Low Power Proximity Sensor Feasibility
2

Question

Low Power Proximity Sensor Feasibility

goldenfab (Aerospace)

(OP)

24 Oct 17 17:15

I am working on a product that is currently using a ~~normally open~~ normally closed [edited] reed switch that when a magnetic field is present turns on off [edited] a small LED. I was told other types of sensors like a hall effect sensor was looked into but the longest battery life that could be expected was 2 months with a 2032 button cell battery. The primary reason for wanting to get away from a reed switch is to be able to adjust the trigger point (the distance a magnet has to be to turn on an LED).

Does anyone have any experience with any types of magnetic field proximity sensor that could have circuitry that would allow for adjustment of the trigger point by something like a potentiometer and have a battery life > 2 months (hopefully a year or more) assuming ~~the it is not triggered turning on an LED~~ a magnetic field is present and the LED is turned off [edited].

Thanks in advance.

IRstuff (Aerospace) · Answer 1 · 2017-10-24 18:47:39

The datasheet: http://data.energizer.com/pdfs/cr2032.pdf says 240 mAh, and that's to 2V, which is fully 1V below its nominal starting point.

If you are running this continuously, then you cannot draw more than 27.4 uA. That's a pretty tiny budget.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! https://www.youtube.com/watch?v=BKorP55Aqvg
FAQ731-376: Eng-Tips.com Forum Policies forum1529: Translation Assistance for Engineers Entire Forum list http://www.eng-tips.com/forumlist.cfm

goldenfab (Aerospace) · Answer 2 · 2017-10-24 19:56:53

Yes I agree a very tiny budget hence why a reed switching is currently being used. The only plausible solution I have come up with is using a reed switch in conjunction with something else where the reed switch turns on when close and the digital logic does the fine triggering. This seems like an over engineered solution. Putting a feeler out there to see if there is some new wiz bang sensor that would do this on such a small power budget.

Compositepro (Chemical) · Answer 3 · 2017-10-24 22:22:56

Have you considered making the reed switch position adjustable or using an adjustable iron pole piece between the magnet and reed switch to focus the magnetic field differently.

goldenfab (Aerospace) · Answer 4 · 2017-10-24 22:26:11

Adjusting the location of the reed switch would be troublesome to incorporate into the design but an adjustable iron pole could be a great solution. Thanks for the idea!

itsmoked (Electrical) · Answer 5 · 2017-10-24 23:56:59

You are not going to come anywhere close to the zero power drain of a pure reed switch solution. Work with it somehow.

Keith Cress
kcress - http://www.flaminsystems.com

IRstuff (Aerospace) · Answer 6 · 2017-10-25 02:10:13

True, but there are some alternatives that are not too horrible:
http://www.allegromicro.com/en/Products/Magnetic-D...
http://www.ti.com/lit/wp/slyy058a/slyy058a.pdf

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! https://www.youtube.com/watch?v=BKorP55Aqvg
FAQ731-376: Eng-Tips.com Forum Policies forum1529: Translation Assistance for Engineers Entire Forum list http://www.eng-tips.com/forumlist.cfm

VE1BLL (Military) · Answer 7 · 2017-10-25 11:55:05

Duty Cycle is always an option. Work out just how fast that your sensor can power up and make the reading. Let's say (for example) that it's one millisecond. Let's say that you need to check once per second. So, on the face of it, there's about a 1000:1 duty cycle ratio available (it'll be a bit worse than that). Obviously this is combined with the processor's Sleep function and tight efficient code.

The next thing might be to use an analog Hall Effect Sensor (without all the digital circuitry built-in), and see if you can current starve it. See how low you can go and still get a reliable signal. As a pure guess, there might (maybe, perhaps) be an order of magnitude available.

Another idea. Change the simple battery supply into one that uses a "Joule Thief" circuit. Beware the trade-offs inherent in this.

MacGyverS2000 (Electrical) · Answer 8 · 2017-10-25 14:12:03

Using a 16-pin MicroChip processor and their sleep ability, one of my products sipped <50uA (on average). It checked a couple of inputs, and if nothing was active, went to sleep. It repeated this cycle every 10ms (MAY have been 100ms, but memory makes me think 10)... way faster than the user could recognize, but plenty slow enough to save a lot of power. If you can go longer between active sessions, the power comes down even more. This was 15 years ago, and sleep-level power is continually being reduced in the processor world.

Dan - Owner
http://www.Hi-TecDesigns.com

IRstuff (Aerospace) · Answer 9 · 2017-10-25 14:27:50

The TI application note I linked describes exactly that case. They claim an average of 0.9 uA, which would allow the CR2032 to last over 10 years.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! https://www.youtube.com/watch?v=BKorP55Aqvg
FAQ731-376: Eng-Tips.com Forum Policies forum1529: Translation Assistance for Engineers Entire Forum list http://www.eng-tips.com/forumlist.cfm

goldenfab (Aerospace) · Answer 10 · 2017-10-25 14:53:30

I misspoke in my original post and have since edited it. The desired function is to have an LED turn off when a magnet is close and turn on when the magnet goes away.

The TI application note IRstuff shared has a super impressive battery life. Unfortunately as far as I can tell reading the data sheet would only allow for open drain so it would not be able to function like a normally closed reed switch (my fault for misleading you guys in the OP). This one might work though http://www.ti.com/lit/ds/symlink/drv5032.pdf Link. Unfortunately is looks like these are only on/off and can't be made to utilize and adjustable trigger point. Need to check out the analog ones TI makes to see what the power usage is. I'm guessing they will blow my tiny battery budget.

VE1BLL (Military) · Answer 11 · 2017-10-25 15:59:57

FYI (just in case some are not aware).

There are "reverse" ('Normally Closed') magnetic reed switches.

They're built so that the contacts become open circuit when a magnet is brought close.

Otherwise just like the more common 'Normally Open' variety.

It's been a few decades since I've seen them, but presumably they'd still exist as a variation.

[EDIT: Yep. They seem to be readily available.]

IRstuff (Aerospace) · Answer 12 · 2017-10-25 16:08:30

The MCU in the TI application note: http://www.ti.com/product/msp430g2230-ep has ADC capability, so all that's required an analog Hall effect sensor: http://www.ti.com/product/drv5053-q1 The thresholding would require some software in the MCU, and driving a relay, but all else is applicable.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! https://www.youtube.com/watch?v=BKorP55Aqvg
FAQ731-376: Eng-Tips.com Forum Policies forum1529: Translation Assistance for Engineers Entire Forum list http://www.eng-tips.com/forumlist.cfm

MacGyverS2000 (Electrical) · Answer 13 · 2017-10-25 17:53:37

Is the ADC included in that low-power figure? Usually not as the ADC not only consumes quite a bit of power, but it also takes a long time for the clock to stabilize. That requirement has killed off a few ideas of mine in the past.

Dan - Owner
http://www.Hi-TecDesigns.com

VE1BLL (Military) · Answer 14 · 2017-10-25 20:13:30

Going back to basics:

OP: "The primary reason for wanting to get away from a reed switch is to be able to adjust the trigger point..."

If another reason for getting away from the reed switch is to increase range, then none of the following would work.

If the reed switch was sufficiently sensitive (range of the trigger point), and assuming that the adjustment of the Hall Effect sensor is to reduce the range of the trigger point (inside the range of the reed switch), and assuming that the magnet doesn't continuously come close enough to trigger the reed switch yet just outside the adjusted trigger point, then...

...One could use the reed switch to provide the hard trigger, and then (and only then) check the distance with the Hall Effect sensor. This way the power-consuming Hall Effect sensor circuits would only be active those rare times when the magnet is getting close.

Even with a 'Normally Closed' reed switch, it could still be wired to wake up the sleeping processor only when the magnet is "in the neighbourhood".

IRstuff (Aerospace) · Answer 15 · 2017-10-26 00:27:20

The datasheet puts the ADC current on the order of a couple of mA, but it only needs to be on for a short time, say 25 us. If the system sampling rate is, 10 Hz, the average ADC current draw would be 0.5 uA.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! https://www.youtube.com/watch?v=BKorP55Aqvg
FAQ731-376: Eng-Tips.com Forum Policies forum1529: Translation Assistance for Engineers Entire Forum list http://www.eng-tips.com/forumlist.cfm

Come Join Us!

Posting Guidelines

Jobs

Low Power Proximity Sensor Feasibility
2