A DIY Stevenson screen

[Gina]

What are the holes for in the wind vane part of the casing?

[JamesF]

2 hours ago, Gina said:

What are the holes for in the wind vane part of the casing?

They're just recessed holes for the bolts that hold it together.

James

[Gina]

I see.  Not printed in one piece then.

[JamesF]

I printed a mount for the USB power socket last night and today fitted it inside the screen together with an original RPi that I've just upgraded to the latest Raspbian.

Given an existing 12V supply for a mount, these USB power sockets would quite possibly be a neat and simple DIY way to provide power for an on-mount RPi, or indeed anything else that could be powered from a USB bus.

James

[Craney]

On 16/11/2019 at 18:22, JamesF said:

 

 

That looks a quality piece of work.   Make sure your Observatory wall is sufficiently insulated to avoid heat ingress into the screen.

Your thread prompted memories of when I built my own screen for a science based expedition to Northern Norway in 2012.   Yes, the prices of the official kit are as stiff as a force 6 ....

Edited December 26, 2019 by Craney

[JamesF]

When I took my rain gauge down it was completely blocked up, so yesterday I (hopefully) improved it by adding some spikes in an attempt to stop birds perching on the rim and (ahem) blocking the funnel.

James

[JamesF]

My anemometer and wind vane stopped working this week, as did the temperature sensor on the same board.  This morning I have taken it down and opened up the case to find the PCB is wet despite my attempts at sealing all the joins in the housing.  I assume the water must have been forced in during some of the torrential rain we've had this week, so I clearly need to make a better job of that.  I think the possibility of rain getting inside was always one of the weaknesses of this design.

Or I could design a better housing I guess, though at the moment I'm not sure I can see how I might avoid having to split it vertically which is (I think) part of the problem.

James

[JamesF]

Oh, on the positive side, my rebuilt rain gauge appears to be working a treat.  At the moment I just have a little script running that reads all the weather sensors and prints the data out in a terminal window.  The rain data is shown as totals for the current hour and current day.  On Thursday afternoon between 3pm and 4pm we had over almost 3/4" of rain!

James

[Gina]

24 minutes ago, JamesF said:

My anemometer and wind vane stopped working this week, as did the temperature sensor on the same board.  This morning I have taken it down and opened up the case to find the PCB is wet despite my attempts at sealing all the joins in the housing.  I assume the water must have been forced in during some of the torrential rain we've had this week, so I clearly need to make a better job of that.  I think the possibility of rain getting inside was always one of the weaknesses of this design.

Or I could design a better housing I guess, though at the moment I'm not sure I can see how I might avoid having to split it vertically which is (I think) part of the problem.

James

This is indeed a problem and I'm looking again at my structure.  Torrential rain can be blown upwards by storm force winds and I am on an exposed hillside.  Previous versions worked because they only used sealed reed switches with a multicore cable down the mast and into the observatory.  Unfortunately, the anemometer disintegrated - it was a commercial one, and I decided to make my own.

I'm considering a new design combined with my ASC but not sure about the windage affecting the ASC.  May be better on a separate mast.

Edited January 11, 2020 by Gina

[JamesF]

My anemometer was in a pretty sorry state too.  After quite a few years outside the plastic had degraded on the outside (I assume because of UV), but I've painted the entire thing with several coats of white gloss paint in an attempt to stave off the inevitable.

Rain being driven upwards is one of the possibilities I've been considering regarding how the water got inside because there's an opening at the top to allow the spindle for the anemometer to pass through and whilst it's pretty well shielded it's never going to be watertight.  I might try to make it more difficult to push rain into that space though.  Otherwise perhaps I need to be a bit more "heavy-handed" with the sealant.  I'd tried to make a nice neat job of it, but I guess that in doing so I'd spread the sealant too thinly.

Making sure the PCB still works once it dries out is of course the first stage of fixing things.  If it is dead then everything else is somewhat moot.

James

[JamesF]

Well, the bad news is that I think the weather vane and anemometer are dead

I've dried out the PCB and cleaned off the corrosion.  When I plug my 1-wire<->USB converter into my desktop I can see the 1-wire interface device inside it appear on the PC.  But when I then plug the weather station PCB into that, everything disappears

The PCB is truly 1-wire in that it just has ground and data connections, so I'd have to guess that somehow it is messing up the data path such that not even the 1-wire device in the converter can communicate.  Could be shorting the two out, I guess, or maybe just introducing enough noise that communication isn't reliable.  I'm not really sure there's anywhere to go from here though.

Oh.  Unless I happen to have got unlucky and connected it using a dodgy cable...  I'll go and find another.

James

[JamesF]

It lives again

I had to make a new cable in the end and having done so the PCB has come back to life.  Whilst it can't tell me how windy it is on my desk at the moment, it is at least telling me the temperature is just over 19.5C.

I am wondering if I can find something to waterproof the board.  Most of it appears to have been done already, but the corrosion around the connectors makes it clear that bit was missed entirely.

James

[Gina]

There a spray on waterproofer I used but can't remember what it's called.  Don't know where it is either.

[JamesF]

I couldn't find anything to help waterproof the PCB that wasn't silly money, so I've just reassembled it and tried to get sealant well into the joins by loosely assembling it, applying the sealant and then tightening it up fully.  I'm also tempted to put sealant over the screws holding the case together so they can't provide a path for moisture to get in.

Other than that I don't think I can do much more.  If it leaks again I'll just have to look at making a new case or have a complete rethink about the design, perhaps changing to a solid state system that might be easier to protect from the (unwanted) elements.

James

[JamesF]

Whilst finishing off reassembly of the cups for the anemometer I realised that rain could possibly get in around the washer and nut that holds them onto the spindle and then run down the spindle onto the PCB, so I've put some sealant around those too.

It's now back up on the post and seems to be working ok thus far, though I'm probably going to have to wait for another deluge like the one we had on Thursday afternoon before I have any idea whether it will continue to work.

James

Edited January 12, 2020 by JamesF