Transistorized water flow?

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
perfectbite
Posts: 255
Joined: Tue Jan 07, 2003 1:01 am
Location: California
Contact:

Transistorized water flow?

Post by perfectbite »

An easy (dirt simple) and interesting (to me anyway) project:<p>I am building a front garden (out in the sun) drip fed (1/4" copper tubing) birdbath and would like to have a temperature sensitive device in the range of 50 to 90 deg. F. regulate the city water flow 24/7 so that on a cool night the number of drips per minute is at a slow rate and on a really hot afternoon the drip rate increases to keep up with the evaporation rate. As a clamp/regulator I am thinking of using two pieces of a stiff material one fixed, one movable, with surgical tubing sandwiched between them and using two short pieces of muscle wire on each side of the surgical tubing to pull the movable side of the clamp and squeeze the 'sandwich filling' in response to the temperature thereby regulating the open ended water flow. Power to the muscle wires would be controlled by a temperature sensitive transistor. Would I be expecting too much from the muscle wire? Is there a simpler way to do this? Can anyone recommend a 2 wire thermistor transistor in this 50 to 90 deg.F. range that would feed the base of a transistor supplying power to the muscle wire?
User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Re: Transistorized water flow?

Post by Chris Smith »

All transistors [old types] are temperature sensitive. Look them up and find a sharp curved one that fits your temps? <p>For example a 250 MW transistor can handle 100 watts at minus 70. You can use this temp coefficient to bias your circuit. <p>The muscle wire however will work, but lots of amps???<p>Proportional valves, pulse timers, and other ways might use MUCH less power?
User avatar
jwax
Posts: 2234
Joined: Mon Feb 09, 2004 1:01 am
Location: NY
Contact:

Re: Transistorized water flow?

Post by jwax »

Call me lazy, but a small float switch to open a solenoid valve with a flow restrictor seems easier. Could be battery operated, solar powered.
WA2RBA
L. Daniel Rosa
Posts: 400
Joined: Tue Dec 25, 2001 1:01 am
Location: Bellingham, WA
Contact:

Re: Transistorized water flow?

Post by L. Daniel Rosa »

Does it have to be a transistor? The LM334 is temperature sensitive current regulator. The LM335 is a temperature sensitive voltage reference. They're both in a TO-92 package so they look like transistors.
perfectbite
Posts: 255
Joined: Tue Jan 07, 2003 1:01 am
Location: California
Contact:

Re: Transistorized water flow?

Post by perfectbite »

jwax, I am trying to keep the bowl clear so that it easily detaches for easy brushing out so no floats. Just a bowl and a water drip ionto it. <p>L.Daniel, Thank you for your suggestion. That was what I was looking for. I am researching the Nitinol springs. It is one of those websites where there are tons of information and very little informative overview. i.e. Cautions to have good electrical contacts but little on avoiding clamping stress fractures. They have a 2 Amp. stretch/pull and a 3 Amp. compression/push spring about 8mm in dia. I don't know if the Nitinol is all or nothing like a zener diode. <p>Chris, The weather around here, maybe twice a year gets past 90 deg.F. during the day for 3 days each time. I have amended the temp range from 75 to 90 deg.F Below 75, no Amperage, above 90 full Amperage.<p>[ May 07, 2004: Message edited by: perfectbite ]</p>
perfectbite
Posts: 255
Joined: Tue Jan 07, 2003 1:01 am
Location: California
Contact:

Re: Transistorized water flow?

Post by perfectbite »

Apparently Nitinol changes its shape due to what is called 'Ohmic heating' and is not a gradual event. Conceivably Peltier junction blocks could be used to serially activate a 'staged' Nitinol response but that is far too expensive, complex and energy intensive for my use. Thank you for your inputs. A set temperature at which the drip rate suddenly speeds up would seem to be the way to go.
User avatar
haklesup
Posts: 3136
Joined: Thu Aug 01, 2002 1:01 am
Location: San Jose CA
Contact:

Re: Transistorized water flow?

Post by haklesup »

Don't know much about muscle wire but it sounds expensive and hard to get. One waekness is that it requires constant power to regulate (squeeze) to less than max flow. If you loose power, the water goes to max.<p>Do you think a servo motor from an RC toy would have enough torque to turn a screw on a tubing clamp or some sort of cam or valve to pinch off the water flow. Lower power and retains steady state when power is off. This system takes Analog input.<p>Another idea would to put several solenoid valves in parallel, each set with a different flow rate. 3 valves gives 7 flow rates when you add them in different combinations. This system requires digital input which can easily be derived from an analog signal by using multiple comparitors.<p>For the sensor part. just use a regular 2 terminal RTD and associated analog circuit and apply the output to a comparitor (1 or more) so you can set trip points. In fact it may just be easier to use a wall thermostat (battery powered, digital type) then you can select a model with enough programmability to suit your needs.<p>All that aside, if all you want to do is keep a constant level in your birdbath; Fill a large closed container with water and place it at the same level as the birdbath but out of sight. Now put a small tube from your tank to the birdbath. Create a siphon with this tube so that the tank and bath maintain the level you want. Now you can put a more unsightly but cheap and reliable float valve on the tank to maintain its level remotely. You can even hide the tube inside the birdbath base.
techno
Posts: 9
Joined: Sat Apr 24, 2004 1:01 am
Contact:

Re: Transistorized water flow?

Post by techno »

Sensing the temperature isn't going to work. It isn't taking in humidity at that temperature. Imagine a 75* day thats raining? Or one that is close to raining.There isn't any feedback but a whole lot of finagling to get it tuned.
Sense the water level and have that sensor open/close a valve. It doesn't have to be a float you can use a pressure sensor and place it where it reads the activation pressure. At X depth the water is fed, at Y depth it stops.
A light sensor might also work.
A conductor of like a rain sensor would work to, in reverse though. Turns on when dry.<p>I had an electric faucet once. Small motor and batteries. You could hit a button to turn the thing on or hit one for watering, thats what it was for -timed watering. Cheap at the hardware store.
Sure beat turning a valve. Wire that into some sensor.
User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Re: Transistorized water flow?

Post by Chris Smith »

Wire a 555 touch switch/ Humidity switch just under the soil. And use another for pulse soleniod water control.
dyarker
Posts: 1917
Joined: Fri Aug 22, 2003 1:01 am
Location: Izmir, Turkiye; from Rochester, NY
Contact:

Re: Transistorized water flow?

Post by dyarker »

Have you considered a weight or spring to turn water off; and motor, muscle wire, whatever to turn water on? That way water isn't full on during a power outage.<p>A self-heated thermistor is another way to measure water level. Put enough current thru it so it's body gets to, say, 110ºF in air with a breeze blowing. That's water on.<p>Water is a much better heatsink than air, so when water level rises to cover the thermistor, it's body temperature drops to maybe 5ºF above water temperature. That's water off.<p>In between, water just barely touching the thermistor, is medium water flow.<p>This way measures the water level, instead of guessing evaporation rate from water temperature. Wind and humidity effects on evaporation rate will drive you crazy making adustments to drip rate vs water temp.<p>For a specific thermistor search [url=http://www.digikey.com,]www.digikey.com,[/url] [url=http://www.mouser.com,]www.mouser.com,[/url] or your favorite parts place; drill down to manufacturers spec sheets, pick one that is fairly linear between like 45ºF to 120ºF. The spec will also have the thermal resistance of the package at 25ºC in still air, from that you can calculate how much current is needed to heat it to the "in air" temperature.<p>Drive the resistor/thermistor voltage divider with a constant current source instead of a fixed voltage. This increases the voltage change from hot (in air) to cold (in water). A constant current source can be as simple as one transistor, one zener diode and two resistors.<p>Output of voltage divider goes thru a high value resistor to inverting input of an op amp. Pick feedback resistor so op amp out put has a 5V swing from hot to cold. Also connect wiper of a pot to inverting input (ends to pos & neg supply), this sets offset voltage so the output swings from 0V to 5V. Clamp with a couple diodes for insurance. Now you have an analog voltage that you can feed to the ADC of a PIC, or set the duty cycle of a 555, whatever you decide.<p>Cheers,
Dale Y
perfectbite
Posts: 255
Joined: Tue Jan 07, 2003 1:01 am
Location: California
Contact:

Re: Transistorized water flow?

Post by perfectbite »

Here's my plan. A base and lever fulcrumed off the base squeezing the tubing. An opening spring and a damp rag judiciously placed somewhere on the lever beam should provide enough force to restrict the tubing. As the rag dries it will become lighter causing the spring to open the lever clamp. The rag will be re-dampened when I manually turn the sprinklers on every other day - dirt simple. (I'm sorry the Nitinol wire didn't work out. In looking at the power requirements I'd have to have 110 VAC temperature switched to feed the 3 Amps to the Nitinol) <p>I think another name for Nitinol is memory wire and it is used on the Mars' rovers to open test instrument portholes. The portholes probably latch open, I can't see NASA being so profligate with power. Although a voltage isn't specified, there are cautions on overheating the wire. the 'Ohmic heat' produced by 2 or 3 Amps passing through the wire is enough to cause it to shrink or expand. <p>Haklesup. What is an RTD?<p>Chris. A 555 to PWM a water flow controlling solenoid? And a 555 to pick up the ground humidity? You'd have me taking the bird's temps on the wing next to see if they are hydrated enough.<p>My wife was a wonderful gardener and, if she came back and saw the state I have allowed her gardens to get into she would get a stick and she would chase me till she caught me. I think my time would be better spent pulling weeds than building interlocking 555 control modules to keep dem little birds happy.<p>Thanks again.
User avatar
Chris Smith
Posts: 4325
Joined: Tue Dec 04, 2001 1:01 am
Location: Bieber Ca.

Re: Transistorized water flow?

Post by Chris Smith »

The 555 humidity sensor is one of the simplest designs to build for the beginner or expert, while the 555 pulse drip system or switch is another. Doesn’t get any easier,.... when the ground is dry underneath, the water starts and when the saturation reaches the sensor, it goes on which turn off the water.
zotdoc
Posts: 177
Joined: Tue Jun 17, 2003 1:01 am
Location: Douglas, Georgia
Contact:

Re: Transistorized water flow?

Post by zotdoc »

I like electronics projects also, but why not just manually adjust the inflow valve until the birdbath is just barely overflowing and slowly dripping, so that you water the plants around the bath and it generally stays full?
techno
Posts: 9
Joined: Sat Apr 24, 2004 1:01 am
Contact:

Re: Transistorized water flow?

Post by techno »

When your neigbors water their lawns the water pressure will drop. When they don't it goes up. I almost watered my convertable because of this. Sprinkler adjusted when pressure was low. Pressure went up and so did the water.
perfectbite
Posts: 255
Joined: Tue Jan 07, 2003 1:01 am
Location: California
Contact:

Re: Transistorized water flow?

Post by perfectbite »

Zotdoc. That area of the front garden will be 4 inches deep in pea gravel soon. So no plants. (My wife would kill me for wasting growing 'land'). I am using a post mounted terra cotta 12" dia. 1" deep dish with a raised centre drain so the dish always stays full, the overflow will feed a nearby 1/2 barrel trickle fountain. My concern was that on a really hot and sunny day with a good warm breeze that the water drips eventually may not even wet the bottom of the dish (if and when the birds go splashing around in there) and I was looking for ways to compensate for a low water level and electronically speeding up the drip seemed a viable option. <p>I drilled a hole in the terra cotta and pipe nutted a 1/8" pipe nipple into the hole. (You know you're in trouble when, at the hardware store, you ask to speak to an old timer for how-to advice and they point you to a 25 year old intent on selling you the latest carbide-titanium masonry drill bit. The 12 years old 1/2" carbide bit I had worked fine with good work-piece support and lots of water).<p>But wait!<p>This is an electronics forum. I did hear back from the Nitinol store, (BTW Nitinol is an abbreviated acronym for NIckel/TItanium Naval Ordinance Laboratory) and thermal hysteresis gets in the way of Nitinol wire recovery and therefore efficient PWM control of 'Ohmically heated' springs is iffy but, I did think that cycling an integral Peltier block somewhere in the works would speed the cycle up. Is there a cylindrical peltier junction block?<p>Chris. You have piqued my interest. I would like to see a circuit diagram for the 555 water sensor. Did you have a web page source in mind?<p>Dale Y. I still can't figure out what you are doing in Turkey. Teaching perhaps or with an NGO or GO technical mission? Are you in a large city or out in the boonies? What is rural Turkey like anyway? Baked or verdant?<p>Haklesup. Is that old television's AFT fixed yet? I can hear Edd drumming his fingers from here.
Post Reply

Who is online

Users browsing this forum: No registered users and 22 guests