Energy monitoring - going beyond smart meters...

For the sake of others reading this in later weeks/months, I’m also posting questions about the emonPi Energy Monitor over on the Open Energy Monitor site. If it’s related to the hardware build, internal software, configuration or errors, then I’ll be discussing those details with the experts from the existing user-base.

That seems to be the best approach, allowing us here to concentrate more on the use of the equipment and the data it delivers to us.

A great little aside this one, @Transparent. I’d heard the OVO Foundation had supported similar projects in the past so I got in touch with the team to check if there are any current projects going on in this area:

OVO Foundation's mission is to give kids and young people all over the world a greener, fairer future. OVO staff have previously supported a few electronics/ programming initiatives (including CoderDojo) but the most relevant of the programmes supported by the Foundation is probably Climate Changers, and specifically the work they're doing with Energy Sparks. They teach school students how they can help to make their schools more energy efficient. And once they’ve got this new, energy-saving know-how, they can pass it on to their friends and families too. Part of their activities includes setting up clubs, so the equipment may be of interest to them. If not, OVO Foundation has lots of connections to school groups who have coding/ electronic clubs if that's of interest.

Keen to share your expertise and get involved in any way? @Transparent, @Simon1D 

I think the Climate Changers organisers/mentors should be invited to the Forum.

If they’ve got secondary school students in their group (like an out-of-school club) then we could provide direction and encouragement for them to build a portable home-energy monitor, using the major parts available from the Open Energy Monitor shop.

The design of this is inherently safe. Current readings are taken using clamps, without interrupting the mains power. The voltage is obtained by inserting a 9vAC transformer-plug into a socket.

Once a school had such a piece of monitoring equipment, it could be loaned out to pupil’s homes for a couple of days at a time whilst it collected the data on an SD card.

So this would yield direct, tangible evidence on electricity usage for each pupil taking part.

They key bits where we could help are:

• where to source the parts
• how to make it fool-proof, such as operating from an internal pre-charged battery
• what sort of frequency to take data readings
• how to extract the data to graphs, and produce meaningful information
• how to compare with smart-meter data

This would very much simplify the highly-technical discussions over on the Open Energy Monitor Forums and enable such a project to be more confidently taken on by teacher/parent organisers.

Erm… @MrPuds - did you pursue with your DNO the possible over-voltage problem which you raised here 5 months ago?

If not, then I guess this current topic is heading in the right direction for you!

Not that I’m paranoid or anything, but the current transformer is something that, in the OEM system, can be unplugged from whatever is reading it. Of course I was careful to follow their instructions, which are very emphatic about making sure the current transformer remains plugged in when clamping/unclamping it around the supply cable. Otherwise there is a risk of generating a high voltage at the unplugged connector. I read that it has a zener diode to keep this high voltage at a safe level, but I wasn’t inclined to test this feature.

Would it make sense, in a school project, to adapt this detail so the current clamp remained connected (“plugged in”) all the time?

PS, if my understanding is correct, a “current” transformer like this steps the mains supply current down to the level of mA, and so must involve a turns ratio which is in the 1000s. Given the chance (i.e. with the secondary circuit open) the same transformer would step the voltage up by the same factor from a starting voltage which is the mains supply voltage...

As the emonPi is Open Source we can check the design to satisfy ourselves on the matter of safety.

The current-clamp inputs are on the emonPi shield for which the circuit diagram is available on GitHub.

Here’s the input stage from a current-clamp jack socket to which I’ve added an arrow pointing to the zener diode:

The zener will conduct if the voltage at that point exceeds 3.3v

The current clamps used by the emonPi will pass 50mA along the connecting wire for 100A peak mains current. That’s pretty small!

If the wire wasn’t connected to the emonPi board when the current clamp was placed around the mains cable, there is a small risk of damage to the circuit components when it’s plugged in because the voltage may be substantially greater than 3.3v. But I don’t see how there could be any risk to a human. The current simply isn’t high enough to pass through the skin.

Nor do I see how a current transformer could operate ‘in reverse’ and yield voltages higher than the circuit being sampled. The clamp itself surrounds the cable being tested. This forms an embedded transformer, and the output along the clamp’s wire to the emonPi board comes from the thousands of turns on the secondary.

Current transformer theory is well expounded on the electronoobs site where they also show how to make your own!

Do you see things differently @Simon1D ?

Am I missing a possible safety issue?

Exactly the same clamp is used on my PowerVault Storage Battery which passes all necessary safety checks.

“Current” transformers are transformers which are (approximately) treated as sources of constant current (and the secondary winding is closed to make a circuit so that this current has somewhere to flow).

[Voltage] transformers (with the secondary winding attached to a relatively high resistance, including the possibility of open circuit i.e. infinite resistance) are (approximately) treated as sources of constant voltage.

The ferrite confines the flux (specifically, the flux that links the secondary winding) effectively enough that the same flux links the primary “winding” (I put it in quotes only because it really is a circuit that is closed by the other cable connecting supply and household), and it is fair to use the turns ratio to relate the currents.

From those figures, 100A and 50mA, the turns ratio must be 2000:1

A transformer is a transformer. That’s physics which I understand. This isn’t a matter of the current transformer acting “in reverse”: it’s the same action. Provided the flux linking the primary is the same as the flux linking the secondary, i.e. provided the ferrite does its job, the products of current and voltage in primary and in secondary windings are the same. (This is the point at which it becomes apparent that the ferrite can’t induce the implied current when the voltage gets very large. But, in the absence of the zener diode, it’s not at all obvious to me how large that voltage could get...)

If the secondary winding circuit is left open, then that’s the ratio by which the voltage across the primary winding will be stepped up. That’s what I was trying to say.

I don’t seriously think there is a safety issue (such things couldn’t be sold otherwise).

I’m not sure - do I see things differently to you?

Perhaps it’s just that I’m taking too seriously what would happen if the zener diode weren’t there. Although I would maintain that

“The current simply isn’t high enough to pass through the skin.”

Might indicate a misunderstanding - it is voltage that determines what passes through skin. As you say, there are momentary voltages that are higher than 3.3V. It’s that momentary voltage that determines whether a current will flow. I repeat, I don’t think these things are unsafe.

PS Perhaps it helps to think about the case (without the zener diode) where the secondary winding is closed with a known resistance and that resistance is varied between zero and infinite. (Left as an exercise for the interested reader...)

@Simon1D wrote:

do I see things differently to you?

Well I hope so. Otherwise our collaboration on this project isn’t going to very fruitful!

As I see it, there’s only a certain amount of flux to go round.

If the open-circuit voltage rises towards infinity, the current decreases towards zero. I don’t think anyone’s yet been injured by zero current.

The same is not true of the analogue input to the Atmel chip on the emonPi board. Until that protection zener is fully conducting, there remains the risk of a high voltage frying the silicon gate.

Sorry, I was still editing my post while you replied. To boil it down. If the zener diode were removed, how large do you think that voltage could get?

10V? 100V? more? If not more, why not?

I’ve just checked the specification for the SCT-013 current-clamp on the YHDC datasheet. We have no need to be concerned. It includes a pair of back-to-back diodes to limit the voltage rise!