Is it possible to detect current ramp-up to DC with a coil?

[mikaeleriksson]

Hi,

I am thinking of ways to detect when an electrical device is turned on and off without actually touching or modifying the existing circuitry. More specifically, I have a hand held device powered by 2 x 9V batteries and with 2 electrodes connected through leads and plugs to sockets on the device, and that way connected to the internal circuitry. When turned on, the current ramps up linearly from 0 mA to 2 mA. The ramp-up rate is 0.1 mA/s. When shut down, the current is ramped down correspondingly.

It's been a while since I had anything to do with electronics and the Maxwell laws, and my googling didn't give me any more clarity in whether what I was thinking of doing is possible, so my hope is that someone here could enlighten me.

So, can I use a coil around the lead/plug and detect that the device is turned on due to the changing current? Or should the coil be placed perpendicularly to the lead and plug? I would then connect the coil to a microprocessor unit and send it by GSM from a GSM module to a server and collect the data there.

Or if this is not possible at all, are there any alternative options to induction into for how to do this without this coil solution that I'm considering?

If my idea with a coil works, are there any specific considerations I should think about with regards to technical requirements on the coil. Number of turns etc. Given that the rate of change of current is 0.1 mA/s for 20 seconds, from 0 to 2 mA.

I tried to be as clear as I could. Hope my message makes sense although I'm not very well-versed in this field yet, as you can probably tell.
Mikael

 

[anorlunda]

A simple way to do for a DC circuit it is with a resistor and capacitor in series with each other and connected in parallel with the switch. The voltage across the resistor will be zero except briefly when the switch is turned on or off. I lack a drawing tool just now, but the ASCII art below might help to illustrate.

|----R-----C-------|
|------switch------|

 

[CWatters]

Can you insert a low value shunt resistor in the battery leads? Perhaps 10-100R to give a 20-200mV signal?

I think if you wanted to use some sort of non-invasive method (like a hall effect sensor) you would need to extend and wind/loop the battery leads many times around a ferrite (field concentrator) to magnify the magnetic field.

 

[berkeman]

Welcome to the PF.

If you only need to know when the device is turned on, you could change out the SPST switch for a DPST switch, and use the 2nd pole in a circuit to indicate that the device has been switched on.

Otherwise, I think that the Hall Effect sensor is the right way to go to detect that the application current is flowing.

 

[meBigGuy]

To be non-intrusive, you need to use magnetic techniques.
Hall effect sensor is the best solution.
Can you wrap the leads around a core? It may be required for the low currents you want to detect.
2ma is a pretty low low current for magnetic detection.

http://www.allegromicro.com/en/Design-Center/Technical-Documents/Hall-Effect-Sensor-IC-Publications/Non-Intrusive-Hall-Effect-Current-Sensing-Techniques-for-Power-Electronics.aspx

 

[mikaeleriksson]

Thank you for all your input!

It does indeed seem like a Hall-effect device is my best bet, since the sensing method needs to be non-intrusive. The link posted by meBigGuy was excellent, and made it clear that current sensing methods are either 1) resistive (intrusive) 2) Hall-effect based or 3) current transformer based (AC only).

This leaves HE as my only option, and I need to figure out whether I need to wind the leads around a core to detect a signal as suggested by CWatters.

But this has been great, now I have a direction to go, and things to research and experiment with. Thank you everybody! :) And if anybody has anything to add I would be very grateful to hear about it of course.

 

[CWatters]

I've not actually tried to measure currents that small using a hall effect but thought I'd have a stab at calculating the field without a core...

The field B in the middle of an air coil is roughly given by

B = μ₀nI/2a
where
μ₀ = 4Pi * 10^-7 for air
n = number of turns (say 10?)
I = current (2mA)
a = radius say 0.5cm (0.005m)

So I make B about
=4Pi * 10^-7*10*0.002/(2* 0.005)
2.5 * 10^-6 T

Check my working but the Earths magnetic field is about 50 * 10^-6T

Looks like a core will be required.

 

[sophiecentaur]

It does indeed seem like a Hall-effect device is my best bet, since the sensing method needs to be non-intrusive.

There is a digital 'clamp' type ammeter (Power Stream PST-PR11), which measures up to 40A DC with a 1mA resolution. You may be able to interface with it (on an unofficial basis). The spec does imply that it's do-able.

 

[CWatters]

Try before you buy? Last clamp meter I played with it was very difficult to set the zero current. Open the jaws or move anything and you have to re-zero it. On the 0-400mA range it says accuracy 1.5% which is 6mA.

 

[sophiecentaur]

Try before you buy? Last clamp meter I played with it was very difficult to set the zero current. Open the jaws or move anything and you have to re-zero it. On the 0-400mA range it says accuracy 1.5% which is 6mA.

The need to adjust for offset may not be very important as long as your data logging is intelligent and you just need to spot an abrupt change (on/off) and you could ignore slow drifts. (High pass filter)
Buy from Amazon and they will refund under most conditions.

 

[meBigGuy]

Buy from Amazon and they will refund under most conditions.

As long as you pay all shipping

 

[sophiecentaur]

What would you expect?
Actually, I have had free returns on several occasions. If the unit is not up to spec you should get that.

 

[mikaeleriksson]

Thank you again to all of you for the help, I'll try with the clamp meter, it's worth a shot!