hall sensors to detect slideby position of a magnet

[mook johnson]

I'm using linear hall sensors to detect the position of a small SmCo magnet
about 0.75" away.
There will be several sensors in a line, spaced about .4" apart. The magnet
will be directly below these sensors about and travel along a parallel path.
According to my calculations, the magnet will produce about 75G at this
distance, which will product a signal strength of 375mV from a sensor when
the magnet is directly below it.

What I'd like to accomplish is to resolve the position of the magnet down to
no greater than 0.05". Also, I must take into account that the magnet
strength and hall sensitivity will not be constant through out the life of
the part, so the measurement should take into account varying field strength
measurement. In-situ calibration is a fallback but not desired.

Any hall sensor gurus out there?

 

[Wayne]

I'm using linear hall sensors to detect the position of a small SmCo magnet
about 0.75" away.
There will be several sensors in a line, spaced about .4" apart. The magnet
will be directly below these sensors about and travel along a parallel path.
According to my calculations, the magnet will produce about 75G at this
distance, which will product a signal strength of 375mV from a sensor when
the magnet is directly below it.

What I'd like to accomplish is to resolve the position of the magnet down to
no greater than 0.05". Also, I must take into account that the magnet
strength and hall sensitivity will not be constant through out the life of
the part, so the measurement should take into account varying field strength
measurement. In-situ calibration is a fallback but not desired.

Any hall sensor gurus out there?

Try using two opposing magnets side by side. ie. NS and SN. This will
produce a neutral field between the two.

 

[mook johnson]

How do I resolve the location of the null spot down to .05" with the sensors
..4" apart?

 

[Rich Grise]

How do I resolve the location of the null spot down to .05" with the sensors
.4" apart?

I'd think look at all the outputs simultaneously and fit a curve, and
find that curve's maximum. But, as you've mentioned those things change -
(I went nuts once trying to protect a hall device that was metering cathode
current of an ion gun at about 5 KV. The transient from an arc didn't
_destroy_ the device, but changed its transfer function (mostly offset)
unpredictably.) so I can't see a way to get around _some_ kind of
periodic calibration.

Good Luck!
Rich

 

[Joerg]

Does it have to be Hall? Or could it be optical or inductive? Just thinking
about how to make it low maintenance.

Regards, Joerg.

 

[mook johnson]

Hall (or magnetiresistive) is first way thought of. This issue I have here
is that these electronics will be packaged in a .75" ID stainless (u =
1.005) pipe about 1 inches long). The power available is about 1 watt (24
volts input) and the temperature will be as high as 125C. Outside pressure
can be as high as 10K PSI.

If there are other technologies for detecting the magnet or a magnetic
material (steel) with these constraints, I'm all ears...ahmme eyes .

thanks.

 

[Joerg]

You'd have to share more about the application to see what's best. It also
depends on time schedule, qties, cost etc. But if there is a way to have a
little marker piece of material run through a light detector barrier slot then
that would be very accurate. Same for inductive where you have to create a
magnetic field that is very pointed and gets disturbed by a marker needle or
something like that.

Hall works as well but I have hardly ever used it. It might help to check the
automotive world because they do the ignition timing that way. And this stuff
just has got to be top precision with all the smog regulations these days. I
believe they measure the position of the engine's crankshaft down to fractions
of a degree.

0.75" diameter by 1" long is a whole lot of space. On a recent project I just
crammed 3-4 dozen parts in there plus batteries. And 1W is a lot of power these
days. It can be done much smaller.

Regards, Joerg.

 

[Dave Cole]

Hall (or magnetiresistive) is first way thought of. This issue I have here
is that these electronics will be packaged in a .75" ID stainless (u =
1.005) pipe about 1 inches long). The power available is about 1 watt (24
volts input) and the temperature will be as high as 125C. Outside pressure
can be as high as 10K PSI.

If there are other technologies for detecting the magnet or a magnetic
material (steel) with these constraints, I'm all ears...ahmme eyes .

thanks.

Howdy.
How about LVDT
http://www.efunda.com/designstandards/sensors/lvdt/lvdt_intro.cfm
or magnetosonic?
http://www.temposonics.com
HTH
Dave Cole