Very low friction potentiometer or encoder

Hi Tobias,

If you can find a low friction bearing, you could use a non-contact sensor like the Melexis MLX90316. You just need to stick a magnet on the end of the shaft.

Doug

Thanks Douglas! I have been looking into similar sensors, but as you said, I then have to actually construct the assembly around it. This I would like to avoid.

Hi Tobias,

Perhaps you could just stick a magnet on the shaft of the wind vane...here is an Arduino implementation example:

Lab3 - Laboratory for Experimental Computer Science

Alternatively, some complete sensors would include:

Avago AEAT-6012-A06 (about $35)

Bourns EMS22P50-B28-LS6 (about $45)

Doug

Thanks!

Most wind vanes rotates around the actual axel/shaft. So the shaft stands still and vane moves. 

The Bourns EMS22P50-B28-LS6 looks very interesting since its fully assembled. I can find a sister model here at Element14, but its way more than $45. More like $100.

There is a video showing a wireless direction vane, that feeds into an arduino.

He said he used a magnetic encoder but doesn't give an example.

element14 have a mems sensor for the Freescale that includes a magnetic compass chip.

MEMS Sensors Evaluation Board

Freescale XTRINSIC : Accelerometer, Magnetometer, Pressure and Touch Sensors

I have one but sadly it hasn't gotten to be coded and included into mine ....

Short of sticking a bigger vane on, you'll need to find a lower friction device.

Mark

You could attach a tiny strong magnet to the vane. Arrange a halo of hall effect sensors on the shaft nearby. They're really easy to use, and cheap. If you're pin limited you could multiplex them in groups, otherwise you just need to read which of the detectors is registering a magnet nearby. If you're only after 45 degree resolution you could probably get away with 8. You should select a sensor type which can detect over a relatively wide area, the orientation of some needs to be tightly controlled. This should avoid dead space between the sensors. I see that you don't want to build an enclosure around the detector but you could probably construct this 'dead bug' style in a little over the same volume you would be allocating to a pot.

Charlie

Hi!

I just finished making a prototype wind vane. I used a rotary encoder (like this: http://www.ebay.com/itm/10pcs-New-High-Quality-Rotary-Encoder-Switch-With-Keyswitch-/310911594094?pt=LH_DefaultDomain_0&…) for detection of position, it's got 18 degrees of resolution.

But since it could occationally miss a step when moved around fast to both sides, I added a IR led and IR receiver that is used to calibrate north against the flat side of the rotary encoders shaft, after adding this it has been spot on.

The encoder got a little mechanical friction, but not enough to not move with small winds if the vane is properly designed.

I currently have it display the wind in degrees, and it shows the direction like: N, NE, etc.

Think I have nailed the calibration of north function too.

Would be happy to share the code I have pieced together so far, this is a very cheap way of making an accurate wind vane.

Maybe I will try to replace the IR with a hall effect sensor and a magnet, it may be a better long term solution in harsh conditions?

.. Or use a sensor by AMS (5048, for instance), or RLS. Those just require one magnet (diametrically magnetized, for sale at supermagnete.de), and you'll get absolute rotation as output

Thank you for the tip! Haven't even read about that kind of sensor, just took what I had laying around.

Maybe i'll try that sensor in combination with a brushless motor or use the bearings and axle from a HDD for my next wind vane, don't really need the extra resolution, but hopefully it's a longer lasting solution.

Gents

Just some information regarding bearings and lubrication.

We were experiencing premature bearing failures in fans and anemometers, despite the installed ones lasting many years.

It was traced to the lubrication, and the newer synthetic oils had a shelf life of 1 year before they separated.

There service life was huge, but shelf life was 1 year, and once the bearings had been lubricated and sitting on the shelf for a year, they were effectively expired.

The solution was replacing them with items less than 3 months old, and the bearing issues went away.

So for your projects using something that's been lying around may not work, and if you can, flush out the old lubricant and replace it.

Mark