What are the most accurate ways to calibrate ct clamp values?

I have a CT clamp hooked up to a breaker wire with esphome. I used a lamp with a kasa smart switch with energy monitoring to calibrate but it didnt seem very accurate. I wasnt sure if it mattered to calibnrate on the same wire you’d be using the clamp on or if there’s a more accurate way to calibrate? Tahnks!

Did you use a heat bulb aka incandescent :bulb::fire:? Only that one has a power factor of exactly :one: :raised_hands:

I’ve had poor results using CT clamps with ESPHome. I used a 100W lightbulb to calibrate and honestly the results were so unpredictable it never made it past my bench; no amount of filters seemed to keep it from giving pretty erroneous results somewhat regularly. I even tried a dedicated ADC thinking it was due to the onboard ADC on the ESP and still had similar results. The only thing I could think of as a source of problems is that I never moved past a solderless breadboard.

I had planned to add some CTs to some sub circuits to complement the PZEM-004T’s I have on my main lines coming in. I know this doesn’t really answer your question, but curious if I’m an outlier or not.

Heat bulb? It’s important to use incandescent bulb as all others like CFL, LED, etc. have a power factor lower than :one: :bulb:

It just doesn’t make sense to calibrate current meters with other than a resistive load (PF=1) :man_shrugging:

Correct, a 100W incandescent.

1 Like

how do you even get a hold of one? no one sells these anymore

Dug it out specifically for calibrating the CT. :slight_smile:

I still have a handful of incandescents from over a decade ago that I either replaced but didn’t throw out or are in a fixture that gets essentially zero use.

You want a load in the middle range of your CT maximum, 100w will likely be too small.

I use a toasted sandwich maker (1000w resistive element) and a mains power meter like this to calibrate my 10A smart plugs:

https://www.aliexpress.com/item/32695813719.html

Should work for 10A CTs. Might be a bit low for 50A and 100A CTs.

Hi, sorry for joining but I also am trying to calibrate and learning at the same moment yaml, read and in principle just everything. :smiley:

I have a 150 watt bulb available for calibration but realise this is pretty low in the scales of a CT-clamp.
But luckily for me I have just only a 1v-5a CT-clamp. When I try to calibrate I notice due to sinus wave the voltage jumps all over the place and found this value not very helpful. So tried a different approach:
In the perfect world my yaml (for a 1v-5a) will look like (please correct me when i am wrong):

- platform: ct_clamp
  sensor: ads_ct1
  name: "ampere_ct1"
  accuracy_decimals: 5  
  id: ampere_ct1
  update_interval: 10s
  device_class: current
  unit_of_measurement: A  
  filters:
  - calibrate_linear:
    - 0 -> 0
    - 1 -> 5

However with this yaml the bulb will give me a value of roundish 166.9 W, which makes me wonder if this bulb, roundish 20 years old, is still suitable for calibration. If I assume it is and after a small correction I am now roundish 150 watt but it jumps a bit partly because my voltage is not stable too and works with a 30 sec interval from HA. Oh what the heck.

First: A current meter can’t give power readings. It will only give you current readings. :man_shrugging:

As you already found out the AC voltage isn’t constant in reality (just in the books) so even if you have a load with a power factor very close to 1 (like a good old heat buld :bulb::fire:) you will not be able to calculate real power from your current readings. :warning:

Long story short is to get a power meter like a pzem004tv3 (~ $10) :money_with_wings:

I understand your remark that a CT-clamp only give you a current and anything else derived from this current is just in-accurate were a complete product like a pzem004tv3 is just better.

Returning back to the calibration part I have the following question:
With the lack of better resources I try to calibrate with a 150 watt lamp. Given the fact this lamp is likely over 20 years old (and never been used because it will more than double my base-line of power-usage) it was original made for the range 220-230 volt (as stated on the box).

Wat will be the normal current of this old lamp?

  • 150W / ~225V = 0,669A given a new watt of ~235V * 0,669A = 157W ?
  • 150W / ~235V = 0,638A ?

Hope this question is not too stupid :grin:

You could assume the 150W as accurate and constant. If you have a multimeter you might be able to actually measure it’s resistence (if you have one with such a feature). If you know or can measure the grid voltage, you can calculate the current.

Or, plug it in somewhere and measure the voltage and current with a multimeter.

In the end, grid voltage can vary though. In a country with a solid network (unlike mine) it should be pretty constant.

Why would you use two different voltages in these calculations?

  • 150W / ~225V = 0,669A given a new watt of ~235V * 0,669A = 157W ?

The voltage on the box is just for compatibility. The voltage is dictated by the source, not the device. So, your voltage used in the calculation must be what you know or can measure on your grid. In other words, voltage is a given and the resistance causes a certain current to flow.

1 Like

What should a “normal” current be? Like a theoretical one? What should it be good for when it can’t be applied in real life scenarios? :thinking:

I think you can call yourself already lucky that your measurements are somewhat in the ball park of what’s written on the package. Current meters are current meters and current meters are not power meters. :bulb:

What is your use case at all for your current meter @WJ4IoT? Are you planning using it for a resistive load or even apply it to a something with a power factor lower than :one:?

The production of my solar system (SolarEdge) gives my only on a 15 minute interval data. For a few months I used a Shelly 1PM to measure my solar system. I think it is safe to say that these tiny devices do not like being attached to a a solid core wire and perhaps due to friction the reset button was already slightly pressed. Anyway I experienced some unwanted behavior and, not completely sure about this, sometimes triggered by a restart of HA. Anyway dismantled Shelly and looking for a non-invasive method (to reduce the change my solar system goes down and stays down).

Thought I found my solution in a CT-clamp but I understand this will only give the current and anything else is a just bit make-believe.

Calibration is a bit of a issue as discussed above; can use the Tasmota POW method with my 150w bulb but was just wondering if for this for myt lamp 150 watt is still my starting point.

Anyway the results should not be 100% perfect, I realise this is just an illusion

Use proper energy meter like Shelly EM with current clamp

1 Like

That can be actually even be tricky for proper power/energy meters :grimacing:

:100: