Archived Forum

You should generally get better accuracy with a ring-core CT.

Voltage is irrelevant, the insulation is provided by the insulation on the cable that is the primary winding. The problem is that most modern CTs are wound for a 0.333 V maximum secondary voltage, as we need about 1 V rms for the emonTx (with a 3.3 V supply - more for the emonTx Shield) the 0.333 V models need to be derated by the same ratio. Therefore for your 40 A duty, you'd need a 120 A version. I usually recommend one of the Magnelab range, the UCT-0750-000 should be suitable, because although it is only rated at 100 A, it is accurate to 130% overload. The ratio is 3000:1, thus for 1 V rms you would need a 75 Ω burden resistor for 1 V rms output (and if you have an emonTx V3, you need to remove the fitted SMT burden resistor and add a wire-ended one in its place - there are holes ready for it.)

Thanks. I guess its UCT-0750-100 you think of.

My original plan of thinking in ring-core CT was to get a smaller CT (physical size), but the Magnelab is a lot bigger than the SCT's that are normally used here on this site.

Not the -100, I really did mean the -000 without the built-in burden resistor. Read again the bit about the (implied) VA rating and voltage output!

A ring-core will indeed normally be smaller than a split-core, (a) because there's no hinge mechanism and (b) because the secondary winding can be distributed around the core instead of being lumped on one limb. If you spot one that might be suitable, it's pretty easy to see whether it will be OK, it simply needs to generate about 1 V rms in its burden resistor at your maximum primary current.

Robert,

I was thinking of doing something similar to the original poster but I can't find solid core CTs for less than £40 including VAT, as I want to monitor 10 circuits that's a lot of cash! The OEM shop sells split CTs for £9.60, why are they so much cheaper than solid core? I can't think that you sell so many of them the price you pay is discounted that much.

Thanks,

BraileTrail

Hi Robert.

Sorry my bad. the first page a saw them on there was only the -100.

Now I've looked at Farnell and other supplies but as BraileTrail mention I can only find physical bigger and more expensive than the split CT's. I've saw Hobut has produced some 3 phase CT that fits a circuit breaker, but don't know if the company exist anymore as their webpage is non-existing

Hello Haden and BraileTrail,

I'm not sure if this would be cost effective or not, but here's a link to a 40A solid core CT that sells for 4.99 USD. (about 3.18 GBP according to Google as of 25 Nov 2014) If the shipping costs are reasonable, and its VA rating isn't exceeded by using a 38R burden, it might be just what you're looking for.

http://www.brultech.com/products/ECM1240/packages/moreInfo/infoMicro40.htm (Details)

http://www.brultech.com/products/ECM1240/packages/currentTransformers.htm (Price)

The specs say the CT has an output of 26.23mA at 40A, so a standard 1% tolerance resistor of 38.3R should give just a bit over 1V RMS output. The VA rating of the device is not given, so I've sent an e-mail to their tech support department asking if that if that particular value burden would cause the device to exceed its VA rating.

I'll post their answer when I get it.

Regards,

Bill

Robin Emley uses a small ring-core within his PV Diverters to measure the diverted power, but he only needs about 10 A or so. The difficulty is that for best resolution, the Arduino-based 3.3 V or 5 V designs need 1 V rms or more, whereas the bulk of readily available CTs have an internal burden and a 0.333 V output. And the other problem is, many if not most have little or no supporting documentation, so it's impossible to evaluate their suitability without getting your hands on one and testing it.
I think this 20 A one with a 150 Ω burden is what he first tried, whether he's still using it I can't tell you.
I can't comment on the pricing, other than to say that if you go for a high precision 'metering' spec, it can be very expensive.