Help using SCT-019S


#1

Hi,

I just ordered some CTs and noticed that I made a mistake and ordered SCT-019S units instead of SCT-019-000. They seem to be 33mA for 300A instead of 33ma for 200A. The turns ration is 9090:1 instead of 6000:1. They have already been shipped and the folks at YHDC (through aliexpress) were kind enough to add 3.5mm jacks to the wires.

The question is: What burden resistor should I use? This is a North American setup with a main breaker rated at 200A (split phase). There will be one SCT-019S on each of the 120v leads. Should I leave the original smd burden resistor or change it to get better precision in the 0-200A range? 100A?

Thanks for the help!
Maxime


#2

I would not recommend changing the SMT burden resistor. At the end of the day, there is the possibility of collateral damage, and you are still using a 9000 turns CT. Ive not tested that model, but I think the Magnalab SCT0750 is just a rebranded YHDC SCT019 with 7500:1 ratio and the trend. going from the SCT019 to the SCT0750 is that phase shift increases, particularly at the low end where you will be operating. It’s not possible to come up with a single correction factor in that range because the curve is pretty steep:

image

I expect at 9000:1 it will only get worse. I think your best bet is to just configure it straight up as a generic:

What you use for phase can be informed somewhat by the phase plot above for the SCT0750. 2.5 for that CT would be on target at about 15 amps ~ 1800Watts. By contrast, here’s the phase shift of the ECS24200 4000 turns CT that I’ll be offering:

You can see that it’s an order of magnitude better. So try the generic settings and see how it works. I think you’ll get reasonable results, but may not be able to get to 1%.


#3

So, would you recommend ordering some SCT-019-000? Or is gain in precision too small to bother?

Thanks a lot!


#4

I’m a big believer in empirical results. You have the CTs on the way, and they will work with the generic settings that I posted. Hook it up and compare the results to your meter over a few days. They may be fine. Wait till it’s broke before you fix it.

There’s a lot of theory flying around this technology. If I were commissioned with building something for the National Institute of Standards, I might be more concerned. By the 80/20 rule, you quickly and easily get to a point of reasonable accuracy. After that, you start to push up against things that are more difficult to control: Linearity of the CT, variability of the phase shift with load, tolerance of components, accuracy of the meter you are comparing to. You reach a point of diminishing returns. We’re looking for 1%. After that, it gets geometrically more difficult to improve. Those CTs might just work fine.

Perfect is the enemy of good.


#5

Ok, I’ll wait and try using them.

We’ll see what I can get out of them, maybe try to sum all the output of the other circuits using the SCT-013 I have. If only the damn wires weren’t so big…

Also, comparing to the meter outside is not a bad idea.

Thanks for the input, I just have to wait for the package to arrive from China…