Low cost Cloud Detector released

Thanks for the tip to look at Phidgets. I picked up the 1045 and have it working on my bench (actually, in its box on my living room couch ). I have just ordered the Phidget SBC and intend to use it to monitor the 1045 and then trigger the MaxDome Aux 1 input to go low. This is supposed to make MaxDome keep the dome closed.
Question:
Will MaxDome connection to ACP notify ACP that Cloudy has been reported and the dome is closed?I know I will discover this but if it doesn't work I'd like to know it's either not supposed to work or that I have a bug to fix somewhere.
Is there another way I could notify ACP that it is cloudy?

I intend to add other Phidgets to the SBC. For now I would add an external ambient temperature thermocouple sensor (not the value report in the heated enclosure for the 1045). I would also add the light detector and humidity detector. I'd like to add an airspeed detector but haven't located a solid state product. (Let me know if you know of any solid state airspeed detectors.)

I will be glad to share my solution for a Phidget driven cloud detector. It's not as fancy as John's in that it does not include graphs but it should work.

Thanks for any comments or advice.

Bob

Hi Bob (hillier),

I too use the SBCs (v1 + v2), but only to promote resilience of connectivity (have not yet written code on them, but will). Your going the route I was eventually going to take. I know little of MaxDome so can't help you there, but using an SBC to monitor an installation and safeguarded it and not be at the whim of Windoz is a good think, but it's quite costly but lots of fun and flexible. Do keep us posted on your creation. As I stated here and elsewhere, my 1045 is an integral part of my observatory and has been instrumental in protecting the dome and allowing maximum sky usage by ACP. My observatory/scopes have 12 phidget devices (!) and work very well with their C/C++ libraries. I had removed the barrel IR sensor from the 1045 and replaced it with one that did not have any barrel to increase the sky coverage, remove the need for an IR transparent window (not easy to find) and allow for heating it (to clear it of frost, ice and water), but finally it's been no better than the original barrel with a IR semi transparent piece of plastic bag I had used as a window.

BTW, no need for a thermocouple - just get the Phidget analog thermometer and plug it into one of the SBC's analog A/D inputs, far cheaper and plenty good enough. In fact get two, one for outdoor, and one for indoor Temps. For airspeed, there are lots out there, and Phidget does have costly pulse counters, but you can easily turn pulses into an analog value which you can again connect to the SBC I/O. It shouldn't be too difficult to report all of this to ACP, either by mimicking the Boltwood file (as I did), or using other methods that Bob D outlines in his extensive help section of ACP. I dump all my data about every minute to disk in a comma separated file (csv) which can easily be graphed by various tools, such as MS-Excel. I find it very nifty to graph OAT and IAT, as well as sky T, plus the logic signals and slit opening of the previous night. I also have the SQM-LE sky magnitude per arcsec2 overlay-ed with them. The same data is added to a all sky camera I have that takes snap shorts during the night, further helping me evaluate sensor quality, sky quality algorithms and the night's sky quality version the images ACP snapped up (if it did! because images are taking even of less than ideal skies). www.watchobs.com


Cheers

Hi,

These phidgets are interesting, they look good.

For my cloud sensor I just bought a MLX90164, http://www.melexis.com/Assets/IR-sen...heet-5152.aspx ,
and hooked it up to the SDA/SCL comms pins on an Arduino. The IR sensor costs only 15 euros and an arduino 15 to 20 euros.
I´ve then programmed some code in VB to get the values back over serial comms to the PC and display in a graph.

Mine isn´t installed at present as I had problems with dew and rain so it needs a heater element. I have a rain sensor
with heater that should do nicely if I can find the time to install it.

It is great watching the clouds go by via the temperature changes on the graph!

So....What I should order is the 3807_0-Acrylic Enclosure for the 1045 (7.00USD) and the 1045-Phidget Temperature Sensor IR for the computer (90.00USD)....Anything else ??












(
9USD) ???

So...What I should order is: 3807_0-Acrylic Enclosure for the 1045 (7.00) and the 1045-Phidget Temperature Sensor IR (90.00). Anything else ?

Quote:

---

Originally Posted by Frank Mraz

So...What I should order is: 3807_0-Acrylic Enclosure for the 1045 (7.00) and the 1045-Phidget Temperature Sensor IR (90.00). Anything else ?

---

Don't bother with the acrylic housing, it is not rated for exterior use. You'll have to make your own weather proof box, and since the 1045 comes with the barrel sensor, you can't exposed it to the elements either, thus the need for an IR transparent window, which is quite difficult to find. As I stated previously, I finally un-soldered the sensor and replaced it with an identical sans barrel.

My Phidget is still going strong.

An alternative: http://www.auroraeurotech.com/CloudSensor.php

Cheers

Hi Andre....What you saying is that the -3807 - housing is for ONLY indoor use ?

Yes, exactly. Build your own. If you can't remove the sensor barrel, place the 1045 in a weather tight box with a hole, cover the hole with plastic that is semi transparent to the IR (try as many as you can find). I finally found one that reduced the sensor readings by only 40%, as opposed to most that were pretty much 100% opaque. It was a zip lock back if I recall. You can see my original version half way down the page here: www.watchobs.com/astronomy4.html

Excuse me for trying to revive this thread, but a couple of points:
First, try this...put a sheet of Al Foil on the ground on a nice clear day/night and point the IR thermometer at it...Maybe the sensor does not have to point "up" to get a good reading. Not sure what a front-surface mirror would do.
Second, I've been keep track of sky transparency comparing ambient and sky T with ability to see mountains in the distance (up to 60 mi). Seems there's a pretty linear relationship between the two temperatures and transparency in that the parallel lines in T-ambient vs T-sky delineate different degrees of sky clarity (not just cloudy or not). So, if there we a few additional parameters in the Phidget interface, it might be a viable transparency meter. (The SQM is not useful for this!).

Alex Woronow (alex_at_awkml_._com)

Alex, it is never too late... to share information!

Indeed, foil or mirror look down would fix the issue but then you'd have to keep that mirror clean, from at least insects (with proper geometry in a shielded box, rain can be avoided). Wish you had suggested this before I began with the Phidget so many years ago :)

As for AmbT vs SkyT, can you post some data?

Cheers