My refrigerator was working about a month through ECS1050, 50A CT and according to that sensor most time consumed about 150 W.
Today I have replaced CT with SCT013, 50A model and now it is steadily reads about 100 W.
I have changed a CT type to SCT013 in IoTaWatt Setup->Inputs settings.
I don’t know actual refrigerator power during “normal” cycle.
IoTaWatt displayed 150 W before and 100 W now.
The documentation says it’s “6.5A” but so much (780 W) it takes only during rare defrosting cycle.
Is there any way programmatically to correct the value?
Hi Ben,
The problem is that not all SCT013’s are alike. The SCT013-000 (100A:50mA) is the only true current transformer. That is to say that it generates a current that is proportional to the primary current on the wire it is clamped around.
The other SCT013’s - 050, 030, 005 etc. are actually voltage transformers. They have an internal burden resistor that causes them to produce a voltage that is proportional to the primary current.
IoTaWatt has an internal burden resistor on each input channel that combines with the external burden and skews the reading. If the 150W and 100W are the numbers you are seeing, you would have 20 Ohm burden resistors in the IoTaWatt. The SCT013-050 has 37.2 Ohm resistors. So combining those two you get an effective 13 Ohms.
I would recommend you get rid of these voltage type CTs and either get some more ECS1050s or some genuine SCT013-000. But if you are adamant about using the 050s then you can kludge it as follows:
In the IoTaWatt setup, change the value of the burden resistor for that channel to 13. Then reconfigure the CT as “generic” with turns = 1860 and shift of about 3.
While this works OK, it only provides about 1/3 of the resolution that you would get from an ECS1050 over a 50A range. That’s significant when you are measuring 150W. 1.25A (150W @ 120V) is less than 1% of the range when configured this way. With an ECS1050 it’s 2.5% of the range. I guess I’m trying to say don’t expect a perfect match with the ECS1050. You’ll be in the ballpark.
You should see the wattage return to close to 150.
Thank you for detailed and prompt reply.
Since I saw SCT013 in IoTaWatt settings list, I was thinking they are compatible.
In your store six extra ECS1050 would cost me $9*6= $54+ $8= $62 while on AliExpress the SCT013 price is $3.50/item * 6= $21+ $5 shipping= $26, i.e. more than twice cheaper.
I might buy there 100 A version for the same price but considered 50A have better sensitivity on lines with 20-30 A breakers.
P.S.
You recommendations work.
I’ve got 144 W with SCT013 vs 148 W with ECS1050 before. That’s more than satisfactory for my purpose. Considering it is a single line with permanent load, no problem at all.
Thanks again. IoTaWatt and its technical support both are excellent.
P.P.S.
I’ve read the topic about removal of burden resistor from CT013 and will try that approach as well. Does such modification affect a transformer sensitivity/resolution?
It becomes essentially an SCT013-000 but a little more sensitive because it has a lower turns ratio (1860 vs 2000). Remember to change the burden value in setup back to 20 Ohm.
I’m not pushing my ECS1050s, its just that you drew the comparison in your initial post. If you want to save a few bucks and go Ali-Express, you can get HWCT004 or DL-CT08CL2 for less than $2 each. You will need to add cords and jacks but they work excellent for the 0-40A range.
Your ECS1050s are great, but it’s impossible to buy them in retail.
At least small 20-30 items batch from manufacturer only.
Besides 100% compatibility/accuracy they are more compact and have x1.5 longer cord.
Must admit that I was penny wise and pound foolish. Let my experience to be a warning for others.
glad i’m here. i’ve placed myself in the situation SCT013-050… So far i’m using unmodified SCT and i’ve changed burden resistor settings to 13 ohm as overeasy suggested. Two thongs:
my datasheet of SCT010-050 says 1800 turns instead of 1860. Different manofacturer?
phase shift impart on measurements - how big it is? How accrate should we be? Current measurement using 2 degrees vs 3 degrees will be much different?
I have a mix of CTs, since I have used several different systems over the decades. It really depends on what you are going to do with the data how much it matters. I assume you are NOT using these for revenue/billing anyone. So how much does it matter if the actual measure is 1%, 3%, or 5% off from the “true” value? If it really matters, pay the money to get good senors.
I decided that the actual number doesn’t matter that much, since it isn’t going to have impact on what I do with the data. If my water heater element is using 5500W or 5000W, does it matter? I still am going to want to minimize the amount of time it is on. Where it might matter is if I want to quantify exactly how much I am saving by using a Heat Pump WH. For that, having an understanding of the measurement accuracy is probably good.
The phase measurement is really relevant for loads that are small and not unity power factor (typically motors have non-unity PF). I wouldn’t worry about the phase too much. I guessed at it for some CTs I had that were not on the list and I was happy with the numbers. Eventually I got the CTs characterized and I was off by a couple of degrees. It would not have impacted any decisions I made on the data.
The turns are irrelevant when it’s a 1V type CT. I measured it a few years ago and it was more like 1860 with those samples back then. I imagine YHDC could use any turns with an appropriate resistor. My advice would be to put it on a high current wire with a known CT and play with the turns to calibrate. If the PF is different as well, you can play with the phase to try to calibrate that too.