MadModder

Gallery, Projects and General => The Design Shop => Topic started by: John Hill on March 07, 2017, 12:45:51 AM

Title: Absolute encoder for anemometer direction indicator?
Post by: John Hill on March 07, 2017, 12:45:51 AM
I need to make a direction sensor for a replica WWII airport control tower and all the synchros/selsyns I have are too big and clunky for the lightweight mast.

So, here is the idea.  Make a wind vane in the traditional pattern with a vertical shaft and a (slightly) off centre steel disk mounted on the end of the shaft.  A hall effect sensor is mounted close to the edge of the disk and sees a varying gap as the wind vane revolves with changes in wind direction.  A second hall effect sensor is mounted 90 degrees from the first.  The two sensors are connected to an Arduino.

Now the questions,  are two sensors enough to read absolute position of the disk and hence the wind direction?

John
Title: Re: Absolute encoder for anemometer direction indicator?
Post by: philf on March 07, 2017, 04:21:55 AM
Hi John,

I'd say no it wouldn't work - there would always be two possible positions up to 180 degrees apart giving the same result. Do Hall Effect sensors give an analogue output? I've only used them as a digital sensor for speed or position sensing.

Phil

Edit: Sorry - I was talking utter crap when I said there would be two positions giving the same result. I also found some analogue output Hall Effect sensors which I didn't know existed. However they are looking at a magnetic field and your off-centre disc would have to be magnetised axially so the polarity didn't change as it rotated. The maths involved are also quite complex. How did you propose calculating the angle from the two measurements in the Arduino?
Title: Re: Absolute encoder for anemometer direction indicator?
Post by: awemawson on March 07, 2017, 04:28:35 AM
What is wrong with a standard encoder based on a slotted (or reflective) disk. Quadrature output will give you up/down counts (CW and CCW in your case) to apply to a standard counter and a reference marker can zero your counters at (say) true north  :scratch:

Title: Re: Absolute encoder for anemometer direction indicator?
Post by: John Rudd on March 07, 2017, 06:17:51 AM
Gray encoding?
I think epe magazine did a weather station project that used Gray coding for the weather vane, but it was some years back.....

I came across this too.....might be food for thought...

http://sandaysoft.com/forum/viewtopic.php?t=4570
Title: Re: Absolute encoder for anemometer direction indicator?
Post by: John Swift on March 07, 2017, 08:00:20 AM
Hi John

EPE Dec04  wind indicator
Title: Re: Absolute encoder for anemometer direction indicator?
Post by: Joules on March 07, 2017, 08:12:11 AM
It can also be printed and wrapped round a cylinder or shaft to make a much more compact sensor.
Title: Re: Absolute encoder for anemometer direction indicator?
Post by: gerritv on March 07, 2017, 09:07:28 AM
If you need more resolution, this article describes how to generate larger numbers and decode back to something usable.
https://en.wikipedia.org/wiki/Gray_code

Gerrit
Title: Re: Absolute encoder for anemometer direction indicator?
Post by: John Hill on March 07, 2017, 10:37:25 PM
Thanks for the comments.

John
Title: Re: Absolute encoder for anemometer direction indicator?
Post by: John Hill on March 12, 2017, 07:12:46 PM
The first tests...

(https://c1.staticflickr.com/1/767/32563548304_c6a6b3e71a_n.jpg)

Somewhat like a sine wave but not perfect,  maybe there is scope for tweaking the shape of the pole piece?

(https://c1.staticflickr.com/4/3826/32563554234_1a27d03d72_n.jpg)
Title: Re: Absolute encoder for anemometer direction indicator?
Post by: BillTodd on March 19, 2017, 05:27:35 PM
I've posted this before but it is such a simple clever idea (not mine obviously ) I thought it worth repeating:
Title: Re: Absolute encoder for anemometer direction indicator?
Post by: hermetic on March 20, 2017, 02:53:29 PM
 Speedometer cable? Simple and reliable!
Title: Re: Absolute encoder for anemometer direction indicator?
Post by: John Hill on March 20, 2017, 04:36:08 PM
BillTodd,  that might have been from one of my ideas!  The problem I found in advancing that principle further was the difficulty of designing thoroughly reliable slip rings. 

Speedo cable? Well not so daft when you consider that some 'professional' systems used solid rod drive to the indicator with Hook's joints etc.  Dynes make pressure tube systems with a pitot head and a rather large container filled with water!
Title: Re: Absolute encoder for anemometer direction indicator?
Post by: raynerd on March 20, 2017, 05:41:11 PM
Do you need absolute resolution. I made an anemometer with an arduino but not direction. If you didn't need perfect direction could you not use a series of Hall effect sensors and display the last trigger as the direction. I appreciate this is only an approximate direction but would be good enough for N, NE, E, SE, S, SW, W, NW display with 8 sesnsors. Sorry, probably a silly too simplistic suggestion but thought I'd mention it!
I struggled calibrating mine, have you considered this yet? I ended up with an arbitrary wind speed value which day by day was comparable but struggled to calibrate it with actual wind speed.
Title: Re: Absolute encoder for anemometer direction indicator?
Post by: John Hill on March 20, 2017, 05:53:33 PM
Hi, this is for our replica WWII control tower and will feed multiple displays.  I would use a Selsyn transmitter but I do not have one available and besides it would be 70 years old if I did have one.  I am going for the best possible solution bearing in mind that I do not want sliding contacts of any kind up the mast and I want it to be still going after I am nailed in my box.

I am just about at the stage to mount three hall sensors in the direction sensor then the challenge will be software for an Arduino to simulate selsyn signals to drive the indicators I have on hand.  If the old indicators dont work I will use steppers for the direction displays,  I have to use a stepper in the wind speed displays too as I do not have any 300 degree meter movements.

My idea for calibration of wind speed will be to count the pulses from the sensor (it has a disk and photo diode sensor) over a period of time while running another 'professional'  anemometer of the 'distance run' type.  That will give me a somewhat average value for wind speed.

The stainless steel anemometer cups arrived yesterday  so everything is coming together nicely!

Title: Re: Absolute encoder for anemometer direction indicator?
Post by: BillTodd on March 21, 2017, 07:01:29 AM
Quote
BillTodd,  that might have been from one of my ideas!  The problem I found in advancing that principle further was the difficulty of designing thoroughly reliable slip rings. 

It would not be too difficult to convert the reed switches to LED + opto sensor thus allowing the electronic parts to remain stationary while the light is directed by plastic guides or mirrors.

Or you could make the slip ring from a capacitive (touch) sensor ic