Hi, I'm new to the forum and what I have seen so far is very good.

My problem is that I want to detect a magnet that is rotating - sounds easy - well the magnet is 100mm away from any possible detection device. I have tried Hall effect sensors and the like but have had no joy - the magnet is just too far away. I thought of using an inductor - say 1mH coil and, although I do get a slight ripple, looking across the coil with a scope, my attempts to amplify this have been unsuccessful.

My application is to determine wind speed - the magnet is on the axis of a propeller that the wind turns. The magnet is diametrically magnetised - the bar is N - S on opposite sides, not the normal N - S on opposite ends so the magnetic field is rotating N-S-N-S.....etc so it should be possible to detect it.

Any help would be most appreciated - thanks

 

The most sensitive thing I am aware of that could potentially detect 'tiny', at that distance, changes in magnetic flux would be a magnetic tape head.
I expect someone will tell me that a hard disk drive head is faster and better but I wouldn’t want to try soldering to one.
All that said .... you will almost certainly have far better success if you go optical.
100mm looking at say a patch of white paint or a small strip of foil would be relatively easy.
A modulated light source and tuned detector would be best but what you are trying to sense is so big and rough a signal that a constantly illuminated reflector and a simple photodiode will probably work.
Connect the detector output to a an opamp based ‘integrator’ (I think), which detects the rate of change of a signal, and then generate a known square pulse when the rate of change exceeds some set value.
There are good opamp tutorials on this site that include integrator and diferentiator circuits
( http://www.allaboutcircuits.com/vol_3/chpt_8/11.html )
You will probably find that IR will work best and you will need to restrict the field of view of the detector as much as is practical.
Hope this gets you started
Al

 

Why does the detector have to be so far from the magnet? A picture perhaps?

Ken

 

Many thanks for the suggestion - already thought of that - as they say! The equipment is battery powered so LOW current is essential - not good when powering LEDs. Also, there is no clear optical path to the propeller - that is why I built a magnet into the shaft. I was hoping a sensitive Hall effect device would do the trick but it did not. Optical would have been my first choice because of noise immunity etc but too much power I'm afraid - together with the lack of a clear optical path makes it a non starter.

 

Ken

Difficult to describe, but the propeller is in a housing that is 100mm above the box that any electronics can go in. The housing around the propeller prevents an optical method (see previous reply about too much power). I am stuck with what I have so have to try and make something work. Any suggestions would be most welcome.

Thanks again for looking

 

Without a picture (even a sketch) and a schematic of what you have, we will just be floundering for solutions. How about a weather-proofed Hall-effect sensor on a 10cm post protruding from the top of the electronics box?

Ken

 

Someone did make a good start mentioning a tape head, the one out of an old cassette or 8-Tr deck might be fairly sensitive.

The second thing would be to limit your amplification just to the areas of interest.

Too bad on your limited supply, however a shiny piece of foil, some precision optics and it still might work with simple ambient light - just not at night.

 

Pictures, pictures, we need pictures.


hgmjr

 

OK

Here is a picture of the concept. At the moment I do not have any hardware built to the correct proportions but this picture illustrates the principle. Note the weather vane is detailed for information only, in reality it would be at right angles to the direction shown - so that the propeller and it's housing face into the wind.

The second photo is of my prototype propeller - in the final job the prop will be of smaller diameter - about 20mm.

Having looked closely at just where the sensor could be, it would be just about possible to get a sensor 50 - 60mm from the axis of the prop.

Hope this helps.

 

Does the wind speed turbine (fan) rotate with the wind direction vane shaft, so it's always pointed into the wind?

Ken

 

Yes, that is the idea of having a wind vane attached, the whole thing rotates into the wind. There is another magnet in the base of the shaft that is measured by another chip to give wind direction. It's the speed that the prop measures and that is where my problem lies.

Thanks for looking

 

Why not mount the magnet into the face of the turbine, and the sensor to one of the "struts"(the three ribs supporting the hub)? Run the wires down the "strut" to the pole.

 

you could put the magnet on a blade tip and balance it with another magnet on an opposite blade tip. This would give you two pulses per rev instead of one, and would have minimal effect on the propellers response to low wind speed. It would also allow the hall effect device to be close enough to the moving magnets for detection without complicated amplification circuitry.

 

Problem with that is two fold:-

First the wires from the sensor will inhibit the rotation of the shaft so detection of wind direction will be difficult and secondly, the sensor would be out in the weather, not inside the sealed box.

In the final unit, the prop will be less than 20mm diameter so the struts will need to be VERY small so as not to influence the wind speed measurement. Attaching things to them is not possible I am afraid.

Nice idea though!

Thanks

 

you could put the magnet on a blade tip and balance it with another magnet on an opposite blade tip. This would give you two pulses per rev instead of one, and would have minimal effect on the propellers response to low wind speed. It would also allow the hall effect device to be close enough to the moving magnets for detection without complicated amplification circuitry.

This would be OK for a one off maybe but I would like to produce lots of these so plastic moulded propeller will be small and light and could not take magnets on it's blades.

Thanks for looking

 

Shortbus and Kermit2,

The fan has to freely rotate beyond +/- 360° because it's attached to the wind vane. The tube rotates...so no wires around the tube.
But... with the wires and sensor up the center of the tube, and the tube of a non-magnetic material, the sensor could be within mm of the magnets, and free of rotational limitation. Just thinking out loud.

Ken

 

Shortbus and Kermit2,

The fan has to freely rotate beyond +/- 360° because it's attached to the wind vane. The tube rotates...so no wires around the tube.
But... with the wires and sensor up the center of the tube, and the tube of a non-magnetic material, the sensor could be within mm of the magnets, and free of rotational limitation. Just thinking out loud.

Ken

Ken

Even so, the wires from the sensor will 'wind up' inside the tube as the wind vane rotates it into the wind. This will give resistance to the rotation and give faulse direction information. Also, how could I get to the wires when the bottom end of the tube is full with another magnet that is up close to the direction detection IC?

 

I assume that the shaft/tube is supported on two bearings. One the the top of the electronics box and one at the bottom. If the bottom bearing and tube where not touching the bottom of the box, a post with the sensor on it's top could place axially inside the tube, without touching the tube's internal walls. The sensor would be on top of the post, near the fan. The wires could be run down the inside of the post, so no friction on the rotating tube. The wind direction magnet(s) would need to be placed around the outside of the tube, to allow for the center post. It would all depend on the mechanical limitations you have.

Ken

 

Nice idea Ken. Trouble is I just cannot get there with the mechanical arrangement I have. The bottom of the tube has a magnet inside it and has to be very close to the direction detection IC which has to be on axis. This prevents anything from sticking up inside the tube - again, nice idea!

Thanks

 

I don't think the sensor(hall) needs to be outside of the tube. If the drawing you show is what you will be making, the sensor can be inside the tube and not 'rotate' at all. putting the magnets in the blades will work for more than a one off, plastic can be extrusion molded around anything and there are also wonderfully strong epoxy glues in these modern times. a smaller support rod inside the 'rotating' tube would not have to move at all, and could place the device(hall) extremely close to the edge of the blades shroud. Hope you can picture what I'm trying to say, it would not be impossible to do, but actually very easy.