Archived Forum

1. The voltage input is calibrated for the UK standard AC-AC adapter, which give approx 11 V output for a 240 V input. As your adapter appears to be giving approximately the same voltage out for 120 V input, you will indeed need to alter the voltage calibration constant.

2. For US users, I normally recommend the Magnelab range WITHOUT a burden resistor. Although these are specified for 0.333 V output, Magnelab assure me that with a suitable choice of burden resistor, they are suitable for a 1 V approx output that the emonTx V3 requires, provided that they are derated to 1/3 current. I think the SCT-1250-000 should be OK. You need a 75 Ω burden resistor for 100 A max, which means you need to remove the on-board SMT burden resistors for each current input and fit a wire-ended component in the holes provided. It's the resistor closest to the jack socket.

Thanks Robert,

I was going to order a couple from here (http://www.aimdynamics.com/SCT-1250-Split-Core-Current-Sensor-333mV-Outp...) and noticed they have a "0-1v output option" listed here http://www.aimdynamics.com/product-p/burden-1v.htm

Would this eliminate the need to change the resistor on the emonTX?

Thanks again

M

I'm suspicious. Even though it won't affect you at 100 A, Magnelab say they need derating for 1 V rms out, yet this reseller /  stockist still offers up to 800 A.  If you have the 1 V output option, you would still need to remove the on-board burden as there is clearly a burden resistor inside the CT housing, and you must not have two.

maybe i will order a pair and see how they do after i remove the onboard burden.  I'm guessing i need to add something after removing the onboard burden, just a piece of wire between the holes?

I have another energy monitor (old TED) that i can probably compare with when testing calibration

your up late tonight Robert - watching the vote?

"I'm guessing i need to add something after removing the onboard burden, just a piece of wire between the holes?"

No, but if you like, you can solder in a 3.3 GΩ wire-ended piece of air

[Yes, I was up late; no, I wasn't watching the vote. I'm just nocturnal like my cat.]

Thanks Robert.

I ordered - i'll report back when I get them.

Madas - I'm also in the US and just got my EmonTX3 setup with a single CT and a couple of the 100A ones on the way.  I just ordered the Magnelab SCT-1250-150 Split Core Current to monitor my mains line. (specs: http://www.powermeterstore.com/p2097/mag_sct-1250-150_split_core_ct_150a...) 

How did you end up connecting your CT from your main?   What did you end up needing to modify in your EmonTx?

aaronmds:

Read the thread again carefully, it looks to me as if you've ordered the wrong part. Yours has a 0.333 V output, which isn't enough to fully use the input range of the emonTx.
I'm told that a 1 V output is available as an option, which is much better. This might be direct from Magnelab only, it might not be available through stockists.

I'm suspicious. Even though it won't affect you at 100 A, Magnelab say they need derating for 1 V rms out, yet this reseller /  stockist still offers up to 800 A

Aim Dynamics is the reseller one gets directed to from the Magnelab site.

http://www.magnelab.com/catalogue

Something fishy going on, perhaps? Like maybe someone at Magnelab is having a memory lapse?

Bill, my info came from Jeff Lawson, V.P. of Sales and Support, Magnelab, Inc., 600 Weaver Park Rd, Longmont CO 80501.

It's all down to the VA rating. If you want 3 x the output voltage, you have to accept a limit of 1/3 the secondary current which implies 1/3 the primary current. But to confuse it just a little, Magnelab devices are good from 10% to 130% of the rated current, so you're actually down to rating it at 43.3% of the published value.

Could it be that the reseller isn't quite on the ball? I've just gone back to Jeff asking for more details of their range - I've only got credible data (i.e. direct from him) for the 1250 split-core ones - and clarification on the rating question. i.e. Is a 1250 with the 1 V output option really still rated to 600 A - which I don't believe but a stockist who hasn't a clue might - or 200 A (and good to 260 A) ?

Could it be that the reseller isn't quite on the ball?

That's what I was curious about. Since the Magnelab site has a link on it straight to Aim Dynamics, One would think the info on the AD site was good. Perhaps not...

My experience from the manufacturer's point of view is.... it depends!

Sounds like that's the case here. It'll be interesting to hear what Jeff has to say about the 600A CT w/the 1V option.

Seems like there are several threads on here that have bits & pieces of what is needed to configure a US implementation. I'm setting mine up now based on the emonTX Shield SMT, so the discussion is timely. 

I've noticed at least one commercial vendor (can't find the link at the moment), and also this DIY installation http://www.desert-home.com/p/test-html-code.html recommending using 2 CTs wired in series, one around each leg of the incoming power feed. This seems attractive to me since it would only use up 1 of the 4 CT inputs on the Shield, leaving the other 3 to monitor some key circuits, such as a heat pump, an air conditioner, and an electric vehicle charger. At the whole-house level, I don't really need to know how much was used on L1 vs. L2, just need the total. 

What would be the appropriate calculations to size the proper CTs, burden resistors, and calibration constants in this series-CT configuration? 

Thanks! 

A solution may be forthcoming - I'm thinking of writing a Building Blocks page for US installations.
[Bill Thomson - get ready to duck!]

Putting the CTs in series to get the total could be right or wrong, depending on where the burden resistor or resistors is/are. Remember a CT is a current source, and the burden resistor turns that current into a voltage. Therefore, you can put two CTs in parallel and use a single burden resistor, or you can put a CT together with its burden resistor - external or internal - in series with a second similarly connected to its burden. In both cases, the voltage that the input sees represents the total current in the two legs.

You need to look at the worst case to calculate the resistors, bearing in mind that's when both carry maximum current and there's nil current in the neutral. Let's say that is 200 A per leg, and the CT is 200 A : 25 mA (no internal burden). Let's say (as you mention it) that you're using a Shield at 5 V.
The max rms input voltage is 5 / 2 / √2 / 1.06 = 1.667 V (the 6% is to allow for component tolerances).
Therefore, if you're adding currents, the max current is 200 A in the primary, 25 mA in each secondary, or 50 mA total.
Hence the burden is 1.6667 V / 50 mA = 33.33 Ω, so a 33 Ω 1% resistor should be fine (we've already taken it's tolerance into account in the 6%).
If you're adding voltages, the max current is still 200 A in the primary, 25 mA in each secondary. But you only want half the voltage, hence the burden is 0.8333 V / 25 mA, which comes out as the same (unsurprising really - but if your CTs aren't quite equal, separate burdens allow you to tweak the burden value with additional high-value resistors in series).

You can read about calibration theory in Building Blocks to find out what your current calibration constant is - once you know which CT and burden you're using. The simplest way is to regard the assembly of CTs & burden(s) as a CT with a built-in burden (voltage output).

A solution may be forthcoming - I'm thinking of writing a Building Blocks page for US installations.
[Bill Thomson - get ready to duck!]

As Scotty would say to Kirk, "Aye, Captain!"  ;-)

Robert, count me as one who is looking forward to the US (should really be North America) Building Block. If you need any help fine-tuning the language, I'd be ecstatic to help.

Robert,

I got the following response from Jeff Lawson at Magnelab today: 

Patrick, the SCT-1250 must be burdened in order to achieve a voltage output.  The reason Magnelab produces the SCT-1250-000 is to allow the customer to determine the burden and the voltage output.  Magnelab can custom design a resistive burden to output a particular voltage at a specific current.  Provide your current rating and voltage output and we can design accordingly.

Please let me know if I can be of further help or assistance!

Jeff Lawson

V.P. of Sales and Support

Does this contradict what you thought about the SCT-1250-000?

Update: A follow up email got this response:

If you know what input the meter will accept, in this case 1.0Vac at rated current, Magnelab will custom design a SCT-2000-600 which will output 1.0Vac at 600A.

I think we have all the pertinent information about the items in the Magnelab range that are of interest. The 1 V output (a custom option available through their distributors) is a little lower than we normally aim for, but perfectly adequate for most purposes. Bill Thomson pointed me at their White Papers and there they make it clear that saturation is not an issue even at well over 1 V output.

The advantage of having Magnelab produce a custom unit is it will be calibrated to within 1%, but if you need accurate measurements that won't negate the need to calibrate the emonTx itself because there are many other factors that also affect the overall calibration. There's a Building Blocks article about that.

I guess Jeff's email confused me some. Do I need to custom order the SCT-1250-000 and specify a 1V output?