Could a Raspberry Pi usefully help me create a wind direction meter with physical/analogue face and hands...

[JoeSmith]

Hi All

This wind direction meter in Coleton Fishacre in South Devon in the UK inspired me to the point that I've wanted to build one for years. My apologies for my poor video work but I think this shows the idea...the dial moves silently to show the wind direction. I need this because, due to the layout of our house, neighbours and trees etc we don't have an unprotected location where I could put a weather vane that we can see from inside the house. This means that in winter we can get caught out when it looks bright but there is, in fact, a sharp Northerly blowing and we have two dogs who like to walk a long way across moors.

Sharing Album

New item added to shared album

photos.app.goo.gl

This particular system runs via magnetic switches relayed to a motor on the back of the dial via radio waves and is a beautiful piece of kit and horribly expensive.

Initially I started down the route of a mechanical system that uses cogs, rods and gears to transfer the movement of a weather vane to an indicator dial on my kitchen wall. Ultimately this proved to awkward due in no small part to the wonkiness of our chimney up which I was planning to run it.

So I asked on here a couple of times and "met" some fantastic and helpful people who gave me wonderful suggestions. This took me to the world of Selsyns which was interesting but they're quite hard to get hold of and pretty expensive and, if I'm honest, the electronics was a bit beyond me.

I'm wondering now if I can use a Raspberry Pi to process movements from a wind sensor like this one that I can site on top of my chimney and feed wires down it (it's no longer in use) to the Pi which I can use to run a stepper motor. I had considered a servo motor would be better but I was put off by the noise they make (at my budget level) and finding a silent one seems hard. That said, putting it in a "sound proof" box mounted further back inside the chimney and using it to power a longer shaft to the dial on the indicator's face in the room could be helpful as it would also help protect the motor from dirt in the chimney so I'm open to suggestions on this...

I'm completely new to the world of the Raspberry Pi. I had thought an Arduino would be better for this but I have a RP and don't have an Arduino.

My initial questions therefore are firstly whether this is a reasonably achievable thing to create with a Raspberry Pi? Would this give me the ability to "broadcast" the movements of the chimney top mounted sensor via Wifi to another dial in the sitting room if I wanted one? Is a stepper motor a reasonable solution or should I look out for a quiet servo or just "sound proof" an available servo - I'm after as smooth a sweeping hand as I can create? What other hardware might I need?

Any advice much appreciated.

All the very best

Joe

 

[kellys_eye]

a mechanical system that uses cogs, rods and gears to transfer the movement of a weather vane

What about a 'speedo cable' type solution? Mechanical coupling to a chimney mounted vane and reciprocated at the display position?

Tacho and Speedo Cables | Cable Tec Cables and Controls Ltd

www.cable-tec.co.uk

An electronic solution is fairly easy to resolve - with the right approach and skills - but sometime the KISS principle is better. I might be tempted to part with my synchro/display as I won on an eBay auction as this seems to give all the answers in one box! I'll post a few pics of it shortly.

 

[kellys_eye]

Here's the device per the previous post:

The main dial is only some 3 inches across but the bezel could be removed and a larger pointer fitted (along with a dial of your choice). Power and connections are made between the two units via the chock block - you need FIVE cores (the device came with some old CAT5 cable connected and it worked ok with that). 24V at the +/- terminals and connect 1-2-3 of the sender to the 1-2-3 terminals in the rear of the display unit. Of course you could run a 2-core for power and a 3-core for the 1-2-3 wires (2-core mains and 3-core mains would be ideal).

Voila! Turn the 'bar' on the sender (which you couple to your weather vane) and the dial follows - accurately.

I've tested the unit and it works perfectly.

If this is of interest then PM me with an offer!

 

[JoeSmith]

Hi @KellysEye
Many thanks for this and for your kind offer. I did investigate the speedo cable approach to this a while back but was concerned that our house is thatched and the cable method could increase the risk of a lightning strike and/or fire. I think a data cable is potentially safer in this respect. I did go into a bike shop locally to ask about such things and was openly laughed at by a group of instant experts who said very confidently that it won't work. I took from that that it will but I will never re-enter their shop again!!
The other reason for going the SP route was that I'm thinking in the future I can extend the system via wifi and add some cool, funky stuff which will turn this into a real hobby that I'm quite excited about.
But you're right, KISS is much easier. There's a beautiful windicator at The Hunting Lodge in Charlton, West Sussex, UK that runs entirely mechanically. It's never been hit by lightning either so I'm keeping that in reserve!
I'll try the SP route anyway. I have a feeling I saw your setup on eBay when it came up and was tempted to bid!
Thanks again.
All the best

 

[kellys_eye]

To do it using discrete (perfectly feasible) and/or Raspi components isn't that difficult. You can source what are called 360-degree potentiometers that will deliver a proportional voltage that is readily repeatable either via conversion to a digital signal or used directly in an analogue form - perhaps a ring of LEDs driven by a chain of LM3914's(?) as but one idea.

You need to choose and settle on your solution before we can go around the houses again on this subject.

Decide (1) how you want to display the result, (2) whether you're happier with a digital or analogue solution (don't over-think it. It's too easy to imagine what you 'could' do with it rather than what you just 'need' to do).

When (1) is resolved let us know and we'll expand on what's required to achieve it.

 

[JoeSmith]

Thanks for that. I'll look into 360 degree potentiometers.

My plans are as follows...

Create an analogue compass face with painted bearings with a single, physical needle that is wall mounted above the fireplace in my kitchen. The needle will be coupled to a stepper motor/potentiometer/servo motor that is connected via Raspberry Pi (if applicable) to an RS485 wind sensor that is attached to the top of the chimney above.

I'd prefer it if the needle on the dial can "sweep" (like the second hand of a Rolex wristwatch) between bearings rather than "jump" between bearings on the dial (like a quartz Timex watch's second hand) as the wind direction changes. I think this requirement could dictate, to an extent, what type of motor drives the needle. I also want the needle to move silently.

In the future I'd like to be able to drive additional analogue dials in other rooms by broadcasting the signal from the RS485 to receivers on them. This is an advantage of using the Pi I guess.

Any input on the above is most welcome.

 

[kellys_eye]

Your best (only) choice is to use a stepper motor and the associated controlling circuit/software.

Adafruit do a small steeper as per this link.

Automotive Gauge Stepper Motor

This stepper motor is a little different than the large NEMA-17 types you may be used to. These are often used in gauges for motorcycles and cars to replace the old-style fully-analog type. ...

www.adafruit.com

The short video therein shows the needle moving 270 degrees but it is quite capable of 360 degree rotation - it's all down to the software. The size/weight of your pointer may be the determining factor for this particular motor as it can't wield anything too substantial (inertia). Regular steppers (NMEA series) can, of course, drive 'any' load when the right version is used.

That sorts the display. You then need some way to 'send' the wind direction data to the controller/display and that could be via the 360 degree potentiometer (using an analogue port) or something somewhat more complicated i.e. a synchro resolver.

There's nothing quite so simple as the synchro option though. Ebay have a load of old gyro compass repeaters (dials) like this:

Ships Compass 7" Repeater Ex Royal Navy highly collectable (No1) | eBay

Find many great new & used options and get the best deals for Ships Compass 7" Repeater Ex Royal Navy highly collectable (No1) at the best online prices at eBay! Free delivery for many products.

www.ebay.co.uk

but they require the synchro transmitter to match - something like this:

Muirhead Vactric 11C1G1 Mod.A Synchro Control Transmitter 400Hz | eBay

Find many great new & used options and get the best deals for Muirhead Vactric 11C1G1 Mod.A Synchro Control Transmitter 400Hz at the best online prices at eBay! Free delivery for many products.

www.ebay.co.uk

You'd have to get the right pair though. Alternatively you can build an interface between an Arduino and the compass repeater by simulating the synchro output signal (3-off 8-bit D-A converters will do it).

The advantage of the synchro version is that there is no calibration required. You turn it on, it points to where it should. Any digital option will require a 'reference' signal so you know how far from 'zero' the pointer is. Another advantage is that you can buffer a synchro signal and send it to other repeaters very simply - but it does need a wired connection (disadvantage).

Best to start with the 'top end' - decide what you're going to use then come back to us for more details on where to move to next although such finery as the actual software writing (if required) will only be on advice-based response, not the actual writing of it.

 

[bidrohini]

Here I found the Arduino sample code and library:
https://wiki.dfrobot.com/SKU_SEN0482_RS485_Wind_Direction_Transmitter_V2
For Raspberry Pi, I think you will not find an existing library.