Is there a noticeable decrease in resolution of sampled data by using a 50A CT on a 20A circuit? Would I be better off with a 20A CT, or does it not really make any difference?
The accuracy of the 50A is fine.
I must admit I was very worried about this too, but after reading several comments by overeasy on this site I took the plunge and got some 50A CTs with my order, most of the circuits I will monitor would max out below 20A.
I just finished a bench testing session after receiving it and was amazed how accurately it reads a load as little as 50W through the 50A CT that is capable of measuring 12kW at full current capacity, it is as accurate as I can cross check with my meter and matches the readings from a small commercial power meter that I have also!
It does defy what common sense tells you about matching the sensor size to what you are measuring, but i suspect its due to fact that its all transformer based as well as very accurate ADC in the IotaWatt hardware, but whatever it works way beyond my expectations!
I was also worried about this as it hadn’t dawned on me that this may be an issue until after I placed my order and looked at the H040 datasheet that states:
RATED CURRENT 5A, 20A, 30A, 40A, 50A, 60A
Current Range 10% - 120% of rated current
@markw I’m glad you did some testing to alleviate my concern. Thanks! I’m going to try to use these on 15a circuits as well.
A friend of mine is planning to order a kit when he told me about it. I suggested that he could loop his 15 A breaker wire through the 50 A coil 3 times, more if it was always a light load. Does the software allow you to put in a scaling factor if someone does this?
That is completely unnecessary. The 50A CTs are linear to 0.5%. Even the AccuCT 100A would be fine on a 20A circuit. Consider that my mains are measured with a 200A CT yet the average current is 3.6Amps for the past 4 weeks.
That’s 1.8% of capacity, yet my kWh matches my meter to less than 1%. The 50A CTs can measure the power of a 10W LED bulb to its nameplate rating.
But the answer to your question is yes, you can create outputs that show the measured values divided by your turns multiplier.