I'm looking for ideas/suggestions for a completely non-intrusive detection mechanism for an AC motor. I intend to monitor the duration and timestamp of an AC motor. I believe it is a 1/7 HP permanent split capacitor single phase induction motor. I have no access to any of the input wiring, i.e. no sensing or current probes.

My thought was to use a hall effect sensor to detect alternating magnetic fields generated by the stator coils. I connected a latching hall effect sensor, HAL102 with +- 2.5 mT sensitivity, to an oscilloscope and moved it around a mini AC fan. This generated a 60Hz square wave, which was the exact behavior I was desiring. The problem was when I tried to detect the motor I wasn't able to generate any signals.

A qualifier: I'm by no means an expert w/ AC/DC motors. I'm more experienced in digital circuits and embedded software development. I would appreciate any help or ideas. Things I was wondering; What would be the expected magnetic field strength of stator coils at a known distance, thinking inverse square law? What would be the expected magnetic field pattern, i.e. I'm thinking it would be alternating north/south normal to the motor's housing? Any other alternative ideas for monitoring/identifying energized motor. By the way, I don't have access to the rotor and therefore cannot use any encoders.

Thanks in advance for help.

 

On a split phase motor there is two fields (Main and split) which are ideally 90° apart.
A mini fan is generally a shaded coil motor, but basically similar, just a different method of producing the phase shift.
Also on the motor you are trying to detect, are the lamination's exposed or are they encased in a frame? If so this could be offering some shielding.
Do you just require the running time of the motor?
There are also current detection devices made by Honeywell that are non-invasive by detecting motor current by passing one conductor through the device and have low voltage signal output, .
Max.

 

Thanks for the response MaxHeadRoom. I believe the laminations are exposed. Unfortunately, I have no access to the conductors, or I would use the Honeywell current detection. Is it correct of me to assume that the magnetic field is normal to the motor surface? The motor I'm using looks very similar to this.

 

Do you want evidence the motor is actually turning, or just drawing current? What is the motor turning?
I'm just thinking how to expand the options.

 

I'm not so concern that the motor is turning, if its working correctly, which I'll assume it is, than if it is powered it should be turning. This is a motor for burner and it connects a blower fan. I'm just interested in knowing that it is "on". I'm specifically interested in knowing the start/stop time for run-time statistics.

 

A typical HVAC solution to this is to measure ∆P, the pressure difference developed by the blower. For instance my furnace won't light unless it sees a small vacuum pulled by the exhaust fan.

 

My two cents: I set up a current monitor for a PSC (permanent split capacitor) motor using the ACS712 linear hall effect sensor, which measured current flowing from the neutral leg of the motor to mains neutral. I used an inverting amp to scale the output to my micro's ADC, similar to the example circuit they show in the chip's datasheet.

I mention this because you could utilize a similar solution, and it could be even simpler than what I did. The ACS712 is a readily available chip, Sparkfun even has a breakout board for it: https://www.sparkfun.com/products/8882

You could feed the low voltage output to the cathode of a >2.5V zener (the chips 0A voltage is 2.5V), which would prevent current flow until the motor is running. Inrush current will produce a high voltage output at startup, ensuring motor start is detected. From there, the output could energize a mosfet, which would toggle a micro input.

 

I believe the laminations are exposed. Unfortunately, I have no access to the conductors, or I would use the Honeywell current detection. Is it correct of me to assume that the magnetic field is normal to the motor surface? The motor I'm using looks very similar to this.

Have you tried a uni-polar or omni-polar instead of the latch type?
Honeywell have them in the ss400 series.
Max.

 

My two cents: I set up a current monitor for a PSC (permanent split capacitor) motor using the ACS712 linear hall effect sensor, which measured current flowing from the neutral leg of the motor to mains neutral. I used an inverting amp to scale the output to my micro's ADC, similar to the example circuit they show in the chip's datasheet.

You could feed the low voltage output to the cathode of a >2.5V zener (the chips 0A voltage is 2.5V), which would prevent current flow until the motor is running. Inrush current will produce a high voltage output at startup, ensuring motor start is detected. From there, the output could energize a mosfet, which would toggle a micro input.

b1u3sf4n09 thanks for the feedback. I'm not sure how your PSC is setup, but I don't have easy access to the the motor neutral/hot as they go directly from the burner into the motor wrapped in a sheathing (similar to the image below). I also don't want to cut or strip any sections of this mains voltage. Although this is a neat idea, I ideally trying for a completely non-intrusive solution, that any layman with minor instructions could attach to their burner.

I will keep in mind your zener diode if I use a hall effect to try and monitor the magnetic coils. I feel as though, based on Faraday's law, that the field strength created by the coils should be magnitudes larger than the single conductor due to the N turns multiplier.

 

Have you tried a uni-polar or omni-polar instead of the latch type?
Honeywell have them in the ss400 series.
Max.

I will look into the uni/omni polar hall effect sensors. I have a box of samples provided by a Micronas rep. At a quick glance I thought the latching one had the best sensitivity and I liked the open-collector digital output. This is my first project where I'm responsible for designing in a hall effect sensor, so I'm still becoming familiar.

One observation that I noticed, when using a an analog magnetic compass immediately next to the motor is the south pole always pulled towards the motor when running. When the motor stopped, it appeared to balance out a little better, but still had some pull. I'm not sure what I was expecting to see, but it wasn't that. I figured the responsiveness of the analog compass wouldn't be good enough to keep up with the 60Hz field. I figured if I scrambled the compass it would be a 50/50 shot if north or south would "lock" on. In addition, if anything I would think north would have pointed towards the motor, since that is earth's magnetic north.

 

I get the Honeywell from DigiKey they usually have a selection of the SS400 series, all open collector as well.
Max.

 

Just came to me a latch type may not work if you want steady state output when the field is present as it will try to latch/unlatch every cycle, a unipolor may be the best to stay on while the motor is running.
Max.

 

Just came to me a latch type may not work if you want steady state output when the field is present as it will try to latch/unlatch every cycle, a unipolor may be the best to stay on while the motor is running.
Max.

I was thinking about just setting up a timer and if I don't see any edges after ~3-4 cycles I would consider the motor off. With that being said a steady field would be even easier.

 

I have a few of the different Honeywell switches If I can find a suitable motor I may try it!
Max.

 

I did not have suitable motor handy, but I tried a unipolar on a 1kva power transformer, lightly loaded and I could not get it to trip, which surprised me, I imagine the surrounding field was not sufficient or it did not activate on a AC field?
Max.