Using Cumulus with a rotatable AWS (on a windmill)

[Stuart2007]

As a newbie to AWS, I am exploring options, preferably using Cumulus, for an online AWS which may end up mounted on a windmill. To test this out, I have just bought the pre Christmas offer on a Clas Ohlson WH1080 Fine Offset.

Problem is, the mill is routinely rotated to head into wind, so the wind bearing from the AWS won't be relative to north. I can read the "Mill Heading" electronically eg using a robot compass module, so in principle can use software to do the simple arithmetic (CORRECTED wind direction = wind direction + Mill Heading). But how best to do this, and can it be integrated into the data recording and webpage creation by Cumulus.

I've been reading here and in the Wiki but not come across anyone trying to do this, yet.

First thought is, can Cumulus accept a 'trick' compass deviation - for example something like [true deviation+180] to indicate the mill is pointing due south not due north?

Alternatively, from reading this forum, I know the webpage templates can be customised. But can I use Cumulus to create the pages, but not upload them, leaving that to some custom software which corrects the wind direction, adds info on Mill Heading, then does the uploading?

Or even, use my software to repeatedly generate the template, writing in the corrected data every minute, so Cumulus picks up a template which ALREADY has some data in place...but which will lack the <#bearing> and similar webtags for wind direction.

I'd prefer to base a solution on Cumulus but may end up having to consider the 'roll your own' approach eg looking at a Raspberry Pi/pywws solution like http://www.weather.dragontail.co.uk/index.php.

I shall start testing some of these ideas shortly, but would welcome any suggestions to avoid me heading off down too many blind alleys!

[steve]

Cumulus allows a calibration figure to modify the wind bearing. But this figure is only read at start up (or when the calibration setting screen is used), so it's probably not practical to use this for your purpose.

[AllyCat]

Hi Stuart,

It depends if you want a purely "software" solution and/or using off-the-shelf components, or if you're prepared to get your hands dirty with a soldering iron, etc..

Given that the Fine Offset direction sensor is so poor (very "lively" and in practice usually only 8 directions are resolved - see the numerous posts in the FO section), perhaps a solution would be to use your "mill facing direction" AS the wind direction sensor?

The FO transmitter just reads one of 16 resistance values (8 resistors with 1 or 2 in parallel for each direction) as described for example in the "Photos of the insides..." thread. I have (very) long-term plans to use a PICaxe microntroller to read, process and re-insert wind data directly into the A/D converter of the FO transmitter. But I'm afraid that project is currently "stalled" by more important projects (currently one concerned with my Central Heating system ).

Cheers, Alan.

[Stuart2007]

Cumulus allows a calibration figure to modify the wind bearing. But this figure is only read at start up (or when the calibration setting screen is used), so it's probably not practical to use this for your purpose.

Thanks Steve - would my suggested 'custom template' approach*, updated every minute or so, be a viable option?

Stuart 2007

*"use my software to repeatedly generate the template, writing in the corrected data every minute, so Cumulus picks up a template which ALREADY has some data in place...but which will lack the <#bearing> and similar webtags for wind direction."

[Stuart2007]

Hi Stuart,

It depends if you want a purely "software" solution and/or using off-the-shelf components, or if you're prepared to get your hands dirty with a soldering iron, etc..

Given that the Fine Offset direction sensor is so poor (very "lively" and in practice usually only 8 directions are resolved - see the numerous posts in the FO section), perhaps a solution would be to use your "mill facing direction" AS the wind direction sensor?

The FO transmitter just reads one of 16 resistance values (8 resistors with 1 or 2 in parallel for each direction) as described for example in the "Photos of the insides..." thread. I have (very) long-term plans to use a PICaxe microntroller to read, process and re-insert wind data directly into the A/D converter of the FO transmitter. But I'm afraid that project is currently "stalled" by more important projects (currently one concerned with my Central Heating system ).

Cheers, Alan.

Thanks Alan - I have my soldering iron out, tweaking those pesky reed switches up towards the magnet...

Stuart

[steve]

Thanks Steve - would my suggested 'custom template' approach*, updated every minute or so, be a viable option?

Yes, that would work. You could also use javascript to calculate the bearing from the cumulus figure and a figure you supply yourself, perhaps.

[beteljuice]

You say "windmill", but that doesn't tell us the size or the build.
Do you mean Ye Olde Dutch Windmill design where the entire body rotates, or do you mean a turbine 'head' mounted on a mast or tower ?

For weather readings the anemometer and direction vane should be approx. 10m above ground level. However; they shoud not be "shaded" by either the structure or the blades.

On static tower or mast designs they are usually mounted on a stand-off of at least 400mm, on larger structures there would be two (or more) diagonally opposed and the lower reading ignored (In structure shadow).

Again, you don't say if the windmill can report any information back to a PC and / or generate a CSV data file or beter yet html. In both cases should be do-able without too much pain

[Stuart2007]

You say "windmill", but that doesn't tell us the size or the build.
Do you mean Ye Olde Dutch Windmill design where the entire body rotates, or do you mean a turbine 'head' mounted on a mast or tower ?

For weather readings the anemometer and direction vane should be approx. 10m above ground level. However; they shoud not be "shaded" by either the structure or the blades.

On static tower or mast designs they are usually mounted on a stand-off of at least 400mm, on larger structures there would be two (or more) diagonally opposed and the lower reading ignored (In structure shadow).

Again, you don't say if the windmill can report any information back to a PC and / or generate a CSV data file or beter yet html. In both cases should be do-able without too much pain

A post mill, whole body rotates. 10m would be achievable. Mains power available, but not broadband so contemplating using a 3g mifi. Thanks for the point regarding use of twin sensors with the lower reading ignored.

[beteljuice]

Thinking aloud ..........
Are you really interested in 'weather', or do you want to report yawing winds (with respect to the mill) ?

ie. If the direction vane is lined up with the sail rotor (regardless of North) then it will report the plus /minus degrees* of the wind to the sails

*With the FO style kit you have bought direction is only truly 8 sectors resolution with a further 8 "iffy" intermediate points.
Is this accurate enough for your needs ?

[Stuart2007]

Thinking aloud ..........
Are you really interested in 'weather', or do you want to report yawing winds (with respect to the mill) ?

ie. If the direction vane is lined up with the sail rotor (regardless of North) then it will report the plus /minus degrees* of the wind to the sails

*With the FO style kit you have bought direction is only truly 8 sectors resolution with a further 8 "iffy" intermediate points.
Is this accurate enough for your needs ?

Hi beteljuice

Well, wind speed and 'jaw' are first priority, mill heading second and other factors 'nice to have'. I'm now up to speed regarding the 8/16 point FO resolution issue. Can you tell me what resolution Davis gives for wind heading? In practice +-10 degrees is OK (so 16 points is about OK, 8 not very good). It's tempted to build a replacement PCB with 16 not 8 reeds (just about room on the PCB), to give a more accurate 16 point resolution. The issue may be that this will give good resistances for the 16 points, but introduce unexpected values for the intermediate 16 points were multiple reeds activate... I've lifted the reeds closer to the magnet, and it is noticeable that the vane now tends to settle on one of the 8 points due to magnetic attraction to the reed switches.

After Steve's feedback, I'm now thinking in terms of a custom Cumulus template to incorporate the Mill Heading info, either based on a manual parameter or an electronic magnetic field sensor.

Thanks for acting as a sounding board!

[beteljuice]

One of my concerns is how are you going to prevent the direction vane / anemometer being 'shaded' by the sweeps or being affected by swirling eddies.

Having the anemometer as far back as you can will probably give near enough (but reduced) speed readings, but I think the FO direction vane might literally lose the plot.

The more I think about it the more I believe you will need to mount the kit on a suitable mast or pole in the vicinity or perhaps a long rearward stand-off.

Then again just try it - might be OK

[BCJKiwi]

Would a forward standoff be better than a rear standoff - less turbulence??

[beteljuice]

"Would a forward standoff ..."

Think about it ...

[BCJKiwi]

Yes I know you need some way to mount it but we don't know how big this thing is or if there is a static central shaft.

[peterh]

When considering a forward standoff, you would have to take into account that those vanes, at anything above 2 Bft, are going to ferociously deal with anything that gets in their way, including forward standoffs, however you interpret these. Since, depending on the mill type and location, these vanes may come all the way to ground level, that could limit our options for a forward standoff

I'd agree with betelgeuse - you would want to steer clear of the windmill altogether.
But I see another problem... windmills tend to be the highest object in their immediate surroundings. As a rule, a miller is genuinely pissed off when someone plans a higher building in his immediate vicinity, because that potentially puts his mill in a wind shade. So you'd have to come up with a bloody long mast to mount your wind sensors on and I think that's why the topic starter wants to mount his stuff on top of the mill. Which is where the direction vane would loose its orientation, aaaand we're back to square one.

A central static shaft would solve our problem, but I've never seen a windmill design that uses a central shaft. I think the dome rotates the same way as an astronomic observatory dome, with the bearing surfaces at the outside of the structure.