Gina's Observatory Roll-Off-Roof Automation

[Gina]

[Gina]

Just realised that I've forgotten a couple of inputs I need on the RPi to feed back to the indoor control system - limit switch states. Need to know for sure that the roof is closed properly or fully open.  Doesn't need extra Arduino lines though - it can be commoned up with the limit switch inputs to the Arduino.  Does need 2 more optocouplers and two more GPIO lines on the RPi.

[Gina]

I'm wondering though if I need to know the motor current remotely.  As long as it's roughly normal, I don't think I do.  The only dodgy condition is if the roof is jammed and the motor is taking excess current.  The Arduino sketch can take care of this but a fault condition would need reporting remotely.  At least this would mean only one information line about the motor current - Normal or Fault.

[Gina]

Forgot the rain detector too!!  Hope I'm not losing it - I seem to be getting very forgetful!

With the Arduino Nano being powered through the Vcc terminal rather than Vin there is slightly less danger of a fault causing destruction of the RPi and I'm wondering if I need all those optocouplers.  If I were to feed the RPi output lines into the Arduino Analog inputs there is little chance of +5v being fed back to the RPi to kill it.  Inputs to RPi could be fed through a resistive divider or maybe a Schottky diode pulling the input to Gnd and blocking any positive input current.  This would simplify the RPi to Arduino interconnections considerably.

Another point I've thought of is the remote Abort signal.  There could be a conflict between the Rain Detected signal and the remote Abort.  I was thinking an Abort is necessary for safety but an Abort signal could be caused by a fault and prevent roof closure when rain is detected - we don't want that.  I think rain detection should take priority and Abort only recognised following a manual opening or closing.

Edited August 7, 2019 by Gina

[Gina]

New diagram.

Edited August 7, 2019 by Gina

[Dr_Ju_ju]

Make sure all the GND's\0v's are connected together, or you might still need the opto-couplers...

[Gina]

Been searching out the input impedance of the Arduino Analog inputs and it seems to be 100MΩ in parallel with 14pF so that's fine.  For extra protection I could include Schottky barrier diodes to prevent current being fed into the RPi as I have proposed for the RPi inputs.  I have several BAT43 diodes - 30v 200mA.  Though with the logic 1 level being 3v, any ordinary signal diode would do.

[Dr_Ju_ju]

Its' just that some of the buck converters I've had, 'fixed' the +ve rail and raised\lowered the 0v line thereby producing a potential difference between the incoming 0v\GND and the output, but with the correct, regulated output....

[Gina]

Mine have a common -ve rail - a solid strip of copper from input to output.  I only connect one end to avoid any earth loops.

[Gina]

Here's a modified diagram showing the buck converter better.

[Gina]

The rain detector is a very strange unit.  When rain starts it detects it as long as it's heavy enough but after a while it stops reporting even the heavy rain we have at present.  That is actually alright as once rain is detected the roof will be closed and stay closed.  I can only guess that initially the rain is a bit polluted (or there's dirt on the sensor) but later the rain has washed the pollution out of the air (or off the sensor).

[DaveL59]

have you figured yet how many sensor bars need water bridging them in order to trigger? Could be why it doesn't detect light rain so well and not having any sensitivity adjust makes it a lot harder to tweak for your requirement.

I know it's not ideal but are you able to see how the water is laying on the sensor once its wet? Is some sort of coating affecting it from bridging the bars perhaps? Or is it maybe figuring short so stop triggering...

Hard to figure much without a cct diagram or adjust much without access inside - is it potted or just in an openable IP66 box?

[Gina]

At present the rain detector is on the roof but I can place it lower down so I can see it.  I'll do that tomorrow morning if the weather forecast is right and it's fine.  Meanwhile, here a quote from the web site.

Quote

M 152

If the sensor plate comes into contact with rain or slushy snow/hail, it switches on a relay.

This allows you to retract awnings, close skylights or acknowledge that it's raining.

• The automatically heated sensor surface prevents any freezing or wetting of the sensor surface.
• 2 built-in LEDS indicate the function.
• Waterproof encapsulated electronics.

Technical data
Operating voltage 12 V/DC
Current consumption without heating approx. 8 mA
Current consumption with heating approx. 120 mA
Relay contact 1 x On, max. 25 V, 2-A rating
Sensor heating automatically when rain contact occurs
LED 1 Display that the rain sensor is in operation
LED 2 Display that the rain has been reported and the relay has switched
Relay switch-on duration as long as the sensor is wet
The module is encapsulated and waterproof.
Active sensor surface, gold plated approx. 29 x 30 mm
Dimensions approx. 64 x 44 x 36 mm

Note: The electronics of the rain sensor reacts to the electrical conductivity of the water. Now we have found that there are areas where absolutely uncontaminated rainwater falls (distilled water). The sensor does not respond to it. The rain must have a slight amount of contamination (dust particles, smoke, etc. ), so that the water is electrically conductive and triggers the sensor. In 99% of the areas in Germany, the rainwater is conductive. If the sensor does not work for you, please install it so that the rain water runs off a small canopy or something else before the water comes into contact with the sensor. If the water falls directly from the cloud in its purest form on to the sensor and it is not triggered, then it is sufficient, if the water runs onto the sensor via a small board or from a canopy. Then the water will have absorbed enough contamination to be electrically conductive and to trigger the sensor. Of course, install the sensor at an angle so that the water runs of it.

 

[Gina]

I think I shall try a tray beside the rain detector so that water runs off onto the sensor.  This would still allow rain to fall directly onto the sensor as well.  I'll try some ideas in the next few days of showers.

[DaveL59]

ahh ok, that's kinda limiting if it needs contaminants. Perhaps roof run-off from the roofing felt or down the guttering would be more effective to trigger, tho that'd also mean v light drizzle/rain may take a bit longer before it fires. A plastic tray may not hold much contaminant, at least initially, texturing the surface might help it build some quicker tho 🙂

 

[andrew s]

14 minutes ago, Gina said:

I think I shall try a tray beside the rain detector so that water runs off onto the sensor.  This would still allow rain to fall directly onto the sensor as well.  I'll try some ideas in the next few days of showers.

Dew forming on the tray may be an issue if it drips off onto the detector.

Regards Andrew 

[Gina]

Yes, there's a problem of detecting rain/drizzle but not dew.  It may be a question of getting the temperature right - just above dew-point.

Edited August 8, 2019 by Gina

[sloz1664]

Note: The electronics of the rain sensor reacts to the electrical conductivity of the water. Now we have found that there are areas where absolutely uncontaminated rainwater falls (distilled water). The sensor does not respond to it. The rain must have a slight amount of contamination (dust particles, smoke, etc. ), so that the water is electrically conductive and triggers the sensor. In 99% of the areas in Germany, the rainwater is conductive. If the sensor does not work for you, please install it so that the rain water runs off a small canopy or something else before the water comes into contact with the sensor. If the water falls directly from the cloud in its purest form on to the sensor and it is not triggered, then it is sufficient, if the water runs onto the sensor via a small board or from a canopy. Then the water will have absorbed enough contamination to be electrically conductive and to trigger the sensor. Of course, install the sensor at an angle so that the water runs of it.

Once you affix your rain sensor in position, air pollutants will settle onto the sensor. If, and it's a big if, in this country "pure" rainwater hits the sensor it will mix with the contaminents anyway.

Steve

[Gina]

Yes, that seems to be what's happening.

[Gina]

"Oh what a beautiful mornin'!"  🎶  Ha ha.   Neither of those in fact - tongling down with rain, blowing a gale and it's not morning either!

So why am I happy?  Because I've just seen my doctor and got a clean bill of health with a recommendation to increase my salt intake a bit as I've gone too far in cutting down.  That's why my blood pressure is too low, causing dizzy spells and lack of energy.  Hopefully the answer to my problems health-wise. 😀

Edited August 9, 2019 by Gina

[DaveS]

That's good to hear Gina.

[sloz1664]

Excellent news Gina

Steve

[Dr_Ju_ju]

👍  Great to hear it, just don't go too far the other way......

[Gina]

Don't worry, I wont.  I shall continue checking my BP.  I plan to increase my salt intake gradually and watch what happens.  This morning's readings were 121/65 and 130/47 and I'm feeling tired.  I know that any change of diet will take a while to take effect so I'm not expecting an immediate cure.  I shall continue to pace myself according to how I feel.

[DAVE AMENDALL]

I don't know if you can still get salt tablets. They were issued to us when I was living in Malaya. People took them by the handful before meals.They were affective and economical................Dave