Chook-House Project, another new thingy...

M

manie

Senior Member
Since I felt left behind by Stan et al, I decided it was time (litterally) to look at checking wind direction, wind speed was already mentioned..

I have tried my hand at this device, totally home brewed with the help of a friend who can do turning and milling (am i lucky ?).

Basic idea:
1. Use 4 digital input pins on the 40x1, pins kept low by usual 10K's to 0V
2. Have "high" +- 4.5V signals come in via micro switches.
3. Micro switches operated by a 5/8th round cam (3/8th's = straight face)
4. Have 4 switches but be able to indicate 8 wind directions (good enough resolution for this application).
5. Read the binary value on the 4 pins to know wind direction.

With this device, either 3 switches at a time, or 2 switches at a time will be "closed" to put a high signal on the corresponding pin. With 3 switches closed, the Picaxe can assume wind is comming from a compass direction corresponding to the "open" or "low" switch, either N/E/S/W. With 2 switches closed, the wind is comming from somewhere between the two "low" or "open" switches.

Attached are two pics to demonstrate. The first pic is of the top side with the bearing which will carry the vane on a 10mm threaded rod. The bearing centre (shaft diameter) is 30mm. The bearing is +- 60mm diam.
The second pic is the business end with the cam and 4 micro switches. The position of the "flat" on the cam in the pic will indicate the wind is comming FROM between those compass direction switches.

As always, comments/suggestions/improvements are by now almost mandatory ? Thanks for input.
Manie
 

Attachments

  • Direct1.jpg
    Direct1.jpg
    69.8 KB · Views: 74
  • Direct2.jpg
    Direct2.jpg
    69.9 KB · Views: 94
Well, it's nice and shiny manie.

If it works then fine.
But have you ever wondered why we moved to optical and magnetic sensing?

Looks like you are squashing those microswitch levers a bit(?)
Maybe one day when the wind changes those levers will stay bent?

Anyway, I'll bow out to the million wondericulousful suggestions that are bound to follow...
(We'll see if we can get Andrew up to his kilo-posting in this thread :) )
 
Dippy: Switch no.1 lever does seem a bit strained, but is not really, It was the first switch to go down on some glue and it did shift ever so slightly. The bent ends are to ensure the cam does not snag on a sharp or blunt end. Optical switching yes, I thought of it, but then you need a light tight box with a moving curtain etc. Quite a bit more involved than this arrangement. HOWEVER, me and my lathing/milling friend are working on one, if it is the better one then that is what I'll use in the final setup. Any thing else you can think of that might foil the reading of wind direction ?
Manie
Edit:
Come on Andrew, don't dissappoint now !
 
Under the Readtemp command in Manual 2, page 142, the following is stated:
This command cannot be used on
pin0 or pin3 of the PICAXE-08M/14M, or pin6 of the 20M.
Click to expand...

Question: Can it be used on the 40x1 on pins portc 4 or 5 (physical pins 23 or 24) ? I would like to read the wind direction on input pins 0/1/2/3.
Manie
 
nice job, sometimes the simple ideas are quite satifactory, especially if high resolution isn't required..well done
 
manie said:
Under the Readtemp command in Manual 2, page 142, the following is stated:


Question: Can it be used on the 40x1 on pins portc 4 or 5 (physical pins 23 or 24) ? I would like to read the wind direction on input pins 0/1/2/3.
Manie
Click to expand...

Sorry mate, but you must use the "standard" input for the 40X : PortD
(I've just tested, on the simulator, and importantly, on
a breadboard. ;-)

Alternatives?
If it's simple digital for the anemometer, then 16 alternatives.

PortA and PortC

e
 
Quite a way to 1000!

Microswitches have springs in them, which will cause friction as the wind turns the shaft. However, this effect could be reduced by using a larger wind vane - what size vane are you going to use?

In fact, the friction may prevent it spinning due to small changes in wind wirection, which will improve its accuracy (although remove precision).

That is the advantage of optical (and slighty of magnetic) systems - they add little/no friction. That looks a bit like a ball bearing in the first photo - if it is, that should make it spin more freely.

A
 
I too would have concerns about the friction caused by mechanical sensing such as uSwitches. Also, very limited resolution. Even if you make the cam opperate both switches when at the mid point.
How about a nice simple encoder wheel. Even just four sensors (same number of inputs) would give 16 positions. Only two more would take that up to 64 positions.
 
Manie-aside from mechanical switching issues, IMHO the nature of your local wind needs considering. Is it gusty & continually shifting direction, or a more reliable trade wind? Day-night breezes present? What kind of resolution do you need anyway? Are you hoping to use the wind for ventilation?

Suggest you look at techniques used by off the shelf weather stations. Stan
 
Thanks for input all.

Yes, its a ball bearing, so very little resistance there.

The vane itself would be about 1m (3ft) long, plenty of leverage, this one spins between two fingers its so light.

Resolution ? I basically need to know if the wind is A. South(erly) or B. North(erly).

Ventilation by means of wind yes, no fans.

If wind is South(erly) then South-side curtain = less open than North-side and vice versa.

Nature of wind ? You tell me, its on the "Highveld" (5500+ ft above sea level). In winter, definitely Southerly to SE'ly, quite strong gusts and cold ! In Spring, sort of NW'ly, a little gusty. In Summer mostly N'ly to NE'ly, more steady. I'd say more steady winds than gusts bouncing around. There will be quite a bit of cam/vane movement due to local eddies etc.

I did look at the encoder from e-bay mentioned in another post, would be nice but TOTAL overkill for this. With 8 direction resolution possible here it is almost overkill. Currently, wind direction and strength is measured by "wet finger" method and curtains cranked by hand. It works, so the 1 to 10 second measurements and adjustments possible with the 40x1 is "one giant leap for chicken-kind".

Like I said earlier, I am working on an optical (LDR/LED) setup, which means a heavy duty ball bearing as the only moving part, will last forever ! But I still only need S'ly or N'ly indication, nothing more. Any more will be a waste as this is not a weather station but part of the "curtain raiser" decision tree only. Hope I am explaining it OK here ?

The anemometer is simple "high" or "low" signals, so fine, I'll use portc 0-3 then. Thanks.

Manie
 
So if you don't need a fast responce or precision, then microswitches sound like the way to go. Your vane sounds easily big enough to not be affected by gusts, and big enough to activate the switches.

A
 
Andrew: The threaded hole in the shaft is 10mm diam. Almost 1/2 inch. With a locknut to keep it down, it should be able to support quite a large vane, larger than the plastic housing can support anyways. I also thought the longer the better (to a point that is).
Manie
 
5500' = high velt indeed - are oxygen levels relevant too?! For delicate yet such simple wind direction needs maybe consider the "hot wire anemometer" idea- even though they're normally just for wind speed. Mmm- READADC using thermistors out in the chookhouse jetstream? Stan
 
Last edited:
My weather cock is quite small but always points in the right direction.

If I were to design one I'd make it fairly light with some damping.
Glad you're moving to optical.

Go for it manie.
You've obv got the engineering skills. That's the hardest bit. The code is a piece of cake as you know.
... you don't need people to hold your hand do you? ;)
 
@Manie

If I have understood the situation correctly there is(are) existing building(s) with adjustable openings on opposite sides (or is it all 4 sides?).
If the openings are only on the northernmost and southernmost sides, then all that is required is to decide which way is most favoured.

If a tube is FIXed across the ridge of the chookhouse so the centreline is at right angles to the ridge and pointing in the same direction as the wall with vents faces, then the wind speed fan can be mounted inside this tube.

The effective wind speed on the wall can be measured, and, by sensing the direction of rotation of the wind speed fan, which side it is coming from.

If four directions need to be managed, then repeat the setup along the ridge.

The only moving part is the fan.
 
Is it a big or poorly ventalated chook house? It's not something I worry about with the small house for our 4 birds... more worried about it getting too cold in winter but they where fine - they huddled up a bit!

There are a varitey of cowls that direct themselves into (or out of) the wind, would these give a better alternative? A quick image search for snorkel cowl chimney should show the extraction version that rotates out of the wind.

regards,
colin
 
Dippy: Funny you should ask.......(and some more.......) I am using portc pins 0/1/2/3 for the four wind bits. I got this error:
winddir = portc & %00001111
^

Error: #Directive error - unavailable #com port number
Click to expand...
The carat sign(^) in the error window sits under the space between "portc_^_ &". The pesky PortC again ! What does the error mean ? what is wrong with the command ?

Manuka: yes, O2 is down too, cars have 14% less power on the High Veld than at the coast ! You grow up there, you don't notice it.

BCJkiwi: Existing buildings, 30x20m with ventilation curtains (up/down) on South and North sides only. East/West are gable ends. The wind speed is a different matter in conjunction with BOTH direction and outside(wind) temperature. These three factors decides which curtain (N or S) moves most (up or down), which curtain is highest (least open curtain faces wind direction) and how much curtains must adjust to regulate temperature. All this again in conjunction with chicken activity (more heat generated when active). Quite a complicated scene !

Slurp: If I get this DARN PortC sussed out, the wind direction is the least of my worries ! If I can read it I can make a decision.....
Manie
 
Last edited:
@Manie.
Try something like this,
as a starter .......

Code: 
symbol windir = b0
main:

readportc windir

windir = windir AND %00001111

pause 2000

goto main


; manual 2 page 140

Man 2, p. 140

e
 
Thanks EC, that one did it ! What is wrong with the original command though ?
Manie

Edit: Ahhhhhh ! he exclaims, there is a separate command to get portc values, its called "Readportc" ! If you remember all that you read in the manuals, which I do, just not that page this grand evening !
 
Last edited:
Chook-House project

Why not try using green LED's as photovoltaic light sources? Frictionless, no mechanical wear and tear, etc. You could just have four or more green LED's set up in a ring around the shaft of the wind vane, and a slotted disk above them attached to the vane; the picaxe then could check them for voltage, i.e., daytime wind direction:

for day and night time wind direction, maybe cad sulphide cells in a ring with a source of illumination on the shaft; picaxe checking for resistance change;

Just my $.02

Here's someone who uses the LED priciple for heliostats:

http://www.redrok.com/electron.htm#led3x
 
TravisH: The next unit will DEFINITELY use LDR's and LED's ! Quality of the micro switches I got from Rev-Ed (sorry Tech Supplies) are not very good. New switches installed and they are either O/C or intermittent, a pity. I'm working on the next unit which will be optical...
Manie
 
Replaced one micro switch with only spare I had (bought five). Operated by hand simulating gusting through 90 degrees etc. Works fine, serout gives actual direction eg. "NE-" etc. Another one down and done.... Wind speed next, Rel.Humidity after that, then if I can get gas sensors or gas sensing modules I'll add that. Then comes motor control for curtain/shutter movement and audible alarms. At least the RF serial comms is working and out of the way. Will keep you posted.
Manie
 
Well it was Rel.Humidity first, wind speed will be next. Updated the code also and tested again using Polygon 433MHz transcievers. Results:

This is comming together very nicely now. Temp. reported and request for new alarm setpoint sent/received.
Rel.Humidity also reported and request for new alarm point sent/recieved.
Wind direction reporting beautifully too.

So now its just wind speed and then gas sensing. Getting some feedback from some suppliers but the locals who import are three to four times the price I can pay direct in UK.

"Hanwei" is the manufacturer of the sensors sold by Futurlec... they are willing to sell low quantities (1's or 2's) for evaluation at between 3 and 5 Euro's each.

I'm still waiting for the LCD's from eastbay to arrive form Sure Electronics.

Stan: This Saturday I'll be doing range tests "on site" at the actual chook-house. Will keep you posted...
Manie
 
You must log in or register to reply here.
Back
Top