Pre-Purchase Question

Karl521 · 2018-01-04T21:16:51.088Z

I mean if the whole house is flooded with Wifi so everywhere has a good signal, could the Tx talk to the Pi using the home wifi connection ?

Robert.Wall · 2018-01-04T21:25:38.413Z

Yes, but then you’ll need to add an ESP8266 module and a 5 V d.c. power supply to your emonTx.

Karl521 · 2018-01-04T21:40:59.304Z

Brilliant, I’ll get the Pi first then add the Tx after, thanks.

Robert.Wall · 2018-01-04T22:06:12.957Z

Your emonPi will give you (possibly after doing some maths, depending on exactly where you put the c.t’s) nett grid power and direction, and Economy 7 power; your grid meter will give you imported energy, and your generation meter will give you PV energy generated. You’ll have everything you want, just not in the same place and the same units.

Karl521 · 2018-01-04T22:17:57.143Z

I was having a look at the CT sensors, the 100amp ones are spot on for the clamping of the live in the meter cupboard but the solar panel they are much larger than the cable. In fact the connection from the solar to the fuse board is a combined cable, n,l,e so to get access to the live only, I’m going to have to strip the outer cable sheathing to separated the 3 cables, its almost like a household 13amp cable.

Robert.Wall · 2018-01-04T22:32:34.215Z

Karl521:

I’m going to have to strip the outer cable sheathing to separated the 3 cables, its almost like a household 13amp cable.

Hmm. That’s not legal, unless it happens inside an enclosure of some sort. Can you get access to the individual cores inside the inverter, or how and where does it connect into the rest of your system?
As it’s likely to be 16 A max, you could have a much smaller ring core c.t. (but that’s likely to need a small modification to the emonTx). Robin Emley uses uses either the 20A:10mA version here: http://www.yhdc.com/en/product/425/ or this 20A:10mA one Eppep.com - A Trusted Product Reviewer from http://www.ebay.co.uk/itm/281436762683?_trksid=p2059210.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT Both would need a resistor change inside the emonTx (and a plug!).
~~If space is not at a premium or you don’t want to pull a wire out, you could use the c.t. sold for the Iotawatt. All three would need a resistor change inside the emonTx (and a plug!).~~

Karl521 · 2018-01-04T22:47:41.364Z

It’s pretty much standard, inverter to an isolator then into the small lcd that displays the reading you provide for your feed in tariff. From there into the fuse box, the cables that connect the inverter to the lcd box and box to the fuse board are 3 core 13amp cables. I could simple insert a junction box between the fuse board and lcd readout and put the ct around the cable in that. Would the 100amp ct work ok around that small cable ?

Robert.Wall · 2018-01-04T23:04:18.611Z

Karl521:

the small lcd that displays the reading you provide for your feed in tariff.

That’s your generation meter!

Karl521:

Would the 100amp ct work ok around that small cable ?

Yes, but with limitations: You’re using a 100 A c.t over the bottom 16% of its range - that’s mainly why I suggested a ring-core one that’s both more accurate at low currents and physically smaller. The aperture size (too large) isn’t significant.

Is there room inside the isolator box? Or at the other end inside the fuse box?

Above, I wrote:

Robert.Wall:

you could use the c.t. sold for the Iotawatt

I’ve just noticed that the data sheet gives a questionably low value of burden resistor, so until I can test one of those, its VA rating might not be enough to drive the emonTx input. So best to delete that from the list.

Karl521 · 2018-01-04T23:12:51.861Z

It seems kinda strange that the smallest ct they sell is 100amp surely most domestic pv setups don’t output anyway near that, yet the pi solar kit includes two of them.

I also wonder how accurute clamp meters are in general, I know my fluke one isn’t as accurate as an inline reading. Aren’t there any sensors that you can fit inline with the live feed, both at the panels and the input to your fuse board from the meter ?

Robert.Wall · 2018-01-05T07:39:14.742Z

I’m only a moderator here, so I can’t answer for what Glyn & Trystan decide to stock.
When I looked around at meters in general a few years ago, clamp ammeters were broadly in the same accuracy league as any other in the same price range. However, that accuracy will depend on the magnetic circuit (i.e. the jaws) remaining undamaged and fitting closely, to minimise leakage flux.

Karl521 · 2018-01-05T11:47:36.103Z

It’s the tight fitting bit that worries me, I’m guessing as the 100amp ones are sold in the kit they must be what most people are using, therefore is it just my panels that use a 3core cable from the generation meter to the fuse board ? Or is that common practise.

Karl521 · 2018-01-05T11:53:31.670Z

I notice you have tested the 100amp CT Robert at the lower ampage range, what size cable was you using ?

stuart · 2018-01-05T12:17:48.000Z

@Karl521 rather than using a CT clamp to read the solar PV output - you could use the generation meters pulse output (or infrared output if you want to get technical).

Additionally, what brand is the solar inverter - its often better to get accurate readings directly from this over serial or bluetooth etc.

Robert.Wall · 2018-01-05T12:17:54.649Z

The SCT-013-000 fits the UK 100 A (25 mm²) single core PVC/PVC meter tails. Being a loose fit on a 2.5 or 4 mm² core isn’t a problem. There should be a tiny effect if the wire is not centred and normal to the plane of the core: I could not see it, let alone measure it. (If you think Biot-Savart’s Law, the effect is a secondary one related to the amount of leakage flux that escapes.) You’ll get far more perturbation from an adjacent current-carrying cable.

I can’t comment on how common your arrangement is, I think it will depend very much on the circumstances of the individual installation, and on the preference of the contractor who did it.

You can use the 100 A c.t. on the PV infeed, and many people do. You might not be able to because your problem as I understand it is finding somewhere to put it on a cable, that’s why I suggested a much smaller ring core c.t. The construction of the c.t. and the lower rated current both contribute to better accuracy, which is a bonus.

Karl521:

what size cable was you using ?

You don’t want to know! My test rig uses a 20-turn coil of (I think) 0.75 mm², which gets warm at 5 A, and I use as many turns as necessary to give me the current I want. Or when I want to go to 250 A, a 50-turn coil of enamelled copper wire again carrying up to 5 A. Having tried (and failed) to measure an effect due to positioning, I don’t try any more.

Karl521 · 2018-01-19T15:20:57.255Z

Thanks for the info guys.

I spoke to one of the guys at their office, he was very helpful the only thing he mentioned which I don’t think will be an issue is delay on the 2nd device transmitting data from the solar panels. Although he mentioned the overall data will be good, if I wanted a real time reading it may not be accurate. I guess if I wanted a daily plot though this would be good enough ?

Robert.Wall · 2018-01-19T15:27:55.774Z

Karl521:

I spoke to one of the guys at their office

I’ve got crystal ball failure. Who and which office? That makes no sense to me. Are you talking about the emonTx and the emonPi not taking their respective readings synchronously?

Karl521 · 2018-01-19T15:43:02.431Z

The number on the open energy site, here. The tx sending reads to the pi is what I was told.

Robert.Wall · 2018-01-19T15:59:40.926Z

OK. If both work at the default rate, which is one report every 10 s, then one set of measurements will be up to 10 s adrift compared to your reference. If things change quickly and you’re doing maths comparing readings from two sources, then you will get an error, one that’s going to be proportional to the time difference and the rate of change. So if you only want daily totals, or even hourly or half-hourly ones, I don’t think it will be significant. You’ll need to look at what you’re measuring and the amount and the rate of change, and try to evaluate the effect it will have for you.

Karl521 · 2018-01-19T16:25:52.795Z

Thanks Robert, that is what I thought.

Robert.Wall · 2018-01-29T14:40:56.915Z

A post was split to a new topic: Pre-purchase question - metering gas & electricity