[tyntop]

    Hi to all.
I am looking for some help and advise, I am thinking of using hall effect sensors instead of the reed switches that are commonly used in fine offset sensors. ie Anemometer, wind direction.

Anemometer : I was thinking of using 2 hall effect sensors 180 degrees
opposite each other, so that i would get 2 pulses, as if still
a reed switch.

Wind direction :I was thinking of using 16 hall effect sensors equally spaced
at 22.5 degrees which would give hopefully 16 pulses.

My query is, do hall effect sensors need to be connected separately which would mean a very complicated pcb or can they be connected in series using a common rail.The way I am thinking is that supply and ground can be in seriies and then the output kept separate but to a common point as per diagram.

ANY help and advise will be appreciated.

regards
tyntop.

[AllyCat]

Hi,

do hall effect sensors need to be connected separately which would mean a very complicated pcb or can they be connected in series using a common rail.The way I am thinking is that supply and ground can be in seriies and then the output kept separate but to a common point as per diagram.

"Hall effect sensor" is a generic name and there are many different types, so you probably need to quote a specific device part number.

They generally have three pins (ground, supply and output) and you would normally connect them "in parallel" across the ground and supply rails. But generally you cannot just connect the output pins in parallel (I'm not sure how it could be done "in series") because they have low output impedance and will "fight" each other, possibly causing damage. Basically you need an "OR" gate (or gates) with one input for each "output" signal.

The only exception is if the devices have a "Wired OR" output (which normally pulls the output pin Down but not Up), or you add a diode in series with each output (effectively making a passive OR gate). Then they can be linked together, but need a Pull Up resistor to the supply rail. That might be the way to do two sensors for the anemometer, provided that the "inactive" magnetic state gives a High output level (otherwise you need an AND gate).

For the Vane, each switch needs a separate "identity" which is done by using a different value of series resistor (and again each switch really needs to go "Open Circuit" when Off, not "High"). Fine Offset use eight reasonably "normal" resistor values (although IMHO not very well chosen), but for 16 you will probably have to fabricate some values by putting resistors in parallel (or series). You may also need to take into account the output impedance of the Hall devices (reeds of course have "zero" resistance).

But IMHO the major issue with the FO sensors is NOT the reed switches but the "design" of the magnet fields (paricularly for the Vane). I think you will be very fortunate if you manage to choose or design 16 Hall devices where one and only one is activated at EVERY possible vane angle. At least the FO design "guarantees" that either one OR two reeds will be closed at every angle and thus a "valid" direction is always reported.

Finally, I first worked with reed switches literally 50 years ago when they were being used for modern telephone exchanges. In those days, IIRC, the "design lifetime" for telephone exchanges was more than 20 years, so I don't think that you need to worry about the reliability of (good quality) reed switches !

I do have my own ideas and methods for "improving" the Fine Offset sensors and hope to post them here in due course (but don't hold your breath).

Cheers, Alan.

PS: The use of OR and AND gates above may seem rather confusing. That's partially because a "Wired OR" system is usually "active low", which means that strictly (in "Positive Logic" terms) it's an AND gate (where any one Low input makes the output Low).

[werk_ag]

... That might be the way to do two sensors for the anemometer, provided that the "inactive" magnetic state gives a High output level (otherwise you need an AND gate).

Hi AllyCat

To avoid such "complications", I'm testing a modified FO anemometer with just one hall effect sensor, which gives me only one pulse per revolution. Since some weeks, I'm testing it side by side with a standard FO anemometer. The gust readings are very very identical, and the average speed readings are exactly the same.

To read both anemometers I'm using two WeatherDuino Pro TX units, one of them with a slight change on the software to adapt to one pulse per revolution instead of two.

Modify a FO wind vane to use hall effect sensors instead of the reed switch seems to me complicated, and maybe not worth.