# What sensor for a homemade greenhouse rain guage?

Status
Not open for further replies.

#### bigcanuknaz

##### New Member
Hi all:

This is my first time at this forum. I spent a bit of time reading, and understood some of it, but most is over my head. I think this will be a fantastic place for me to get advice, on designing my greenhouse automation projects.

My problem:

We just rented a new greenhouse that has gutter opening roof vents, instead of peak opening vents. When it rains, you typically set the vents to open only a few inches, and less on the windside than on the leeside. This works fine with peak vents. But for gutter vents (vents that pivot on the peak of the greenhouse and open from the gutter with another greenhouse connected to the other side of the gutter), if it rains very heavy, then the gutter can overflow into the greenhouse making a mess. In this case, you want to close the roof (gutter) vent, open side vents, and turn on exhaust fans.

I am looking for a rain sensor, that measures how hard it is raining, with a reasonably fast response time on the order of 20 sec to a minute.

My search:

I looked and inquired from our equipment suppliers to no avail. Typical "tipping bucket sensors" measure on the order of 1/8 of an inch of rain, and are used primarily for measuring the amount of rain that falls over the period of about a day. They are accurate, but have a slow response time. The main one i looked at also had a very short pulse, that would require a frequency rather than just a dry contact input into our controller.

I found this homemade sensor in my search:

Electronic Rain Sensor

It looks easy to build, and will be quite easy to adjust (either the funnel size, and/or the size of the dual bucket). The only question I have is what kind of sensor should I use?

My control system (Argus Control Systems Inc - Titan controller) will take as inputs:

1. 0-5 vdc
2. 4-20ma (same input using a resistor)
3. resistance measurement (dry contact) (same input)
4. frequency input (different input)

My first thought is to just hook up resistance measurement from the aluminum tipping bucket to a contact point on one side of the bottom of the PVC tee (where the inductive sensor is in the homemade sensor referenced above). I could easily adjust the resistance level, to get an on/off input depending on whether the bucket is in contact or not with the contact point.

The control program would be easy for me to set up.

My question:

Is using just a "dry contact" (ha ha ha, in the rain)the best way to go? Have you any other comments?

Thanks so much. I look forward to posting other questions in the future.

naz

Last edited:

#### bigcanuknaz

##### New Member
You can get low cost 'tilt' switches, the contacts are enclosed, all you would have to do is seal the take off leads.
I like the idea of keeping the wires and possible sensor off of the tipping bucket. Less possibility of wear and tear, and of non repeatable performance.

I think I am going to try it out with just using resistance, with the tipping bucket itself closing and opening the circuit.

I'll use a long bolt (instead of the copper wire in the link above) as the fulcrum. I'll then use an extra couple of nuts on one end to secure one lead of the wire.

And in place of the inductive sensor in the linked page, I'll use another bolt and nut.

Bolts and nuts will be stainless, to keep working in the rain, and the tipping bucket itself will be aluminum.

All together maybe $5 or$10 in parts and some fun making it.

Any other ideas or insights will be appreciated.

naz

#### Boncuk

##### New Member
There are flow meters on the market used in conjunction with trip computers for cars. They contain a wheel which interrupts an IR-beam and their output is almost a square wave.

All they require is any liquid for a minimum of lubrication of the wheel bearing (needle type shaft, therefor not recommended to measure air flow.)

I guess they would just be perfect for fast response.

Boncuk

#### Rolf

##### Member
Bugs !!!!!!!!!!!!!!

{snip}
My question:

Is using just a "dry contact" (ha ha ha, in the rain)the best way to go? Have you any other comments?

Thanks so much. I look forward to posting other questions in the future.

naz
I have a commercial unit that uses the tilt bucket system and a reed switch / magnet to count the tilts. Works great indoors!
The problem is the bugs build nests in it and the mechanism gets gummed up every few months and has to be cleaned out.
Have not figured out what to do about it. :-(

After taking it apart for the second time in a month, another bug, a spider who's cocoon had gummed the tilting mechanism to a holt. Then after cleaning and a closer examination of the bottom mounting plate I found four (4) relative large holes that had no part of the drain system. just piss poor engineering of the casting mold. I puttied them closed with somme HandyiTAK.
Now; maybe it will stay in operation for a few months, I will try to remember to report back.

Last edited:

#### dknguyen

##### Well-Known Member
I have a commercial unit that uses the tilt bucket system and a reed switch / magnet to count the tilts. Works great indoors!
The problem is the bugs build nests in it and the mechanism gets gummed up every few months and has to be cleaned out.
Have not figured out what to do about it. :-(
CHlorine tablets? Some substance that does not evaporate with the water and remains behind as water is added and evaporates...like a ton of salt?

EDIT: Oh wait, this is running water.

Last edited:

#### Mikebits

##### Well-Known Member
How about using a humidity sensor in the green house? You can control the set point.

#### dougy83

##### Well-Known Member
Do you even need a tipping sensor? What about just having the funnel feed into a leaky bucket; the bucket has probes somewhere towards the top - if the rain flow rate is sufficiently greater than the leak rate, the bucket will fill the the probes will conduct current. For fast response, use a smaller bucket (or pipe, jug, bottle, etc.).

This idea's not original: There's a Hole in My Bucket - Wikipedia, the free encyclopedia

#### bigcanuknaz

##### New Member
How about using a humidity sensor in the green house? You can control the set point.
yes, we already measure humidity and VPD. we also have a heated ring conductivity rain sensor, and light (w/m2) sensor. So we know when it is raining, just not how hard it is raining.

Do you even need a tipping sensor? What about just having the funnel feed into a leaky bucket; the bucket has probes somewhere towards the top - if the rain flow rate is sufficiently greater than the leak rate, the bucket will fill the the probes will conduct current. For fast response, use a smaller bucket (or pipe, jug, bottle, etc.).

This idea's not original: There's a Hole in My Bucket - Wikipedia, the free encyclopedia
Yes i thought of that. But I was concerned about problems calibrating, and inconsistency over time as the hole(s) got gummed up with dirt and living things. The continuation of this idea is actually directly measuring the water level in the gutter of the greenhouse. A float switch might be possible. Just a continuity check would be possible, although complicated because of the low conductivity of rain water, and the fact that the gutter is aluminum.

I have the above home built tipping bucket built. Took me about 3 hours, and less than \$10 in parts. I'll install this weekend. I do *not* hope for rain, as the tomatoes would prefer sun. I'll let you know how it works.

naz

#### bigcanuknaz

##### New Member
btw, it should have taken me 1.5 instead of 3 hrs. (incl. 45 min to pick up the parts)

I epoxied the tipping bucket together from pieces of aluminum sheeting, and did not get continuity as the bucket made contact....

duh, epoxy does not conduct. a few drops of solder solved the problem, but set me back an hour at least, having to take it apart, fix it, and reassemble.

naz

double post

Last edited:

#### bigcanuknaz

##### New Member
Hi all:

It works!

It took a bit of tuning.

because of water, the contact resistance goes anywhere from close to zero up to about 4000 (ohms??) when closed, but when open it is well above 5000. I had to set the "flip/flop" point to 4600. Also, it takes a while for the control system to sense the switch (the conductivity reading moves somewhat slowly), and to work through the logic.

As it is set now, I am looking at anywhere from a 15 sec to 40 sec response time, depending on how hard it starts raining, and exactly when in the cycle it starts raining hard.

As the greenhouse gutters have some capacity to fill up before they overflow, this seems to be a reasonably good match. If we need faster response, I can always go to a proximity switch, or something like that. But it will only gain me 2 or 3 seconds (the time that it takes the conductivity to fall or rise). a lot of the lag is in the control system (as much as 8 or 10 seconds!)

I got it done with just one extra "equations" program. My control logic is:

Heavy Rain trigger (new 4 line equations program):

(If _contact_, and
_on_time_ long), or
(If _no_contact_, and
_off_time_long),
then _high_

Heavy Rain Pickup (an extra line in an existing program):

Not raining unless _low_ for too long a time.

The logic is that the rain gauge will flip flop back and forth slowly whenever it rains. Checking for both contact and no contact, indicates the change in state, either when making or breaking contact. In the light rain case the trigger will go low for (set to 15 sec now) and then back high again, when the timer counts past 15 sec. However, in the heavy rain case, if it flips faster than 15 sec, then it will stay low.

The pickup looks for low, but delays for 15 seconds plus another 15 second offset to allow for control logic lag. (It looks like I may be able to get the offset down to 11 sec, but I'd rather have a bit more delay, than risk too many false positives)

Thanks,

naz

edit: "low rain case" ==> "light rain case" for readability

Last edited:

##### New Member
I built a rain gauge for part of a project in college. I had used a bunch of capacitors in series "stacked" along a thin piece of PCB. It was connected to a 555 timer circuit, so when the capacitance changed (with the level of the water shorting out specific capacitors) the frequency changed.

#### Rolf

##### Member
I have a commercial unit that uses the tilt bucket system and a reed switch / magnet to count the tilts. Works great indoors!
The problem is the bugs build nests in it and the mechanism gets gummed up every few months and has to be cleaned out.
Have not figured out what to do about it. :-(

After taking it apart for the second time in a month, another bug, a spider who's cocoon had gummed the tilting mechanism to a holt. Then after cleaning and a closer examination of the bottom mounting plate I found four (4) relative large holes that had no part of the drain system. just piss poor engineering of the casting mold. I puttied them closed with somme HandyiTAK.
Now; maybe it will stay in operation for a few months, I will try to remember to report back.
It has now been working flawlessly for over two months and we had quite a bit of rain.

Status
Not open for further replies.