Arduino Based Weather Station

[Gina]

Here is a cross-section of the reflective version of the wind vane mounting.  Now designing a transmission version.  The previous one was before I had a 3D printer so a lot more difficult!

[Gina]

Here is a rough cross-section of the wind vane hub with separate IR LEDs and photo-transistors either side of the encoder disc making the light path much shorter than reflection plus no loss due to less than 100% reflectivity.  It does mean that more separate parts will be required in order to make it possible to assemble.  Not to scale.

[Gina]

Small change in design to take account of assembly.  The encoder disc (black) will just be a flat plate with slots plus holes for screwing it to the hub (yellow) that takes the vane shaft and magnet support.  Actually, the latter could be printed as part of the hub rather than being a piece of bent aluminium.  The IR LEDs will be mounted on a disc (green) which is screwed to the fixed part (blue) that is attached to the aluminium pan lid.  The LEDs will be connected in series and the two wires fed to the power supply (via resistor).

[Gina]

I'll solder the LEDs and photo-transistors to stripboard (aka Veroboard) so either 0.2" or 0.3" apart.  That's either 5.08mm or 7.62mm so the 3 wide would be 10.16mm or 15.24mm.  I think 5mm separation between tracks would be sufficient but I guess 7.6mm would not be beyond the pail and less chance of crosstalk from stray light.  The current system has the 0.3" spacing.  This would mean increasing the size a bit to allow for the extra hub space for attachment.

[Stub Mandrel]

IIRC (it's at the top of a pole) my encoder is built into some sort of two-part cosmetics box with two round halves that screw together. I must get it down and have a look.

Prototype water bucket works, but needs some sort of 'lip' or 'spout' to stop it holding back a couple of 'raindrops'. Any ideas on shapes?

[Gina]

Do you mean the bucket holds drops of water?

[Gina]

Here is the new design for the encoder disc.

[Gina]

And the plate that holds the IR LEDs.

[Gina]

Further thinking has resulted in a reduced LED plate removing what's not needed.

[Gina]

That's still more than required just to hold four little LEDs.  This is enough.

[Stub Mandrel]

13 hours ago, Gina said:

Do you mean the bucket holds drops of water?

There was a fair bit of water not draining from the tip of the lower end, but now I've used it with the base the water runs right off. I don't think I'll use the setting screws as it is quite sensitive, requiring about 2cc of water, and in its box it was happily running at about 1 tip a second which should keep up with a downpour.

I can't bear the thought of sending this machine back. Worst case I will buy a Prusa kit and build it into a more rigid enclosure.

[Gina]

Rain gauge looks good

The Prusa kit sounds like a plan - several people have them and have modified them to make them better but I have no experience of them myself.  Seems you need to spend a lot of money to get a good 3D printer but if you buy a cheap kit you can use it to make parts either to modify it or to make a completely new printer.  I have done both.  Making or modifying 3D printers can be a bit like astronomy - very frustrating at times!  As long as you stick to a RepRep type kit you can get parts dirt cheap.

Tekkydave has a Prusa kit and seems to know a goodly amount about 3D printers and I think I can say I know a fair amount too   So anything you want to know about making a RepRap printer...

Edited January 6, 2017 by Gina

[Gina]

Back to the my weather station...  The latest LED holder plate has the property of being able to clear the encoder disc on the zero position.  This may simplify the construction.  I think the LED holder plate can be glued to the fixed part (shown in blue above) and the hub can be assembled by sliding the parts together as long as the vane is in the "zero" position.  The otherwise exposed connections to the LEDs could be covered with hot melt glue to keep the weather out.

[Gina]

Here is a new cross-section model of the wind vane hub and encoder.

[Gina]

Model of the wind vane hub (yellow part above).  The top takes the encoder disc, the projection to the left takes the damping magnet and hole on the right, the aluminium tube with the vane.

[Gina]

Printing the Gray encoder disc.  Run out of black ABS in 1.75mm so printing in dark blue.  I have ordered more black but the delivery looks like being about a week.

[Gina]

And now printing the LED support plate.  Just a five minute job

[Gina]

I have another idea for the encoder.  Turn it from a disc to a cylinder - just as easy to 3D print but has the advantage that both PTs and LEDs can be fixed to the base part and the encoder can be withdrawn without needing to be in any particular position.

[Gina]

I can take this a stage further so that the optical sensors can be standard slot type.  TCST1103 datasheet  These are extremely efficient and give a minimum collector current of 2mA in the ON state.  The gap of 3mm means plenty of room for the encoder and some tolerance.  The height of the hub unit is much reduced.  (Oh dear, this screenshot has suffered badly from pixel creep!  But I think the general idea can still be seen.)

[Gina]

These ITR9606 slotted sensors from Amazon are much cheaper and delivery is quicker (Monday).  Slightly bigger but I don't think that's of much consequence.  5mm slot rather than 3mm.  Minimum ON current 0.5mA which is still quite adequate.  Datasheet

Edited January 7, 2017 by Gina

[Gina]

Working this out shows the diameter of the hub unit will be just a bit more than the original.  I'll post screenshots of models shortly.

Edited January 7, 2017 by Gina

[Gina]

The base part.

[Gina]

The encoder.  Showing my workings and the final model.

[Gina]

In fact it doesn't need the full disc for support and some filament could be saved.  If it's worth it for a couple of pence worth

[Gina]

This diagram shows the placing of the optical sensors.  Two at right angles read the outer ring and the other two the middle and inner ring.  Separating them makes for a small overall design as it doesn't need 2x the side width of the sensor between rings.