Forums

"Input" on an "output"

Started by Don Y September 15, 2013
On 9/18/2013 12:00 PM, bloggs.fredbloggs.fred@gmail.com wrote:
>> E.g., to fully *configure* the system, you need to specify the >> individual plants (actual species) served in each zone, the types >> of soil in which they are planted, sun & wind exposure patterns, >> runoff/drainage patterns, flow rates of emitters associated with >> the individual plantings, etc. I.e., the same sort of information >> that you would provide to a horticulturalist tasked with setting >> a suitable irrigation schedule (*not* Joe HischSchoolDiploma to >> whom the local Home Improvement store subcontracts irrigation system >> installations). >> >> [Doing all these things doesn't guarantee you a turnkey, optimal
----^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> solution. But, it brings you much closer and makes fine tweeks >> easier -- embed the "expert" in the system instead of relying on >> the "homeowner" to have the desired level of expertise. (he'll >> just keep increasing watering times to ensure nothing dies from >> dehydration -- and, probably never back off on those times as >> the seasonally adjusted needs of his landscape change)]
>> But, you *knew* all this already, right? Perhaps you could point me >> towards some commercially available product offering of your own??
> I know a tad bit about agricultural irrigation and I'll tellya'- that's > the most ridiculous writeup extant.
"Ridiculous"? Because plants have other sources of moisture besides environmental (the ultrasecret "Water Illuminati" that teleports water directly to plants -- bypassing environmental sensors -- directly from Alpha Centauri for the sole purpose of confounding horticulturalists in their secret plot to take over the world!)? Because plants use water based on factors other than solar radiation, temperature, wind, etc. (the colors of the vehicles driving by on the roadside!).
> There are more than a few unknowns making your idea /completely/ > unworkable,
So, you've been keeping current on product offerings? Atmometers? Lysimeters? Current research? Which Ag department are you affiliated with? Co-op extensions?
> but, since you /know/ so much, I'll let you find out what they are.
I have *no* intention of "finding out what they are"! Our yard is very nicely tuned, thankyouverymuch. I approached our problem by controlling the exposure, soil characteristics, (e.g., microclimates) for each planting. And, made sure to keep plants with DISsimilar needs on different irrigation zones so their conflicting goals wouldn't lead to "unsolvable" irrigation problems. Roses need to be watered twice a day during the hot/dry/windy months? Citrus needs long, deep soaks thrice a week during the growing season? (c'mon... "15 zones" in a small, "city" lot? Didn't it occur to you that this was a bit "fine grained" for an irrigation system? Esp given your "tad bit about agricultural irrigation"!) With everything else I'm involved in, technologically, do you think I'm going to spend time *solving* this problem? Perhaps *after* I finish perfecting weather forecasting... (which I'll do just after I find a cure for cancer) I have no intention of wasting my time building trinkets. Or, running a commercial enterprise and dealing with "customers". Can you spell RETIREMENT? :> Rather, I intend to present a framework that lets vendors, researchers, "tinkerers", etc. experiment with control algorithms to determine their efficacy. Hopefully, they will *share* these discoveries in the same spirit that I will be sharing the code base, hardware designs, documentation, etc. I.e., to *improve* upon existing practices instead of just "more of the same". Commercially available irrigation systems tend to be *islands* of automation. They have their own sensors, controls, etc. This forces anyone designing/developing/deploying such a system to take on The Whole Enchilada. And, forces (software/algorithm) development to happen in an "embedded system" environment -- which already limits the number of folks who can effectively "tinker", there. OTOH, if the irrigation *interface* is accessible to other systems (e.g., via an automation network) ALONG WITH other resources (on-site, real-time meteorological data; NOAA weather forecasts; database of historical observations for *this* particular property; database of species specific irrigation needs; live sensor data from atmometers/lysimeters/etc.; and "whatever"), then a "desktop programmer" can comfortably and confidently experiment with different control algorithms and heuristics. And *see* the sorts of performance that can be obtained by drawing on each of these resources in control. The apparent difference between your attitude (expressed in the above statement) and mine, is that I am actually trying to make "the world a better place" (cliche) while you appear intent on hording your *claimed* (unproven) knowledge/experience -- so things "remain the same". I'd rather offer folks a well-established "starting point" from which they can elaborate and extend a system (so they don't have to "build from scratch") No skin off my back. *I* don't have any kids, grandkids, etc. to worry about! :> If *yours* end up drinking treated effluent or dealing with water rationing, prohibitions on certain types of landscape plants, higher food prices, etc. I won't lose much sleep over it (after I'm gone). You only have to answer to yourself. I'm comfortable with *my* efforts! "Grandpa, why can't we flush the toilet after we pee?"
On Wednesday, September 18, 2013 4:32:34 PM UTC-4, Don Y wrote:

>=20 > "Grandpa, why can't we flush the toilet after we pee?"
I'll tell you why, because you flushed it down the drain on a bunch of oddb= all landscaping that required too much water. Modern landscape architecture= is getting away from that. People are learning that they have plenty of op= tions using plants native to their climate zone and soil conditions. This m= eans watering is only required during extreme drought conditions, *if that*= . It also means the plants are immune to attack from the local pests and di= seases (no pesticides, fungicides, microbicides, bactericides), and they su= pply food to the local wildlife in one form or another. You will find that = the native plants will do spectacularly well and with very little maintenan= ce. There are exceptions with some of the chinensis and japonica imports th= at have proved hardier than the natives, but you should check with your sta= te agricultural department for things like invasiveness and legality first.= Things like water intensive roses are out, and it sounds like the citrus a= re too if they absolutely have to have three deep waterings weekly. Sounds = like you have a bunch stuff that needs to pulled up and thrown on a compost= heap.
On 9/18/2013 6:51 PM, bloggs.fredbloggs.fred@gmail.com wrote:
> On Wednesday, September 18, 2013 4:32:34 PM UTC-4, Don Y wrote: > >> >> "Grandpa, why can't we flush the toilet after we pee?" > > I'll tell you why, because you flushed it down the drain on a bunch > of oddball landscaping that required too much water.
"Lawns", golf courses, people using a garden hose to "wash" the cruft off their driveways instead of dragging out a *broom*, businesses watering lush lawns during daylight hours, hydraulic mining, fracking, leaving the water running while brushing teeth, legacy plumbing fixtures, swimming pools, spas, crops that use disproportionately high amounts of water, etc.
> Modern landscape architecture is getting away from that. People are > learning that they have plenty of options using plants native to their > climate zone and soil conditions. This means watering is only required > during extreme drought conditions, *if that*.. It also means the plants > are immune to attack from the local pests and diseases (no pesticides, > fungicides, microbicides, bactericides), and they supply food to the > local wildlife in one form or another. You will find that the native > plants will do spectacularly well and with very little maintenance.
Our yard is xeriscaped with the exception of the citrus and roses. I (personally) removed/dug up the "lawn" that was here previously. We use no chemicals to control weeds, pests, etc. (because our domestic water comes from wells and we don't want to add more cruft to that water supply!) But, you'll find truly "native" landscapes (here) are very boring, lack color, etc. And, you'll find things like blood oranges very expensive (and having to be *shipped* here) instead of growing them yourself (ignore the difference in the *quality*, lack of pesticides, freshness, etc.)
> There are exceptions with some of the chinensis and japonica imports > that have proved hardier than the natives, but you should check with > your state agricultural department for things like invasiveness and > legality first. Things like water intensive roses are out, and it > sounds like the citrus are too if they absolutely have to have three > deep waterings weekly.
Oranges grown in California/Florida also require water. The difference is the amount available to them as a consequence of local environmental conditions. And, "subsidies" from other water sources. If each locality was required to survive with their own precipitation, lots of crops would just disappear. Or become vastly more expensive. E.g., trade diesel fuel (shipping products) for water.
> Sounds like you have a bunch stuff that needs to pulled up and thrown > on a compost heap.
We offset our "excessive" water needs with rainwater harvesting and general water conservation. Not letting water *leave* the property easily. E.g., using *all* of the precipitation that falls on the property instead of relying on "municipal water" to attend to those needs. Amending the soil to retain moisture instead of letting it seep through directly to the aquifer. We *surely* aren't wasteful enough to fill a swimming pool/spa and watch hundreds of gallons evaporate *daily*. Or, have a *lawn* watered with "sprinklers". Or a "mister" that cools *outdoor* air (OUTSIDE!) around a patio by blindly pushing water into the air. Koi pond, "water feature", fountain, etc.
On Wednesday, September 18, 2013 10:41:14 PM UTC-4, Don Y wrote:

> > But, you'll find truly "native" landscapes (here) are very boring, > > lack color, etc. And, you'll find things like blood oranges very > > expensive (and having to be *shipped* here) instead of growing them > > yourself (ignore the difference in the *quality*, lack of pesticides, > > freshness, etc.)
Where are you?
> > Oranges grown in California/Florida also require water. The difference > > is the amount available to them as a consequence of local environmental > > conditions. And, "subsidies" from other water sources. If each > > locality was required to survive with their own precipitation, lots > > of crops would just disappear. Or become vastly more expensive. > > E.g., trade diesel fuel (shipping products) for water.
The imported citrus, mostly coming from underdeveloped countries, actually costs less than the domestic produce. Practices like the jackasses in California growing rice and cotton should be abolished, but the idiots are waiting for an ecological disaster first.
> > > > > Sounds like you have a bunch stuff that needs to pulled up and thrown > > > on a compost heap. > > > > We offset our "excessive" water needs with rainwater harvesting and > > general water conservation. Not letting water *leave* the property > > easily. E.g., using *all* of the precipitation that falls on the > > property instead of relying on "municipal water" to attend to those > > needs. Amending the soil to retain moisture instead of letting it > > seep through directly to the aquifer.
You must have a small lot.
> > > > We *surely* aren't wasteful enough to fill a swimming pool/spa and > > watch hundreds of gallons evaporate *daily*. Or, have a *lawn* watered > > with "sprinklers". Or a "mister" that cools *outdoor* air (OUTSIDE!) > > around a patio by blindly pushing water into the air. Koi pond, "water > > feature", fountain, etc.
Yeah, no end to the ways people waste resource. My area is investing in a $250M augmented river flow reservoir to meet projected demands of the future, but that water was just going into the Atlantic anyway.
On 9/18/2013 8:05 PM, bloggs.fredbloggs.fred@gmail.com wrote:
> On Wednesday, September 18, 2013 10:41:14 PM UTC-4, Don Y wrote: > >> But, you'll find truly "native" landscapes (here) are very boring, >> lack color, etc. And, you'll find things like blood oranges very >> expensive (and having to be *shipped* here) instead of growing them >> yourself (ignore the difference in the *quality*, lack of pesticides, >> freshness, etc.) > > Where are you?
So AZ. 10-11" precip annually. Think: "desert". Sage brush, creosote, cacti, etc.
>> Oranges grown in California/Florida also require water. The difference >> is the amount available to them as a consequence of local environmental >> conditions. And, "subsidies" from other water sources. If each >> locality was required to survive with their own precipitation, lots >> of crops would just disappear. Or become vastly more expensive. >> E.g., trade diesel fuel (shipping products) for water. > > The imported citrus, mostly coming from underdeveloped countries, > actually costs less than the domestic produce.
... picked a week ago -- so, either ripened in the hold of a ship *or*, overripe, by now. Pick an orange off the tree and eat it 10 minutes later. Compare to *anything* store bought and you will see why we spend the time and effort we do to grow them. Along with the *size* of the fruit (our limes are as large as store bought oranges; oranges as large as grapefruit, etc.)
> Practices like the jackasses in California growing rice and cotton > should be abolished, but the idiots are waiting for an ecological > disaster first.
I am told cotton is grown "just up the road". (Did I mention 10" annual precipitation??) Along with a fair bit of other "farming".
>> We offset our "excessive" water needs with rainwater harvesting and >> general water conservation. Not letting water *leave* the property >> easily. E.g., using *all* of the precipitation that falls on the >> property instead of relying on "municipal water" to attend to those >> needs. Amending the soil to retain moisture instead of letting it >> seep through directly to the aquifer. > > You must have a small lot.
City lot. 1/3 - 1/2 acre?
>> We *surely* aren't wasteful enough to fill a swimming pool/spa and >> watch hundreds of gallons evaporate *daily*. Or, have a *lawn* watered >> with "sprinklers". Or a "mister" that cools *outdoor* air (OUTSIDE!) >> around a patio by blindly pushing water into the air. Koi pond, "water >> feature", fountain, etc. > > Yeah, no end to the ways people waste resource.
Because policies don't encourage (reward/punish) better stewardship. E.g., we are now approaching the "treated effluent" stage of demand for potable water. One would think "conservation" would be a big deal! *Stressed* by "policy", pricing, etc. If not for "moral" grounds, then for the economic reason of AVOIDING the need for a new treatment plant, etc. Ah, but that means new development (mindless way to stimulate the local economy by borrowing/stealing? from the future) would be hard to support by policy. So, can't come out and say that! And, if you *do* conserve, city complains they need to raise rates because not enough water is being used (to offset their existing cost structure!). Install gray water recycling and they complain not enough water is flowing through the sewers to keep them clear. Instead, build a water treatment plant. Tax folks to pay for it. Then worry when you will EVENTUALLY run out of water and the whole mechanism grinds to a halt. (I think municipalities here have to certify a 100 year water supply. Failing to do so puts bonds, etc. in jeopardy)
> My area is investing in a $250M augmented river flow reservoir to meet > projected demands of the future, but that water was just going into the > Atlantic anyway.
In the recent past, water from the Central Arizona Project (CAP) has been introduced to potable water supply (historically from wells). (CAP water flows through a manmade canal from "way up north") But, past attempts to do this met with some problems (?). To address these concerns, the water is pumped *into* the aquifer and then drawn *out* of the aquifer (as well water), locally. Apparently, the "in" and "out" are within eyeshot of each other. Until you have reasonable (not heavy-handed) policies in place, folks have no incentive to behave "responsibly". I can be fined if the "water police" happen to catch irrigation water running onto the street *or* across the sidewalk on my property. Yet a business can use spray irrigation to keep a green lawn lush during the heat of the day? 24/7/365?? Neighbor hand washes their *four* vehicles every two weeks. At least if he went to a "car wash", the water would be recycled... I think people (locals) lose track of just how little precip falls here, each year. "You get used to it". Coming from a much lusher environment ("*God* waters the yard"), the difference is far more obvious. You are constantly aware of how much *less* you have than you had previously. E.g., none of the "locals" that I know would even *consider* rainwater harvesting -- regardless of the number of citrus trees on their property, swimming pool, lawn, etc. Yet, come Monsoon season, everyone drops what they are doing and walks outside to watch it rain. ("Hello! Does anyone see the inconsistency, here?")
On 9/18/2013 5:30 AM, Don Y wrote:
> Hi Ed, > > On 9/17/2013 10:29 PM, ehsjr wrote: >> On 9/17/2013 2:29 AM, Don Y wrote: > >>>> For that, you'll need a low voltage >>>> power supply in addition to your existing power supply to the >>>> solenoid. (You could derive the low voltage from the existing >>>> supply if you want.) The low voltage supply will allow you to >>>> sense the status of the switch without forcing enough current >>>> through the solenoid to energize it or to cause a lot of heat >>>> in it. >>>> >>>> N/O >>>> -------- +-----o o-------+ >>>> | New | | | >>>> | Power |---+---[Solenoid]---+ >>>> | Supply | | >>>> | Low V |-----+---[R]---+----+ >>>> -------- | | >>>> |<---V--->| >>>> >>>> R is high wattage low resistance, chosen to allow reliable >>>> operation of the solenoid with your existing supply. V >>> >>> You are sensing V *behind* the solenoid's resistance -- O(120 ohms). >> >> I have no idea what you mean by "behind" the solenoid's >> resistance, nor why it makes a difference in your mind. >> >> With a given supply voltage, the voltage across the resistor >> will be higher when the switch is pressed than when it is >> not. Detect that higher voltage with your circuit to >> initiate switching state from on to off or from off to on. >> >> The resistor is physically located at the voltage source >> location, wherever that is. > > You haven't thought this through, completely.
Correct. I limited my thinking to a way you can detect a switch state where the switch is installed near the solenoid location and connected using the 2 existing wires that go to the solenoid, _because that is what you asked about_ . The rest of the functionality you want is up to *your* circuit. Ed <snip>
On 9/19/2013 1:59 AM, ehsjr wrote:
> On 9/18/2013 5:30 AM, Don Y wrote: >> Hi Ed, >> >> On 9/17/2013 10:29 PM, ehsjr wrote: >>> On 9/17/2013 2:29 AM, Don Y wrote: >> >>>>> For that, you'll need a low voltage >>>>> power supply in addition to your existing power supply to the >>>>> solenoid. (You could derive the low voltage from the existing >>>>> supply if you want.) The low voltage supply will allow you to >>>>> sense the status of the switch without forcing enough current >>>>> through the solenoid to energize it or to cause a lot of heat >>>>> in it. >>>>> >>>>> N/O >>>>> -------- +-----o o-------+ >>>>> | New | | | >>>>> | Power |---+---[Solenoid]---+ >>>>> | Supply | | >>>>> | Low V |-----+---[R]---+----+ >>>>> -------- | | >>>>> |<---V--->| >>>>> >>>>> R is high wattage low resistance, chosen to allow reliable >>>>> operation of the solenoid with your existing supply. V >>>> >>>> You are sensing V *behind* the solenoid's resistance -- O(120 ohms). >>> >>> I have no idea what you mean by "behind" the solenoid's >>> resistance, nor why it makes a difference in your mind. >>> >>> With a given supply voltage, the voltage across the resistor >>> will be higher when the switch is pressed than when it is >>> not. Detect that higher voltage with your circuit to >>> initiate switching state from on to off or from off to on. >>> >>> The resistor is physically located at the voltage source >>> location, wherever that is. >> >> You haven't thought this through, completely. > > Correct. I limited my thinking to a way you can detect a switch > state where the switch is installed near the solenoid location > and connected using the 2 existing wires that go to the solenoid, > _because that is what you asked about_ . > > The rest of the functionality you want is up to *your* circuit.
You make a good point with that. The issue with the failure is not really such a big deal. His controller will be able to sense if the button is shorted and notify the operator. The button is very unlikely to fail shorted, but there are plenty of other failure mechanisms. If the system needs to be robust then these other failure mechanisms should be detected. Probably the most important thing would be to just plain detect if water is coming from the spigot or not rather than detect all the individual failures. Yes, a pressure switch *after* the valve could do that and the pressure switch could be connected to the power wires in a similar manner to provide a result back to the controller. The push button shorts the solenoid, the pressure switch has a resistor in series. So now you have three levels of voltage you need to sense to detect the button and the pressure switch. I would do this with a current driver circuit. It would have a constant current drive allowing the voltage to float depending on the resistance. Sense the voltage out of the driver and you are done. 24 volts is solenoid operating normally, 16 volts is solenoid operating and pressure detected in the output line, 0 volts is button pressed or solenoid shorted. You could use a low level of current to sense the switch when the water if off, or you could pulse the full current briefly, too short to activate the solenoid. Don is a tough cookie to discuss things with. You offer some help and he pulls more requirements out of thin air. But for the most part he is brainstorming and that can be part of what we are seeing. He likely didn't think about the failure detection as an initial requirement, it occurred to him after he started thinking about how to implement what he wanted. So it became a requirement because he thinks it will be easy to include whether it is very valuable or not. I like your idea. But I'm not sure that it will do what he wants if you use a common resistor as a sensor. That would require that you periodically turn off all solenoids and poll each switch. In theory you could do this fast enough that active solenoids do not drop out, but you would then be sending some moderately fast pulses along long lines potentially generating a lot of RFI. I bet it would mess up AM radio pretty badly. Maybe there would be a way to shape the edges to prevent this. I guess sampling 4 times a second would be adequate, maybe even just once a second. That wouldn't be so bad, so maybe my concerns are not valid. -- Rick
On Thursday, September 19, 2013 12:34:38 AM UTC-4, Don Y wrote:

>=20 > So AZ. 10-11" precip annually. Think: "desert". Sage brush, creosote, >=20 > cacti, etc.
There are a bunch of resources on ultra-low water requirement landscape mat= erial there. I found lots of stuff from your state agriculure dept as well = as something called Arizona Municipal Water Users Association. There are to= ns of ornamental desert plants that do perfectly well with little to none s= upplemental watering.
> >=20 > Pick an orange off the tree and eat it 10 minutes later. Compare >=20 > to *anything* store bought and you will see why we spend the time >=20 > and effort we do to grow them. Along with the *size* of the fruit >=20 > (our limes are as large as store bought oranges; oranges as large >=20 > as grapefruit, etc.)
Those plants require tons of water, doesn't make any sense to grow them in = a desert climate. But I suppose you could get away with a small number if y= ou do things like select grafted root stock dwarf variety, possibly use in-= ground grow bags to mechanically condense root mass, use subterranean drip = irrigation, surface mulch/anti-evaporation measures, and possibly shading n= ets for the extra hot months.
>=20 > City lot. 1/3 - 1/2 acre?
That's so tiny there's no excuse for not just manually turning your water o= n and off.
>=20 > Instead, build a water treatment plant. Tax folks to pay for it. >=20 > Then worry when you will EVENTUALLY run out of water and the >=20 > whole mechanism grinds to a halt. (I think municipalities here >=20 > have to certify a 100 year water supply. Failing to do so >=20 > puts bonds, etc. in jeopardy)
It's all about job preservation with those people.
>=20 > In the recent past, water from the Central Arizona Project (CAP) has >=20 > been introduced to potable water supply (historically from wells). >=20 > (CAP water flows through a manmade canal from "way up north") >=20 >=20 >=20 > But, past attempts to do this met with some problems (?). To >=20 > address these concerns, the water is pumped *into* the aquifer >=20 > and then drawn *out* of the aquifer (as well water), locally. >=20 > Apparently, the "in" and "out" are within eyeshot of each other. >=20 >=20 >=20 > Until you have reasonable (not heavy-handed) policies in place, >=20 > folks have no incentive to behave "responsibly". I can be fined >=20 > if the "water police" happen to catch irrigation water running onto >=20 > the street *or* across the sidewalk on my property. Yet a business >=20 > can use spray irrigation to keep a green lawn lush during the heat >=20 > of the day? 24/7/365?? Neighbor hand washes their *four* vehicles >=20 > every two weeks. At least if he went to a "car wash", the water >=20 > would be recycled... >=20 >=20 >=20 > I think people (locals) lose track of just how little precip falls >=20 > here, each year. "You get used to it". Coming from a much lusher >=20 > environment ("*God* waters the yard"), the difference is far more >=20 > obvious. You are constantly aware of how much *less* you have >=20 > than you had previously. E.g., none of the "locals" that I know >=20 > would even *consider* rainwater harvesting -- regardless of the >=20 > number of citrus trees on their property, swimming pool, lawn, etc. >=20 >=20 >=20 > Yet, come Monsoon season, everyone drops what they are doing and >=20 > walks outside to watch it rain. ("Hello! Does anyone see the >=20 > inconsistency, here?")
That 10" is ridiculous, the place is not exactly a garden of Eden.=20
On 9/19/2013 9:37 AM, bloggs.fredbloggs.fred@gmail.com wrote:
> On Thursday, September 19, 2013 12:34:38 AM UTC-4, Don Y wrote: > >> So AZ. 10-11" precip annually. Think: "desert". Sage brush, creosote, >> cacti, etc. > > There are a bunch of resources on ultra-low water requirement landscape > material there. I found lots of stuff from your state agriculure dept > as well as something called Arizona Municipal Water Users Association. > There are tons of ornamental desert plants that do perfectly well with > little to none supplemental watering.
Yes, we spent the better part of a year researching what to plant. And, tried many different specimens that, ultimately, we opted against using (e.g., "African Sumac" is a popular tree, low water use, etc. But, you end up with a *yard* full of seedlings with tenacious roots.). At the time, we had a pair of dogs so that put a constraint on what we could plant -- many plants are toxic, have spines, etc. And, with such small/close lots, you want to choose specimens to enhance privacy (so you aren't looking at a "wall" each time you glance out the window). E.g., mesquite trees (signature plant for this region) are reasonably low water use, provide filtered shade, etc. But, produce bajillions of 1/4W resistor sized "leaflets" that track into the house, etc. An established pine tree with deep roots out front -- but, didn't yield any usable shade, pine needles everywhere (they trap water when they accumulate on the roof and "rot" the roofing material), AND have a tendency of snapping off at the base in the microbursts we frequently experience here (at least one or two fall each year... and these are 18-24" diameter specimens!). Oleander is a favorite as a hedge/privacy screen. But, it's toxic (think: dogs & fires) and pretty boring. Cactus? <yawn> We settled on what we have found to be a good mix of plantings that are evergreen, dog-safe, don't encourage too many bees, flowering, don't produce too much litter, etc. And, in the process, removed all of the original plants as they were either inappropriate for the property (tall trees on small lots don't offer much of anything!), no longer permitted within the city limits (as new plantings... so, "established" is just as bad), offered nothing of value to us (peach, pear, almond), etc.
>> Pick an orange off the tree and eat it 10 minutes later. Compare >> to *anything* store bought and you will see why we spend the time >> and effort we do to grow them. Along with the *size* of the fruit >> (our limes are as large as store bought oranges; oranges as large >> as grapefruit, etc.) > > Those plants require tons of water, doesn't make any sense to grow > them in a desert climate.
Doesn't make sense to have a back yard that is 60% swimming pool, either! Yet about half of the homes here do. Few of them use "covers" to control evaporation. So, the water they consume just "increases local humidity" (i.e., doesn't produce fruit, greenery, etc.)
> But I suppose you could get away with a small number if you do things > like select grafted root stock dwarf variety, possibly use in-ground > grow bags to mechanically condense root mass, use subterranean drip > irrigation, surface mulch/anti-evaporation measures, and possibly > shading nets for the extra hot months.
We had mostly "semi drawf" specimens, originally. They proved to be larger than we expected. We lost three of them (lemon, lime, blood orange) in a particularly cold (for here) winter ~2 years back. We took that opportunity to replace them with dwarf varieties. Soil here has very high clay content. E.g., dig a hole, fill with water. Come back next day and you've ONLY lost water to evaporation (no drainage). As such, each of the citrus trees were planted in what amounts to as a "large pot" (horticulturist at nursery provided that answer when I asked how large a hole I should dig: "Pretend you are sculpting a terra cotta pot, *in* the ground, in which the tree will *live*..."
>> City lot. 1/3 - 1/2 acre? > > That's so tiny there's no excuse for not just manually turning your > water on and off.
The point of the electric valve on the hose bibb was so the hose can be used to bring water *to* a particular location in the yard without having to add or move drip lines/emitters. E.g., the three new citrus plantings can't be serviced by the existing irrigation lines -- their rootballs don't "reach" the location of the emitters (and won't for several years). Standing outside *holding* a garden hose in 100+ temperatures daily is not my idea of retirement! :>
>> Instead, build a water treatment plant. Tax folks to pay for it. >> Then worry when you will EVENTUALLY run out of water and the >> whole mechanism grinds to a halt. (I think municipalities here >> have to certify a 100 year water supply. Failing to do so >> puts bonds, etc. in jeopardy) > > It's all about job preservation with those people.
Actually, I think they just don't consider it *their* money (in terms of "coming out of THEIR pockets) and, as such, are very comfortable rationalizing the spending of other folks' money. They are putting in a "street car" downtown (or, what passes as "downtown", here). It goes from nowhere to nowhere_else. Has torn up the roads. Cost millions. No one expects it to see any ridership. Heavily subsidized. Etc. "Um, what's the point of this exercise? To cut down on traffic? To add a more 'economical' means of moving people (who don't want to walk those few blocks)? To cut down on exhaust emissions?"
>> Yet, come Monsoon season, everyone drops what they are doing and >> walks outside to watch it rain. ("Hello! Does anyone see the >> inconsistency, here?") > > That 10" is ridiculous, the place is not exactly a garden of Eden.
"Sonoran Desert". Key word is "desert".
On 9/19/2013 8:50 AM, rickman wrote:
> On 9/19/2013 1:59 AM, ehsjr wrote: > You make a good point with that. The issue with the failure is not > really such a big deal. His controller will be able to sense if the > button is shorted and notify the operator. The button is very unlikely > to fail shorted, but there are plenty of other failure mechanisms. If
I've not claimed that I *couldn't* detect a "shorted" button. Rather, my objection to this approach is that the "short" implicitly propagates to render the solenoid unusable. As such, the "problem" has to be fixed *now* (or, "before the solenoid would likely need to be actuated"). This, IMO, is just a crappy design philosophy. Failures should have *confined* consequences (Would it have been acceptable if that failure "took down" the ENTIRE irrigation system? If not, then why is it acceptable to take down that *valve's* functionality when the button's role is not inherently related to that?) Imagine I can *telephone* the user AS SOON AS this condition is detected. What good does that do him if he's not "at home" to address the issue? If he *is* home, does he have to hire someone on "emergency" service to come right out and fix this problem? Or, can he wait a week/month? *Or*, even decide to live WITHOUT the ability to "signal for water" with that button -- all else continuing to work as intended?
> the system needs to be robust then these other failure mechanisms should > be detected. Probably the most important thing would be to just plain > detect if water is coming from the spigot or not rather than detect all > the individual failures.
There's a difference between "check engine" and OBD functionality. The latter has far more value than the former. The more information you can provide to a user, the more useful you can be (even if the user doesn't have the inclination/skillset to effect the repair, he can at least convey the information to a vendor/service provider who could then give a more detailed estimate of the cost of the repair ("Well, it'll cost you $100 for us to drive to your house. Then, whatever else we need to actually fix the problem -- time and materials") Also, the more the system can discern about the consequences of a failure, the better it can adapt to that failure. E.g., if the valve is shorted to another, then it can avoid "over/under-watering" at least one of those zones whereas acting as if both valves were still isolated (from each other) would result in both zones being overwatered. Or, possibly, *neither* getting watered properly -- depending on the consequences of trying to "open" those two hydraulic circuits simultaneously (you may not have enough pressure to cause water to flow as intended from each circuit)
> Yes, a pressure switch *after* the valve could > do that and the pressure switch could be connected to the power wires in > a similar manner to provide a result back to the controller. The push
I've approached this by watching water consumption "at the source". If I open a valve and there is no change in flow rate, then I know water isn't flowing. Because the valve is mechanically defective, because a pipe/filter is clogged, etc.
> button shorts the solenoid, the pressure switch has a resistor in > series. So now you have three levels of voltage you need to sense to > detect the button and the pressure switch.
You actually need more than that to be able to detect all of the *likely* failures you will encounter.
> I would do this with a current driver circuit. It would have a constant > current drive allowing the voltage to float depending on the resistance. > Sense the voltage out of the driver and you are done. 24 volts is > solenoid operating normally, 16 volts is solenoid operating and pressure > detected in the output line, 0 volts is button pressed or solenoid > shorted. You could use a low level of current to sense the switch when > the water if off, or you could pulse the full current briefly, too short > to activate the solenoid.
This is essentially the direction I've been pursuing. The problem is getting the "data" back across the isolation barrier "cheaply". I.e., you don't want to put much on the "exposed" side of the barrier as it represents increased propensity for failure. My first approach was to use an iso-optilator in its linear region. That can provide lots of information -- more detail than you really need! But, coming up with a design that can reliably ensure you are using *only* the linear region of those devices seemed a bit dubious ("select/tweek at test" is not an option; esp if Joe Tinkerer opts to substitute some other device that he happens to have on hand -- not realizing the "special needs" that are imposed on the device in the circuit). So, I'm now looking at relying solely on operating them as digital devices and conveying that information in the time/frequency domain. This should keep the actual interface robust, inexpensive and easy to build.
> Don is a tough cookie to discuss things with. You offer some help and > he pulls more requirements out of thin air.
We *all* make assumptions. I don't even think twice about whether or not something like this would have to be electrically isolated. "It goes without saying". Similarly, the desire to keep the amount of stuff on that "exposed" side of the isolation barrier "robust" (surely nothing that is going to fry the first time there's an electrical storm, a large inductive kick in the cable, etc.) Likewise, that multiple valves could be active at the same time. Or, that "broken wires", "shorted solenoids", etc. are likely to be encountered in this sort of application (from knowledge of the types of mistakes DIYers make when doing these things). And, the sorts of things that a "system" existing in this sort of environment is likely to encounter -- that you'd not consider if you approached it solely as a technical problem. E.g., a little kid standing there pushing the button, repeatedly or continuously, because he can *see* that each button press is causing the water to stop flowing from the hose -- a consequence of Ed's design (OTOH, if the button is an "independent function" that the controller must *interpret*, then the controller, after seeing several sporadic button presses, can opt to treat the button as "suspect" and disable that functionality -- wildlife? packrat chewing on wires?) [*You* may, OTOH, assume only the left limit switch can possibly be engaged while "moving left" (see below).] Folks frequently piss and moan that I'm too verbose. Yet, if I present a detailed specification, do you think that will be any *shorter*? "Why do you need clause 3.8a?" "Can you eliminate clause 9.6b?" "Why does it have to work in 100% RH? Are you planning on using it UNDERWATER??" Also, the more you specify, the more you close the door on other possibilities. Possibly incorrectly! I had a colleague go through a very detailed analysis to convince himself that a certain observed symptom could *not* possibly be caused by something I was suggesting. Not accepting his decision -- he was actually my boss/client -- I *demonstrated* that this was, indeed, the source of the problem! ("Where is the error in my analysis? Let me check my math, again..." "Why bother? Do you doubt what you are *seeing* in my 'proof'?")
> But for the most part he is > brainstorming and that can be part of what we are seeing. He likely
I think folks don't think "twice" before responding. I.e., this isn't a place where folks come to ask "simple" questions. And, for the most part, it's not "high school kids" with no background trying to get answers to homework problems. Or, equally clueless homeowners tying to fix their own furnace, etc. When I respond to a post, I assume the querant has already thought about the problem. So, any *obvious* solution that I see gets double or triple scrutiny: "Surely he would have thought of this, already. Dig deeper for a solution." (e.g., my comment to Joerg, recently, about using a lawnmower engine as an "air pulser") Note my original post explicitly stated "shorting the coil". So, it should be rather obvious that I'd already considered that as a potential solution -- and moved *past* it (I didn't ask "how do I detect when I deliberately short a coil"). Finally, I consider engineering to be the art of finding the LEAST WRONG solution to a given problem (i.e., there are no "ideal" solutions in the real world). So, I find *my* best and then hope someone else will have a *cleverer* approach. Or, see some issue that will eat my lunch that I've neglected.
> didn't think about the failure detection as an initial requirement, it > occurred to him after he started thinking about how to implement what he > wanted. So it became a requirement because he thinks it will be easy to > include whether it is very valuable or not.
Actually, it's the other way around. I *assumed* everyone would have considered that as it's been an integral part of my design process (and the folks I've worked with) for decades. My favorite example: driving a motorized mechanism. Often, "limit switches" (of some sort) exist to signal the controls when the mechanism has reached end-of-travel in a particular direction. Most folks tasked with driving such a mechanism would write: case (command) { "left" => while (!left_limit) { drive(LEFT); }; "right" => while (!right_limit) { drive(RIGHT); } * => # can't happen } I, OTOH, would build a little state machine that modeled the mechanism: "at left limit", "at right limit", "traveling left", "traveling right". Then, monitor *all* inputs in each context. So, if I think I am "traveling left" and I see the left limit switch activated, I know I have arrived in the "at left limit" state. OTOH, if I see the *right* limit switch activated, I know something is VERY WRONG! I was not *in* the "at right limit" state (so, the right limit switch could not have been active). And, I am driving to the *left*. So, how could the RIGHT limit switch become engaged? Clearly, the software's model of reality no longer corresponds with "reality". Because we are actually *moving* things, there is a real risk that physical damage/injury can occur when the software is out of sync with the real world. So, go immediately to a "safe" condition -- without relying on any of these models! I.e., "CAN'T HAPPEN" *has* happened (because someone miswired the drive signals to the motor, it's drive circuit, or the limit switch sensors, etc.). The failure isn't allowed to propagate (cascade, get worse, etc.). And, the "cost" for this protection is just NOT "assuming" that only the left switch is significant when moving left! [This sort of approach has served me well -- VERY well -- over the years. IME, bugs hide in the "CAN'T HAPPEN" parts of designs.] Similarly, the idea of relying on "dead plants" as an indication of a failure wouldn't even enter into my calculations. "Doctor, the patient in room 205 is dead".
> I like your idea. But I'm not sure that it will do what he wants if you > use a common resistor as a sensor. That would require that you > periodically turn off all solenoids and poll each switch. In theory you
And, do that fast enough to catch a "quick" button push (or, do you burden the user by telling him he has to hold the button pressed for 15 seconds to ensure it is "seen" by an ULF polling event?)
> could do this fast enough that active solenoids do not drop out, but you > would then be sending some moderately fast pulses along long lines > potentially generating a lot of RFI. I bet it would mess up AM radio > pretty badly. Maybe there would be a way to shape the edges to prevent > this. I guess sampling 4 times a second would be adequate, maybe even > just once a second. That wouldn't be so bad, so maybe my concerns are > not valid.
And this saves, what -- a few resistors? :> Make the "pin driver" circuit effective; then efficient. As with software, don't "prematurely optimize".