sensitive magnetic switch

[RyreInc]

I'm looking for an electronic switch tripped by a magnetic field in the range of ±10 G. Must be omnipolar and small formfactor, but beyond that requirements are pretty flexible. Closest contenders I've found are magnetoresistive: the Honeywell 2SS52M, which triggers at 25G max, 12G min; Bernstein MEM-M10PS which triggers at 10G nominally (but do not specify range)but is physically too large; and Bernstein MEN-D06PS which triggers at 15G nominally (but again does not specify range).

There are analog/ratiometric/PWM sensors that go below 1G--is it possible/practical to use one of these with a filter and comparator, or would there be too much noise?

Any other leads out there I'm missing? (I've tried Digi-key and Mouser as well as Google)

 

[IRstuff]

You might look at magnetometers:

http://www.magneticsensors.com/magnetic-sensor-products.php

Some of these, like HMC1053 have an upper limit of 6 Gauss, but you'd need a processor to crank the total field to decide if the threshold is met, but you have total flexibility on hysteresis and the like. Obviously, much more complicated than a simple sensor+switch.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[EdStainless]

There used to be micro-reed switches available. Some of them were the size of a grain of rice and very sensitive.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[RyreInc]

Thanks for the replies.

IRstuff: Unfortunately a magnetometer is not on option--this will be used in a production setting to verify whether a container full of parts is magnetized or not--we need 60 or 120 sensors operating at once.

EdStainless: Thanks for the lead--I had considered reed switches previously but was not able to find ones small enough/sensitive enough - I now have (e.g. PIC reed micro switches), and have put in a small order to test these.

I would still prefer to have a solid state solution, so if there's anything else out there let me know!

 

[IRstuff]

?? Those reed switches don't appear to be omnipolar, so you may need a number of them for each installation. But, more importantly, you need to have both poles close enough together to get both reeds to contact each other, since they need to be oppositely magnetized.

Since you need something to monitor these 60 or 120 sensors, the same something could multiplex the outputs from magnetometers. Note that your reed switches will bounce, so there will need to be some level of debounce processing at the controller end anyway. The sensitivity of the magnetometer is roughly 1000 times better than the reed switch in question: HMC5883L

Note that the HMC5883L has I2L output, which would be directly compatible with most industrial controllers.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[RyreInc]

IRstuff, thanks for pointing out the potential pitfalls of a reed switch, as well as the potential benefits of using a magnetometer. I will certainly look further into these. I'm not currently familiar with the I²C interface, I'll investigate that.

For the application in question, I think two reed switches would be acceptable, as long as they are sufficiently sensitive. We need to test whether a magnet is magnetized or not after having been placed in a container. The magnet is a 6-pole radial cylinder, with poles spaced appropriately for the reed switches cited above. The sensors would be consistently oriented to the part, and spaced such that one sensor is at a peak while the other is at a null so that part orientation is not an issue.

Contact bounce will not be an issue, as we plan to either light an LED per part and/or use simple logic to determine that all 60 parts in the container are magnetized, either way on a multi-second timescale.

(My definition of omnipolar simply means the sensor can detect a field of positive and negative polarity BTW.)

 

[IRstuff]

That's OK, given the much more constrained scenario that could have applied. The narrower scenario may actually require sensors that are less sensitive to ensure that neighboring poles don't create false positives.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[mikekilroy]

I once had to design my own sine/cosine encoder that watched linear motor magnets go by. Used simple 3 terminal "hall sensors" avail everywhere including digikey, mouser, etc, for a dollar each. worked fine. dont recall the gauss they worked with but i think it is lot less than u want/need. I built a couple dozen for the project with differential line driver output chip for each; each pcb was about 10mm x10mm but could be smaller.... google "hall sensors."