water flow detection, on filling up small waterfall

Capdiamont

I have a small desktop waterfall, that the cats 3-4 love to drink out of. This drains it pretty fast. My idea is to build a stand for it, with a hidden 5 gallon plastic bucket in the base, and a catch basin under the waterfall which drains in to the bucket. I'm thinking of using a sx, to turn on a pump in the bucket to fill up the waterfall once in a while. I intend to have a button for a user to indicate for the pump to turn on now, and some way to detect when the waterfall is overflowing, and thus full. That is my hard part is the overfilling detection. One idea is a trapdoor underneath the drain, weighted so any weight, water will push it down, and trip a sensor/switch. Others is exposed contacts, or weight of the waterfall. The delay between when the pump is turned on, and it is full will also be used to adjust the interval between pumping. IE, short time to fill it, lengthen the time between fills, and longer time to fill it, shorten the time between fills. I would like to store this in some sort of eprom/flash memory in case of power outage. Since I want the sx for cost, I think this might mean external to the sx storage. Might be good to have a temp/humidity sensor and vary time intervals based on that, more storage/look up tables.

LoopyByteloose

A float of nearly any sort connected to a microswitch is the most time honored method to detect water level.

Microswitches come in many sizes, many voltage ratings, and with many different lever arms. Some are designed for all-weather use.

If a BasicStamp is involved, the low voltage saves you a lot of money as the lower voltage microswitches are both smaller and cheaper. Use a length of piano wire with a float on the end to distance the switch from the water. I imagine just a wine cork would be a handy float.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"Everything in the world is purchased by labour; and our passions are the only causes of labor." -- David·Hume (1711-76)········

···················· Tropically,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan

Capdiamont

I imagine that the microswitch, or PING would work great in the 5gal bucket. Problem with the desktop waterfall is there is no room for a float due to design, and looks. I was thinking that under the waterfall, would be a small catch basin, so when the water overflows the waterfall, it is caught, and sent in to the 5 gallon bucket. So there is the rub, detecting the overflow water. Thanks for the ideas though.

Tracy Allen

A sensor based on conductance could be made from SS wire as done in the Applied Sensors tutorial. Put two electrodes in through the side of the overflow tube to detect the presence of water flow. Or maybe even an LED/phototransistor arrangement across the overflow tube could detect flow, either by changes in refraction, or by looking for the effect of turbulent flow on light transmission/refraction.

The QT114 from QPROX is a nice water level detection chip based on capacitance.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

Franklin

Why not just pump the water directly out of the 5 gal bucket to the waterfall and fill the bucket every once in a while?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- Stephen

stephenwagner

http://www.emesystems.com/EarthM/em5.pdf

I think this sight explains how to use a plastic tube, copper conductors and RCTime to measure water height.

1. Connect the tube to the bottom of and run it up the side of the container. Leave the tube open.

2. Run 2 copper wires up the outside of the tube to form a capacitor. As the tube fills with water, the capacitance will change.

3. Use RCtime and a full and empty container to set your limits and develop your code.

You can even set limits for an alarm.

SJW

·

Capdiamont

I wasn't able to read any of their pdf fies. I think this might be the ticket, but in the way I was thinking. What if I ran the wires on the outside of the drain tube, from the catch basin. Measure the difference between the dry/moist tube, and when water is flowing through it?

Shawn Lowe

Tracy-
I was unable to read any of your PDF files as well. Any thoughts?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Shawn Lowe


Maybe I should have waited to do that......

Tracy Allen

Thanks for the head's up. I'll check on the files. In the meantime, the tutorial text is available on the Parallax web site at the bottom of this page...
Applied Sensors Parts and Text, v1.4

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

stephenwagner

I am not sure if this is the exact sight of the capacitive water height sensor.
·
All the Parallax documents that I have reviewed make reference to an emersion probe. Perhaps I missed one.
·
I like the capacitive approach because nothing comes in contact with the water.
·
You can heat shrink the wires and or copper foil in place.
·
SJW


·

stephenwagner

It makes an interesting sensor.

You can integrate the sensor in any fill / flow / drain tube you like.

Check out the closed loop control systems examples posted on the Parallax web sight. Process Control. http://www.parallax.com/detail.asp?product_id=28176

Good luck.

SJW

pwillard

Long Discontinued Part: You can still get National LM1830's on EBAY if you want to try an analog "self contained" sensor approach.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

---

There's nothing like a new idea and a warm soldering iron.

Capdiamont

I have too wonder about one thing, is drain tube size, vs flow, vs detection.

If the drain is a normal bath tub size, will the small flow be detectable?

LoopyByteloose

Capdiamont,
Well you have come full circle. A good robotic microcontroller design can only be as good as the mechanics of the device. You need to balance tubing size, pump size, drain size, and reserve before you can expect the BasicStamp to be able to fine tune things. Flow rate and pressure are directly involved.

One trick to to put oversized tubing on the resupply, the·input and the output, but have valves than can restrictively tweak the flow to an optimal balance.

If you need fluid flow literature, there is material that originates out the the P.E. license program in the U.S. The Intern Engineer Certificate material is quite useful. I happened to have long ago passed·a Califorina Professional Engineer's Certificate as an E.I.T. [noparse][[/noparse]the old name 'Engineer-in-Training' or nowadays, Intern Engineer]; so I see the broader picture of the problem.

But be forwarned, engineers create a number for everything. One has to do the math to get the results.

In sum, the primary system is fluid flow; the control system is a secondary system. You could just as easily use a mechanical secondary system, though using a microprocessor and electronic control is an equally valid choice. But, if the primary system ain't right, nothing's right.

Because it is a 'water fall' I suspect that you really don't want more pressure involved.· Gravity feed whenever possible and larger tubing will keep you from have water 'squirt' unexpectedly.· The drain need not be huge, but balanced within the context of how much the pump can handle.· If you don't have specs, you have to run tests, just like a real engineer, to get useful data.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"Everything in the world is purchased by labour; and our passions are the only causes of labor." -- David·Hume (1711-76)········

···················· Tropically,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan

Post Edited (Kramer) : 9/28/2007 8:24:13 AM GMT

Capdiamont

The pump is going to be a small low voltage type. Control will be on/off. I haven't tested the head of it yet, to see if it will pump it. Then I can play around with squirt. The drain will have to big enough for the flow from the pump in the 5gal bucket, and larger flows from filling the bucket up with new water. My concern is if there will be enough capacitance, with the smaller flow with larger drains, to be detected.

I'm thinking that SS rods, won't dry out enough in this application, and light method may not be tripped if too large of a drain.

stephenwagner

I found the Article for the capacitive sensor.

I hope this helps.

http://www.parallax.com/dl/docs/cols/nv/vol1/col/nv27.pdf

SJW

LoopyByteloose

Personally, I think that measuring capacitance with a liquid might be quite tricky. After all, pure water does not conduct electricity. Anything that does conduct, would likely vary as the amount of ions in the solution change with evaporation, air pollution, or salt in the water. In the past there have been huge threads on measuring liquid levels with various technologies. Some people want to try IR sensors, others sonar, some claim conduction, and then their is capacitance.

There are tiny microswitches that response to a feather-light touch. By all means, go ahead and try to research all the possibliities. But either a mercury switch or microswitch as a float device is likely to be most reliable.

High tech cannot always find a way to replace low tech. Many solutions are brilliantly simple.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"Everything in the world is purchased by labour; and our passions are the only causes of labor." -- David·Hume (1711-76)········

···················· Tropically,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan

Tracy Allen

Most capacitance measurements of depth do not use water as a dielectric. The capacitance is measured across an insulator such as the insulation on a wire or the wall of the container. Another electrode may be either in ohmic contact with the water or it may form another capacitor through insulation, effectively two Cs in series. The water acts as a conductor relative to the insulators, even at pH7. That is for example how the QPROX QT114 level sensor works, using insulation on wire or surface as the dielectric.

There are other kinds of instruments that measure both the real (ohmic) and the imaginary (capacitive) parts of the dielectric constant, to get at conductivity and concentration of ions.

edit: pH7, not p7

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

Post Edited (Tracy Allen) : 9/28/2007 8:32:19 PM GMT

stephenwagner

I think Dr. Allen has hit the nail on the head with this one. He is an authority on the subject.
·
I recommended capacitive sensor for this application because nothing comes in contact with the water.
·
I have been discussing how to measure/control the height/flow rate of steam distilled water in a decorative consumer fountain that the cat(s) drink from using a home made sensor and uncelebrated units of PULSIN/RCTIME/POT.

The control system as a hole can be calibrated againest a full and·empty fountain with limits set by the user.



SJW

stephenwagner

cat(s) drink from using a home made sensor and uncelebrated uncalibrated units of PULSIN/RCTIME/POT.

pwillard

Well, if it had Champagne in the cat bowl, it might be uncelebrated since you didn't get to drink any.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

---

There's nothing like a new idea and a warm soldering iron.

LoopyByteloose

I suspect the cats will trigger a lot of false data. Either you will get too much water or not enough. But I am willing to be proven wrong.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"Everything in the world is purchased by labour; and our passions are the only causes of labor." -- David·Hume (1711-76)········

···················· Tropically,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan

Shawn Lowe

Yeah, that is the capacitance water level I was talking about. I bet that would work like a charm!

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Shawn Lowe


Maybe I should have waited to do that......

Capdiamont

Well I guess it now time to experiment. I have a unused bs2, and relay already attached to control a 110v outlet.

skylight

is it water depth or as i read it water flow from the drain that needs sensing? if flow then·a simple flow switch would suffice wouldnt it?

Capdiamont

water flow from drain. A simple flow switch would be nice, but not sure how.


looking over the design, I'm wondering why the N&V didn't measure the capacitance directly with rc time. Why not use a 555 chip? Basically I don't have a 4060 chip.