highest I've ever dealt with is 11 kV.

thanks once again yes i will have access to the main tails to clip my ct's to it i will be manufacturing it on three arduino boards as like your self i didn't think one would have the processing power to take continuous readings. is there any where where i can find more information on the method you are suggesting so that i can figure out how to design and make one for my purpose? unfortunately my lecturers back ground is in project management and rather camp hand gestures neither of which are of any use to myself in this scenario hence why i am on these forums and i am greatly appreciating your help on the matter sorry i didn't reply sooner i have been offline for a fortnight  

I haven't made one, but here Jack Kelly found the capacitor calculations were correct (and there's a circuit diagram in that thread too - designed for a 1 V p-p input, so that needs tweaking). The only snag with the design I proposed is it measures the voltage line-earth, not line-neutral, so if you have a significant neutral-earth voltage, you'd need a 4th to measure the neutral voltage and feed the line and neutral into 3 differential amplifiers (more op.amps - 7 in total) to extract the 3 difference voltages between the 3 lines and neutral.