WCS1800 Current Sensor Arduino Project

Creator: EnabrinTain Status: Development Born On: 13:04, 1 August 2019 (CDT) Last Updated: 13:04, 01 August 2019 (CDT)

Overview

I'm hanging info on how to use the WCS1800 Current Sensor Module with an D1 Mini here.

Hall Sensors vs Current Transformers and commentary

Hall Sensors use a generated magnetic field to detect variances when conductive objects break the field lines. They can also be used to detect DC and AC current, where a Current Transformer can only be used to detect AC. (link) I decided to use this module instead of a Current Transformer (CT) because the D1 Mini only has one available analog pin, and I wanted to attach three sensors to a single module.

1. Main Shop Lights
2. Big Fan
3. Office Air Conditioner (using the analog pin, as well as a digital pin)

More about the WCS1800 Hall Effect Base Linear Current Sensor

• The WCS1800 Module has a separate digital output pin that can be tuned to go LOW when a current threshold is detected, based on a trim pot on the board.
• This will give me simple "on/off" info for the two appliances that dont vary in their power use.
• Hall sensors require active power, the WCS1800 can use anything from 3V to 14V, and had an operating current of 3mA.
• The datasheet I found tells me that it returns 60mV/A detected for an AC range of RMS 25A at 5V input (DC range was 35A). Since I'm going to be driving it at the D1 Mini's 3v3, I need to confirm that, and expect it to be different.
  • The graph showed values up to ±60A, but the curve started deviating from the simple linear after the above ranges were passed.
  • It did include a graph showing a mapping of Vout for Vcc of 3v3:
    • the curve was fit to the linear range of Vout = 0.0503*Ia + 1.6456, where Ia is the amperage being detected.
    • This will not be as accurate past ~20A however, for example I'd guess that while the formula would give a result of 32A if the Vout is 3.3V, the sensor would actually likely to be reading a current 50A or greater.
    • These values cycle at 60Hz, so it behooves us to sample our values at a high enough frequency to be able to find the peaks, and calculate the RMS of the current draw. There's a very good writeup about doing this in CT sensors (here). The logic for the WCS1800 will work in a similar way.

D1 Mini (ESP8266)

• I soldered wires for the Ground and 3V3 pins to power all the modules from a small piece of proto board. I'll be encasing it in Jeffweld for safety.
• The first WCS1800 module's AOUT and DOUT pins are attached to the D1 Mini's A0 and D0 pins respectively.
• The other two modules' DOUT pins will be soldered to the D1 Mini's D1 and D2 pins, but I haven't done it yet.
• I think the AC 60Hz cycle applies to digital outputs as well. This means some debouncing logic is necessary, even though we aren't reading voltages for current values.
  • I'm not 100% sure about that claim, however tests showed that the D1 Mini D0 pin was rapidly switching from high to low.

=Arduino Sketch (pin reading, and MQTT publishing)

TO DO (right now its just flipping the BUILTIN_LED when it sees a signal on D0 and printing the value of A0)

Next Steps

• Look at the MQTT messages being sent from the light sensors to see the format that grafana wants
• bump up the serial baud rate so that the loop isn't as laggy there
• Verify the math for the analog signal makes sense
• Add code to find peaks
• Add code to debounce D0
• Add wifi connection, look for timestamp library, configure MQTT client
• Add doobler config
• Add Grafana graph