Clamp Sensor Energy Monitoring - ESPHome in Home Assistant

[@KennethLavrsen]

I think someone commented and this is years later. But you cannot measure the electrical power you pay for on an AC circuit with random equipment. Except for heaters and incandencent bulbs all other modern equipment is not a pure resistor. It has a reactive component. It has a power factor less than 1. What you measure is called apparent power with unit VA. You need the real or true power measured in Watts. That is what your meeter charges you for. Look up power factor, apparent power, true power on Google. There are solutions available where you connect a clamps and the phase voltage to the board and then it measures the right thing, or close to the right thing. The power factor on a phase can vary a lot as things turn on and off. In my house I see power factors reported between 0.6 and 0.99. When you are at a low power factor, your measurement can be double of what you pay for. You can only calibrate it away if you measure one specific appliance and is reactance does not change with load.

[@kierancampion2928]

Will, I enjoy your channel a lot but you are wandering into an area that you don’t fully understand. Apart from the inaccuracies in the power measurement as set out by other comments, there is a major safety concern here. The CT clamp used here does not have a burden resistor before the jack plug, it has one on the board, so if the plug was to be removed while measuring it would lead to dangerously high voltages at the jack plug. A burden resistor should always be installed as close as possible to the CT itself. Split core CT’s are available with a resistor installed with the output in millivolts which is a much safer choice in this arrangement . You should consider pointing to the dangers of using this type of arrangement on this video. Safety is a priority.

[@MahlerLab]

SCT013 voltage type has an internal burden resistor, and current type has an internal voltage suppression diode. However, if you are using a model without internal protection, this is indeed a critical point to notice.

[@ashleycawley5816]

@@MahlerLab Thank you for posting this, so I'm using a SCT013 I shouldn't need to worry about that aspect then.

[@ashleycawley5816]

Thanks for sharing this.

[@Sierra-Whisky]

Thank you for explaining! But there are some remarks about the calibration. First of all I wouldn't recommend a microwave because the fluctuations in the current it draws may be quite substantial. A kettle is way better because it's basically just a metal wire heating up. (Although the current may slowly decrease over time) Also it's better to have some kind of energy monitor to use as a reference. The actual energy consumption may differ quite substantially from what's written on the device. Nevertheless, you made me think of some great new ideas 💡

[@Wrenchmonkey1]

Agreed. A resistive load is probably the best option, or AT LEAST using a calibrated clamp meter to confirm the actual power consumption you are correcting for. Also, "230v is the voltage we use here," is putting a LOT of faith in the local utility. You should probably measure the actual voltage you're getting, because that can vary wildly, even within the same block or two of houses. But nevertheless, this video is a great jumping off point, and I'll be using it to add my own monitors to my system. I'll just be a bit more methodical in the calibration.

[@bikeybikebike]

Followed these instructions to start monitoring my house power, thanks. I think I need to adjust the calibration again as the CT clamp is reporting higher power usage than the electricity meter but even if the absolute values aren't accurate being able to see the relative power consumption between days or when different appliances are on is great.

[@ashleycawley5816]

Thanks for sharing this neat idea, I just bought the parts using your links :-) two things I will say: 1) I thought mains voltage here in the UK was more commonly 240v, I know when I measure it within my house it is 240v. Also 2) I thought perhaps a micro-wave isn't the best appliance to measure, not only because of the WiFi interference, but also because the load/demand isn't always on, a Microwave pulsates using higher wattages here and there for a few seconds, then gives a break etc. I think something that may give a bit more of a predictable load (like perhaps a kettle) might be better. Thanks for a great vid though.

[@jitendrasonavane82]

Hi Ashley Cawley, last week i bought Clamp Sensor Energy monitoring device from DFRobot for monitoring my solar panel but what i am facing issue is that its shows some nominal 0.5 kw value in night too. Can you help me out to resolve this.

[@clives555]

Hi Will, thanks for this vid. I'm getting some odd behaviour, I calibrated mine with the immersion heater, known load, and compared results with a Shelly EM and the power measurement was within a few watts. Happy. I turned the immersion off and normal background power use for the house went to 160w ish. however the esp/ct didn't go below 300w. I took the clamp off the mains feed and measured nothing but I'm still getting a reported current and watt usage. log extract below. [11:57:09][D][ct_clamp:037]: 'Measured Current' - Raw AC Value: 0.000A after 3 different samples (15 SPS) [11:57:09][D][sensor:126]: 'Measured Current': Sending state 0.13084 A with 2 decimals of accuracy [11:57:10][D][sensor:126]: 'Measured Power': Sending state 0.03102 kW with 1 decimals of accuracy [11:57:10][D][sensor:126]: 'Total Daily Power': Sending state 143.78992 kWh with 3 decimals of accuracy any ideas why 0A is showing as 0.13084A here's the config used captive_portal: sensor: - platform: ct_clamp sensor: adc_sensor name: "Measured Current" update_interval: 5s id: measured_current filters: - calibrate_linear: # Measured value of 0 maps to 0A - 0 -> 0.000 # Known load: 5.0A - change this! # Value shown in logs: 0.01229A - change this! - 0.040 -> 12.377

[@richardlyd7450]

thanks so much for this video... your clip is the only one I could find that then shown us how to add the entity to Home assistant grid consumption and I've been looking for days! as other examples of Yaml files the entity doesn't show up in energy tab...subbed

[@alexjarvis9636]

Hi Will, thanks for the video. Was wondering why you'd put the clamp around only the live wire rather than just the whole power cable? struggling to get meaningful results out of my attempt at this project.

[@kalamazoowifi3531]

Hi Will, and great video! Is there a way to connect two CT clamps up to a D1 mini? Thanks, Joe

[@mattjdlewis]

Will quick question! Can I put this around my whole home live wire and measure the houses energy consumption?

[@adityapattyeri4378]

Yes you can

[@MahlerLab]

Yes, that's my current setup.

[@MrSilvann]

Doesn't work for me. I calibrated using hair dryer and tuya smart plug which is showing actual current. After calibration when I connected clamp around my life main wire by power supply readings are not match with my portable display which I got with my smart meter. Is not even close. 1220W (display) 1380W (home assistant), 290W (display) 530W (ha). It looks like when load is higher difference is going down and when load is going down - difference is going up. I do not understand that. I do not know how to make this to work.

[@robkendall9920]

Having similar problems , can calibrate it for 1200W and it works fine in that range but when consumption drops to say 200 W it reads twice that value, any more info since you set it up ?

[@MrSilvann]

@@robkendall9920 I gave up. Usually I am using around 200-300W for the most of the day so I calibrate it for that value. Ignoring high spikes.

[@premyslsamuelorsag3759]

You need to use this filter: calibrate_polynomial, which will allow you to calibrate with multiple values. It's a lot better, but still need to many tests, before you will be sure that it is showing well values.

[@benjamink7311]

The schematic of pcb is not clear could you give a higher resolution link please

[@grommet20002000]

Whats was the neatest way to supply the usb power to wemosd1 when its in the meter cupboard

[@WillSurridgeTech]

I have a normal socket outlet in the cupboard so thats what I did...you wont be ale to run a usb cable far so will have to have a socket installed or something

[@ΓιώργοςΜπαντοέρης]

You are on the top. Thank you so much. Greetings from Athens.

[@WillSurridgeTech]

Thank you :)

[@lycakito1814]

That was awesome, great job - thanks for showing, exactly what I wanted.

[@_Miner]

Any ideas why I get negative A / W readings after I enter these calibrated values (@240v)? filters: - calibrate_linear: # Measured value of 0 maps to 0A - 0 -> 0 - 0.01780 -> 4.72 # 1134w microwave (700w) - 0.03571 -> 11.26 # 2702w kettle (2520-3000w)

[@dalethompson7927]

Great video, I have a Shelly that measures the voltage, could you use the voltage measurement instead of theoretical values?

[@coin777]

Yes you can. It will be more accurate then

[@zyghom]

ups, I totally misunderstood your question - of course you can ;-) @coin777 opened my eyes ;-)

[@joehacker1216]

Will how do you make this work with monitoring solar power, have your esp-home code but not reading my power coming in. Have change the setting to .18729 what the log comes up with and 15.8 amps what I read off my meter attached to the each of 2 lines of solar. The home assistant energy say the system is puting out 1.52 kwh and should say 888 kwh coming off solar meter. Please help getting this right.

[@devopssimon]

Is everyone using old versions of ESPHome for these tutorials? I have just come to Home Assistant / ESP Home in the last few days and it does not allow you to configure a board unless you are using a https connection. That is the first difference. There is a link to open ESP Home in using HTTPS which works and allows me to flash my device, but I can then not connect the device to Home Assistant. Do you know if the latest version of ESP home works for you? I don't know if I am doing something fundamentally wrong elsewhere with my configuration.

[@nwzmedia1941]

I think ESPhome website has been updated a lot recently. I've notice that you can install the initial code to the esp8266 directly via the website, and I dont recall you being able to do that when I looked a couple of months ago. Also, there does seem to be some conflicting behaviour when you get onto ESPhome in HA (for example it keeps discovering the temporary hostname that was assigned to the esp8266 even though that doesn't exist anymore), but a refresh of ESPHome seems to fix that.

[@neilsinclair7814]

Thanks for this. I have struggled a bit with this in Home Assistant / ESPHome for the last month. Your video was really useful and well explained. Many thanks

[@AhmedBahgat]

Good video mate. Thanks

[@WillSurridgeTech]

Glad you enjoyed it

[@bully8711]

Hi, great video, all configured and working however my homeassistant values are about 5kwh per day above meter usage. I used the single port motram board with SCT013 100A 50ma CT. Calibrated using a 2000w wallpaper stripper to calibrate. Is there anyway to recalibrate adjust code to get a more accurate reading using a CT . ? Thx in advance

[@amrheing]

Hello Will, I ordered the pieces you mentioned and do all the steps to bring it on. But when I reboot the d1 with the PWM board attached it will not start. without the board all is fine. what's the matter with that? kind regards, an a nice 2022, gerald

[@amrheing]

shame on me.... I have fitted it on the wrong side top :-(

[@ShichaoChenhongyu]

@@amrheing haha. Now do you get it works?

[@amrheing]

Hi, yes it works. But I ordered 20A clamps instead of the 100A pieces mentioned here. So get wrong data.

[@davedee3089]

In general, great video. But like quite a few youtubers, you do sometimes skip over what you might think are easy things like the commenting out of a block of the code. I have no clue how that was done. Also the uploading of the code. 'Simple' if you know what you are doing, and if you do not get errors, but not so easy if you have a few options and you assume that we know what we should see and what we are doing. I have all of the hardware connected and waiting to rock and roll, but I cannot get the code onto the unit. But thanks for making me feel like a should be able to do this.

[@tsarrr]

I 100% agree with the above i was doing a similar project and had issues flashing the firmware. i figure out from another video you need to select manual download, select legacy. It will download an .bin firmware file. then you use a separate flashing tool to install the firmware onto the device.

[@bgable7707]

@@tsarrr @Dave Dee Yes, this unfortunately is are of the learning curve for ESP and HA. Many KZbinrs make is sound like oh, I just did this, nope! It take alot of work! I ALWAYS check and double check my wiring, first!

[@alortiz01]

How to install ESPhome in my Windows?

[@nehajain2974]

Hello I am new to homeassistant. Thanks a lot for your detailed videos.i have learned a lot from your videos. Please suggest an indoor security camera that integrates well with homeassistant. So that I can automate stuff with homeassistant. Thanks in advance.

[@jacekk6913]

IMOU devices supports Onvif protocol - I have 4 of them in my HA setup.

[@---nw9qu]

Anyone have a source for the PCB or something similar in the US? Shipping is very high from UK to US for the one shown here.

[@dougle03]

Don't bother. You'll need a dual monitoring system as you are on split phase power supply (2x110vac, 220vac both phases)

[@nwzmedia1941]

Is anybody using the SCT013-000V CT clamp that gives out voltage between 0~1V ? You have to remove the R1 22ohm burden resistor from the circuit board to use this, but I'm interested to know what raw values from A0 are returned before applying any calibration factor. I'm also trying the Mottram Labs 4 channel current sensor with ESPhome , which would allow me to monitor power used in the house, power generated by solar panels, and power exported back to the grid, and something else, like an electric shower.

[@ernieb4487]

I'm looking at the 4 channel as well to monitor the charging of my car (3 phases). Have you managed to get it working?

[@daveacorn782]

Am also trying to get the 4 channel sensor to work but can't figure it out. Can someone share their set up for the 4 channel version?

[@grahamharrison6125]

Can anyone confirm that I need to remove the 22ohm resister from my Mottramlabs board if using the SCT013-000V (100A) CT clamp. Ive got everything setup and running but no matter how many times I calibrate it the accuracy is nowhere near as soon as the power varies

[@grommet20002000]

In home assistant energy dashboard, why use total daily power Rather than sensor.measured_power,

[@giorgosiotis1557]

Bravo , thank you , you are great.

[@WillSurridgeTech]

Thank you :)

[@ladams5356]

Are you selling it as a kit ?

[@giuseppeusa45]

This project never works for me.

[@giorgosiotis1557]

How could you enable daily energy statistis so easy , i am waiting one day with the same exactly configuration on an esp32 , i have data from esp32, i added 2 utility meters and nothing happens. How did you do that so easy. Because of the total energy power you show i think you dont see all the story. there are so many discussions about the dificulties and you show that as a miracle. Could you be more detailed

[@gelisob]

microwave cycles on and off while running, bad calibration load. Use toaster or radiator or heater fan instead :)

[@forfengeligfaen]

A microwave probably won't draw the maximum rated current all the time. Not great for calibration.

[@BugCrush3r]

This method is totally unaccurate on anything but a purely resistive load like a toaster because you have not considered the power factor in your calculations. Anything with a transformer like a microwave, heatpump, computer, electric motor can have have a considerable power factor causing your measurement of real power to be off by as much as 20%-30%. Calculating the power factor is only possible with real time processing of the current and voltage simultaneously so that you can measure the phase angle difference between them. In reality what you are calculating with this method is not Watts but Volt*Amps or VA power which is a combination of "real power" + "apparent power". VA power cannot be used to calculate energy consumed in kWh because you need to know the power factor in order to calculate the real power portion of your measurement. Typically residential customers are billed only for kWh (kilowatt*hour) which requires you to know what the real power in Watts is. This is very important. Secondly by using a static value for voltage will further introduce a substantial error as the voltage can vary as much as 5%-10%. If someone is serious about home energy monitoring these factors must be taken into account if the goal is to be fairly accurate with the utility main meter. I would recommend using a very cheap pzem-004t power meter which will do all the heavy calculations for you and sends it to your esp32 or esp8266 through a serial connection.

[@milandzuris]

make video for 3ct, for one is so ez.

[@sudheer7595]

0 = 0 hence provided

[@dougle03]

You really should script your videos; they are quite hard to watch and follow when you constantly seem to be grasping for words. This is especially important when you're talking about a subject you clearly don't know very much about. You get a star for trying, but it's knocked off again for failing to talk about mains electrical safety and the requirement for a burden resistor...

[@ashleycawley5816]

A bit harsh? I mean I'm all for a bit more accuracy around the safety side of things, but beyond that you get the general gist of the things discussed. I personally wouldn't describe it as hard to watch.

[@apc4884]

Come on man, the guys providing free educational content. He really shouldn't have to do anything. How about you script them

[@kdelios]

Hold your 🐎🐎... Totally wrong for two reasons: 1. AC Voltage is not constant. It can fluctuate plus minus 6-7% therefore your calculation is off. 2. Power factor (cosφ) rarely is 1 unless is a purely resistive load. You need reactive power in order to calculate cosφ, and voltage in order to calculate Amps. Your calculations are going to 🚽. Get a Shelly EM instead.

[@zyghom]

at home (flat) most of loads are resistive nature. And those, that are not (washing machine motor i.e.) consumes much less than the heater inside it. So with some estimation it could be enough. With regards to voltage: you are right - my HA shows me fluctuation between 227V and 247V (UK)

[@coin777]

Power companies only charge you for the reactive power.

[@kdelios]

@@coin777 Do you know what "reactive power" is? Take a look at Wikipedia to find out. Your statement is not correct. For house use, power companies will not charge you for bad power factor. They charge you for cosφ=1. But this is not always the case. In most houses exist refrigerators, washing machines, and a LOT devices with a power supply AC to DC converter. Those devices drop cosφ to an average of 0.85. Therefore, there is a amount of "unwanted" reactive power going back to the grid causing some issues. On the other hand if you run a facility with 3 phases and a decent power consumption, then power companies will charge (a lot) you for reactive power also (along with your consumption). That's why factories etc use devices to "correct" cosφ and bring it as close and possible to 1.

[@kdelios]

@@zyghom At my house cosφ is average 0.85. It's far from "nature". The reason is I run refrigerator, washing machine, and a LOT devices with a power supply (AC to DC converters). Good thing is that power company not "punish" for that. So, Will's calculations are way off.

[@zyghom]

@@kdelios wait, if he measures only resistive and the vendor charges only for resistive....- what is wrong then? he does not measure for the "truth" but for what he is charged, right?

[@janpruha3133]

This is more of a guide on how not to do it.

[@PiyushNikam]

We don't have ebay in India, how do I get this circuit? You should do projects such that everyone is able to do the exact same project at their home. Don't do some "proprietary" stuff.

[@PiyushNikam]

1:03 what if I don't have this shit and its not available where I live? You should make a generic video for everyone and not use any special component. If you would have used your teenie tiny brain, then you can just soldered that circuit on your own and shared the circuit diagram with us

[@turboturd7954]

Great work Will!

[@WillSurridgeTech]

Thanks!