so we have a gsm SMS alert system (mod bus) that needs a signal that a wire (anyone wire from 3-phase) is having current (Max 15 amps in ON state) or no, so we can know that system is OFF or ON state, there is non-invasive CT product but need guidance of how to use it so that we get signal +5V or 0V to differentiate ON or OFF. Need to check ON or OFF state of a wire using a simple non-invasive method.
Hi I have a similar (maybe) need to monitor when my sump pump is running. I looked at the clamp approach but found that the clamp cannot go around the whole cable…the hot and common cancel each other out. I can create a cable that breaks out one wire and is inserted in line but that is somewhat invasive.
So, since I don’t care about the actual voltage, I am trying to construct a voltage detector like the ones electricians use to check if a wire or outlet is hot. Found this link https://www.jameco.com/Jameco/workshop/ … ector.html
It uses an antennae to detect the rf from the wire (works on the whole cable, just tape the antennae to the cable and I get an led to light when there is voltage in the cable) uses 3 transistors to amplify the signal but I haven’t figured out how to read the situation in the arduino (see attached schematic) …seems like that ought to be easy to modify for arduino interface but I don’t know how yet…also interference may be a problem.
in case you are thinking current sensor also learned along the way that current sensors can return either current or voltage…i got a voltage one that the arduino can read without addl components…you can do it with the current type but additional resisters are needed…may be other issues but the voltage type seem easier.
I am very new to all this so I may be going in the wrong direction. Would appreciate thoughts on talking to the arduino.
thanks
see Ralph Bacon #114 on youtube for a simple detector
Which ever method you choose, IMHO (based on more than 55 years of experience) you would be wise to include some form of “isolation” or “clamping” to asdsure that the detected voltages don’t exceed the input limits of the processor input pins. Note that if a voltage less than ground is applied to ANY pin not specifically designed to handle it, most processors will go into a “run away” state wherein if you realize it within less than a second and turn all power off you MIGHT save the processor. (I’ve seen them melt sockets!) Remember, the average AC powerline has a LOT of “noise”, and can pick up a lot more noise from thunderstorms, etc.
Note that if the circuit can get an LED to light, you can substitute the “input” side of an optoisolator for the LED and it should work.
I like the optoisolation idea. thanks. any suggestion of what to use…there are more than a couple that show up on a search. I am a real noob to building circuits and only 50 yrs of programming. I have the circuit working but it is real noisy…per ralph bacon added a charge pump after the led which seems to moderate the situation but looking at it on oscope its ugly…almost like pwm created using a micro switch, a cam and motor…which i did one time for fun…micro switch lasts about 5 mins. thanks again
DigiKey https://www.digikey.com/ has literally thousands of different ones. (Put optoisolator into the search window.) Give some thought as to what you need, then go through their “filters” cutting things down. (The “In stock” filter will often eliminate a bunch of stuff.) If what you’re trying to sense is AC, use the “AC-DC” filter. And it sounds like you probably only need one channel. If you connect direct to the AC, you’ll need to add a resistor in series to limit the current to around 10mA (or maybe less). These things can be pretty fast, so you might see 120Hz on the output… (remember, a 60Hz sine wave has 120 zero-crossings per second) so you’ll need to add some sort of “low pass” filter, either hardware on the output or software on the micro that’s connected.
Your microswitch PWM reminds me of a project I did in the early 90s. We’d taken a class in C++ at work, and the “toy” for the class was building a program that simulated a (simple) traffic light (on a SparcStation). A few weeks later I was playing with PEEL chips (a simple FPGA) on my own, and with just one PEEL plus a 555 timer to generate an approx. 1 Hz square wave, I replaced about 120 lines of code with some hardware! 
As I said, learning a lot of new stuff…sharing your thoughts and experience are much appreciated. I did a quick search on digikey (been ordering stuff from them for maybe 40 yrs) and as you said lots of choices…only 6100+ in stock SKUs before filtering. A long time ago I was playing with TTL (old copy of TTL cookbook) fun but it took a lot of chips to do anything useful with flipflops, nand gates, etc. Now have cheap arduinos, i2c and sensors so the limiting factor is really my imagination…started this looking for project to get smart grandson a step up from lego and away from minecraft…still working on that part but I am hooked.