# Detect AC current in series with load

I have a challenge at work to build a device to monitor the health of ceramic cartridge heaters, similar to this. It operates with an AC transformer and there are 5 heaters I'd like to monitor independently. The transformer operates on our 120V mains and is estimated to provide 1A to the heater.

First I picked up a non-invasive current sensor but in the limited testing I've done so far, I think the < 1A current and the gauge of the wires won't provide the necessary field needed for useful monitoring. Additionally I'd have to scale the device's 0-1V output to 0-5V logic of my microcontroller.

Then I thought about sticking an LED indicator inline with the heaters, like this one. This brings me to my main question: if I put this LED in series with the heater, as in Transformer > LED > heater, will the current drawn by the heater just burn out the LED? Would a resistor in the circuit allow the indicator to work properly without also limiting the heater and/or just melting?

• What is the accuracy of measurement needed? Have a look at ACS712 application. LEDs are not rated for that huge current. You can still have an LED indication circuit with the help of ACS712 or similar device. – User323693 Jun 30 '16 at 3:17
• You can increase the apparent current from the clamp-type connector by looping the sense wire through the clamp multiple times. You cannot use a logic input as the receiver for the current signal. An ADC of some sort is necessary, plus the logic to process the AC waveform. – user2943160 Jun 30 '16 at 3:31
• Also, the LED indicator will literally be burned to a crisp if you tried to use it in series with the heater. It will provide no indication other than that of exceeding its current rating and there's no way that series LEDs could effectively demonstrate heater health if they were rated for the current. – user2943160 Jun 30 '16 at 3:34
• Do you want to measure and display the actual current, or just get a go/no-go indication? – Bruce Abbott Jun 30 '16 at 4:31
• It's primarily a go/no-go indicator, as the heaters tend to have inf resistance when they burn out – MechEngineer Jun 30 '16 at 16:01

Current transformers with indicator LEDs are available. (Search for "remote current indicator LED".) The wire to be monitored is passed through the centre of the core. These give the advantage of electrical isolation from the mains wiring.

If required, you could cut the LED off and connect to an opto-isolator to pass the signal through to your controller. Pay attention to see if there is a reverse protection diode in with the LED and keep polarity the same.

simulate this circuit – Schematic created using CircuitLab

Figure 1a: Likely internal circuit. 1b: Possible alternative with back-to-back LEDs in one package. 1c: Replacing the LED with an opto-isolator for connection to your monitor / alarm circuit.

Note that a certain number of ampere-turns (amps in wire $\times$ turns through the core) will be required to turn on the LED. This will be listed in the datasheet. If the device turns out to be not sensitive enough you can fix this by threading more turns through the core.

e.g., Load current is 3 A. Current sensor rating is 10 A. Then thread the conductor through four times as $4~turns \times 3~A = 12~At$.

• The current indicator with LED would actually work great for the application, since all i really need to know is whether current is still flowing. I'll have to check what the typical current is to pick the right turn-off point. – MechEngineer Jul 5 '16 at 14:07
• See update regarding ampere-turns. Don't forget to accept one of the answers on the page if you've got what you wanted. ;^) – Transistor Jul 5 '16 at 20:38
• At this stage I've settled on doing something similar to this, except with a standard sensor and Arduino to handle the minimum threshold. So far testing is promising. – MechEngineer Aug 12 '16 at 14:52

If you are just interested in a visual hint (rather than an integration into an automated system), just grab a bag of cheap current indicators. Save yourself fiddling with a micro controller board.

Like these. Not trying to advertise those specifically, just to get my idea across. They were the first to show up on Google.

• Actually that could be a great solution! It looks like they have the capacity to show what we're measuring, and I could build a box to house multiple indicators together. Thanks for the tip! – MechEngineer Jun 30 '16 at 16:02
• @polwel: Include a photo of the device so that when that link goes dead future readers will know what you're talking about. MechEngineer, be careful with the insulation rating of the displays. They're cheap so if the plastic screen breaks they could be live. i.e., The electronics won't be isolated from the mains input terminals, etc. You may also face problems trying to provide power to them from a common supply when they're measuring on different supplies. – Transistor Jun 30 '16 at 19:33
• I was thinking the OP uses one transformer per heater? – polwel Jul 1 '16 at 9:15

I don't know your budget, but you can look for a current (or voltage) limit sensor or limit alarm. A lot of these are used in temperature control systems, but can be applied to others. They come in a variety of current and voltage ranges and can sense down below 1 V, even in the mA or mV range if necessary.

Here are some examples, not playing favorites with manufacturers:

OMEGA: DMD1080

Moore Instruments: Limit Alarm Trips and Switches

I have more examples including Acromag and Action I/O and more from Omega, but since I don't have much rep, I can't post anymore links, you can search for them if you'd like.

If you really intend to go with a LED as an indicator, here is a KISS approach:

Put 5 or so regular diodes all in series with the heater. Each diode has a voltage drop of about 0.6 V at 1 A, so they give you a nice 3V total (choose diodes beefy enough to eat 1 ampere). The missing 3 volts do not affect the heater's power in any significant way.

Then connect a LED (forward drop in the 2V range) together with a suitable current limiting resistor in parallel to the diode chain.

• Can you explain a little more in schematic or psuedo-schematic form? I'm having trouble visualizing how the diodes bring the voltage down from 120V for the LED – MechEngineer Jun 30 '16 at 16:06
• @polwel: It's an AC circuit. You'd need a forward and reverse strings of diodes. – Transistor Jun 30 '16 at 19:36
• @transistor: You're right. That doubles the number of diodes required. Suddenly this doesn't sound like a simple solution anymore... – polwel Jul 1 '16 at 9:19