Temperature sensing technologies

(sigh) More limited imagination.

A thermistor that always produces the same result for the same temperature environment need have NO bearing on an actual temperature: "this reading corresponds to *some* temperature. When I see it again, then I'm seeing THAT SAME TEMPERATURE (to the resolution of my digitizing device).

The thermistor (in this case) doesn't limit the precision of the measurement -- the "digitizing" circuit imposes those limits independent of the "sensing element".

Say I want to design a thermostat that will keep my house at whatever temperature it happens to be at *now*! Should I find an NBS traceable sensor and install that with a super high precision digitizer?

Or, should I purchase something inexpensive -- but repeatable -- and wire it up, LOOK at the output *now* and say "this is my target". Do you care if it's 73 degrees or 85 degrees in the house? Will the output NOT be monotonic? If the sensed value is "low", htat suggests I have to turn on the heat; if its high, that suggests I've got too

*much* heat.

All the information you need to control is present. And, if you're smart, you can optimize that control by just "taking copious notes" each time you engage an actuator.

What's the stated accuracy on your COTS thermostat? Or, the temperature controller in your freezer? Automobile water jacket? Do you KNOW? Do you *care*?

Our freezers have dials that say "cooler warmer". We tweek it until its correct (20 years ago) and leave it -- confident that it will repeatably maintain that temperature.

If we were interested in the most *efficient* temperature setting, we'd be looking at specific temperatures (that can still be expressed in REPEATABLE bogounits) and specific *power* measurements: this temperature uses this many power units; this other temperature (presumably equally satisfying for the frozen goods contained within) uses more/less power units.

The cars have interior temperature displays. And, that temperature reading probably reflects some point *inside* the dashboard -- not in the passenger compartment (let alone the back seat!) where it is pertinent to the occupants. It could just as easily have been a *letter* as a *number* ("It's warm in here. Turn the temperature control DOWN to F, from G")

Don't overdesign things and impose on the user ("Do you want your car interior at 72 degrees? Or, 73??")

There are no accuracy requirements *specified* because there are none imposed by the problem domain!

Reply to
Don Y
Loading thread data ...

1% Resistors should be adequate ??

...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
 Click to see the full signature
Reply to
Jim Thompson

Please no! For decennia, I was annoyed by fridge dials marked in meaningless units. Now that we have dials marked in degrees Celsius, I *don't* want to go back. Even if the accuracy is poor, it's still much better than a meaningless 1-to-5 dial.

Jeroen Belleman

Reply to
Jeroen Belleman

How many people can really tell you what their refrigerator's temperature is? How many can truly tell you what the temperature in their *home* is?

"1-to-5" isn't meaningless: 1 is colder (warmer?) than 2 which, in turn, is colder (warmer?) than 3, etc.

People here get edgy when faced with "metrification". I.e., temperatures will be "less precise" (if reported in whole degrees). Yet, I challenge anyone to tell me the difference between 25C and 26C *reliably*.

Instead, people think in terms of "warmer" (than THIS) and "colder".

If you want to be "accurate", at the very least, you'd say: "My thermostat THINKS it is 72 degrees AT ITS SENSING ELEMENT. The temperature elsewhere (even a few feet away) is subject to interpretation..."

Note the examples that I offered. Set your refrigerator/freezer/water heater to some particular temperature. Use an NBS traceable sensing/measuring element. *Hope* that it maintains that temperature "dead-to-nuts".

If *my* observational goal is to detect when the freezer has failed (or appears to BE in the process of failing), do I need that same "high accuracy" sensor? If i have a *repeatable* sensor and I "notice" what it typically reports for the freezer during operation, then any deviation from that typical is what I'm concerned with, not the magnitude of the deviation: "Hmmm... normally, the temperature would start FALLING at this point (whatever *it* may be, in SI units, unknown to me). But, I see it rising! Something is behaving in an atypical manner. Perhaps the compressor has failed? Perhaps the thermostatic control inside the device has failed? Perhaps *my* sensor has failed? *Or*, maybe someone just opened the freezer and added a dozen quarts of Marinara sauce that were previously at room temperature and the interior of is EXPECTED to increase in temperature. But, wait -- no one is home! Is it likely that a BURGLAR has done this?? Or, MORE likely that the freezer is failing?!"

[Think in terms of an expert system. Would it really be concerned with ACTUAL temperatures? Or, *relative* temperature changes?]
Reply to
Don Y

So, unlimited imagination: you don't need a temperature sensor AT ALL, because you have no accuracy needs AT ALL.

Or, you're incorrect in your statements, there really is an accuracy requirement, and you need to find out what it is.

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

(sigh) More fantasy from you to generate a discussion whereby you dismiss all the professional posts here to lead them on.

Reply to
John S

OK then. To what should I set a 1-to-5 dial to keep the lettuce from freezing? I don't want to iterate to get the right setting. That what we have technology for!

As for your freezer goal, I apologize in advance, but I think you're just a little crazy. I just want it to maintain a set temperature, ignoring brief transients, but have some indicator lit if something goes really wrong. Freezers aren't very good at detecting burglars.

Jeroen Belleman

Reply to
Jeroen Belleman

Don't believe that. It IS meaningless when you can't even tell the sign bit. My fridge (helpfully) has 'higher' and 'lower' indicators next to the number dial. Trouble is, I don't know if the TEMPERATURE gets higher, or the refrigeration does.

It would have been OK if the indicators said 'colder' and 'warmer', and even better if they indicated which was CW and which CCW. They don't.

Reply to
whit3rd

Honestly, have you NEVER altered the setting in your refrigerator -- despite having it *calibrated* in pseudo-accurate units? Did you iterate -- even JUST ONCE?? Why?

Does yours have such an indicator? Have you already *paid* for

*similar* technology? (Ah, I see... just buy new appliances and select from only those with these features. How smart is your furnace? Air conditioner? Washer/dryer/water heater? Or, are those on your "soon to be replaced" list?)

I want to detect when the freezer is failing or likely to fail (gee, its taking longer to cool than normal).

My goal is to enable folks to live independently for longer periods of time despite decreasing abilities.

How does a blind man decide that his freezer is on the fritz? FEEL every item to see if any of them are "unusually soft/warm"?

How does a man confined to a wheelchair accomplish this?

How does an elderly person with cognitive impairment do it?

Ans: someone else ASSISTS them with these responsibilities, despite the fact that they can otherwise maintain their own lifestyle independant of others (feed themselves, cloth themselves, remember to bathe regularly, etc.)

For folks without nearby assistance (friends/family), that means moving into a facility that caters to these "needs" (responsibilities) for them. At a greatly increased cost and sense of lost independence -- even if they are allowed to inhabit their own "quarters".

For folks WITH nearby friends/family, it puts them in a position of having to either pester that care giver (possibly needlessly)

*or* preserve their pride and live "suboptimally".

I have several friends/neighbors that fall into these categories. They are perfectly capable of wiping their own asses. But, can't handle even routine "problems" without calling for "professional services". At which time, they are often exploited: "You need a new furnace" "Your roof needs to be replaced" etc.

Twice a year I make the rounds converting from winter (heating) to summer (cooling) for them -- despite the fact that this is a no-brainer operation. They don't even *remember* that this must be done!

One neighbor came to me, confused, wondering why he had a $300 gas bill -- in October (heat is barely needed). I asked him if the furnace had been running a lot lately -- if he'd had the thermostat set too high...

Of course, he couldn't understand what "running a lot" should mean...

Realizing that his "faculties" were obviously in decline, I went to his house and verified that he'd not installed the baffle to block the air flow INTO the ductwork from the swamp cooler located on his roof ("winterize"). This is a no-brainer operation: slide a sheet of tin into a slot, turn off water supply to cooler, switch from COOL to HEAT. Maybe 3 minutes if you're lazy...

As a result, all of the heated air from the furnace was going up into the ductwork -- and then out through the roofmounted cooler! Of course, the thermostat saw nothing wrong with this: house is still not warm enough, continue to call for heat!

A *smarter* thermostat could have told him of this likely problem

*or* "phoned" some "responsible party" of the problem before he'd incurred $300 worth of "heating The Great Outdoors". [And, that thermostat wouldn't need to know what the actual temperature of the house was during that period! "Still not warm enough!"]

The *city* lost its natural gas supply one night a year or two ago. Plumbers were inundated with calls for service. They gladly scheduled all of those house calls -- and invariably said "your gas supply is out; there's nothing we can do about it. That will be $200, please..."

Of course, I suspect they also "suggested" new HVAC systems to a fair number of those folks!

*I* didn't bother calling a plumber because I could observe how the furnace was reacting to calls for heat: there's something wrong with the gas supply; nothing I can do about it other than report it to the gas company.

I then called other friends/neighbors to try to gauge the extent of the problem -- is it my block? street? neighborhood? (the idea of it being citywide never occurred to me).

If one of these people had been an individual for whom I had some *responsibility* (beyond "the kindness of my heart"), I could have queried their furnace to see if it was exhibiting similar behavior. Or, their "furnace" (thermostat == furnace controller) could have reported the problem saving them the expense and anxiety/uncertainty of the plumber's visit (many hours later!)

[*Our* furnace has a set of idiot lights -- INSIDE the blower compartment. PERHAPS they would have told me what I could deduce by observing the overall performance. *If* I could locate the manual, etc. Had I not been home, I'm *sure* SWMBO would have called a plumber!] *Hope* you retain your physical and mental capabilities. *Or*, plan on your kids (or your bank account) taking care of you when that time comes that you can't perform these tasks that seem almost laughably trivial, NOW! The $100-150/day (FOR THE REST OF YOUR LIFE) alternative is far too costly (and embarassing) for many!
Reply to
Don Y

formatting link
perhaps
formatting link

which seems to be the lastest version of the DS1820

--
This email has not been checked by half-arsed antivirus software
Reply to
Jasen Betts

The one with the meaningless numbers on the dial, yes. Because it froze the lettuce. The one with the display in degrees was just set to 5 degrees at that was it.

In this age of talk about internet connected fridges, we should at least get the basics right before the fancies.

Yes, my freezer has a green and a red status light. It's not even a recent freezer.

Jeroen Belleman

Reply to
Jeroen Belleman

That's still not adequate unless you check the lights hourly.

I had a freezer fail that I didn't know about until I opened the door... whew! What an odor!

An audible alarm or one that texts you would be nice. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
 Click to see the full signature
Reply to
Jim Thompson

In our case, we have a lot of money tied up in the freezer's contents. We'd hate to lose that "investment" just because something that could have been FIXED, failed without our knowing!

Repeat for:

- washing machine hose rupture (happened to us)

- toilet fill hose rupture (while I was asleep; thankfully light sleeper!) [both of these PROBABLY caused by excessive municipal water pressure; subsequently installed PRV *and* pressure gauges -- to be replaced by sensors -- on each side thereof]

- freezer failure (breaker; unnoticed because we didn't access it, then)

- condenser fan failure (caught because I noticed ACbrrr performing poorly AND HAD THE EXPERTISE TO KNOW WHAT TO CHECK; $9 for a run cap)

- natural gas supply failure (described elsewhere this thread)

- water heater failure (unnoticed because located in seldom visited store room)

- natural gas leak (noticed because of sensitive nose and verified with portable mass spectometer? -- some gizmo borrowed from Safety Officer at local hospital -- traced to faulty repair by plumber)

And, we're "on top of" things -- the house is almost always occupied. I can recount horror stories of neighbors who've had pipes burst in their attic, water heaters leak (located in interior rooms), etc. while they were away at work...

All of these things can quickly be addressed -- and often inexpensively -- but only *if* you know they are happening, WHEN they are happening! Sure would be nice if someone (something) was "on duty" 24/7/365 to watch for these things AND SUGGEST POSSIBLE REMEDIES!

Reply to
Don Y

I can. I look on the door.

I look at my cell phone (or the thermostat).

That's the problem with my fridge at work. I don't know which way the numbers go. I just turn it until it's the right temperature (thermocouple stuck under a soda can).

1F is about the limit of senses. I can certainly tell the difference in two degrees, at least in the Winter.

One degree warmer *is* one degree warmer and it is noticeable.

OK, accuracy isn't all that important but resolution is.

It's relatively important for it to be constant. Milk will last longer the closer you keep it to freezing (without freezing it, of course).

Reply to
krw

Platinium thin film resistance thermometers - the substrate is usually alum ina IIRR - are compact and pretty robust. Platinum doesn't oxidise, so the parts are durable too.

The AD590 is a temperature to current transducer, which makes wiring resist ance variations something of a non-problem (as long as it isn't high enough to cut into the supply voltage).

formatting link
f

The LM334 is nice, but you might want to use a Kelvin - 4-wire - connection to save worrying about lead resistance. "Interchangable" thermistors aren' t all that expensive, but again you might need to go 4-wire to keep lead re sistance from being a complication.

And there are digital temperature sensors

formatting link

not that I've ever used one.

--
Bill Sloman, Sydney
Reply to
bill.sloman

ading

enger

They buy a fridge thermometer - you can buy them everywhere, so people do s eem to buy them - or wall thermometer - ditto.

That's not metrification, it's quantisation.

Sure. And that perception varies from moment to moment.

at

e

It's not subject to interpretation, but to thermal gradients, which can be measured. If you want to know the temperature at a specific point, put a th ermometer there. If you don't want too much temperature variation across a room, insulate the walls (so there's less heat flow across the room) or sti r the air (so the heat flow is carried by a larger mass of air across a sha llower temperature gradient).

ater

The nut involved is you.

Reply to
bill.sloman

Previous house had best solution for that... recessed floor for washing machine, with floor drain.

Garp, "Pre-disastered" ;-) ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
 Click to see the full signature
Reply to
Jim Thompson

Don Y has very little imagination, yet criticises people who respond to his posts for not "imagining" the data he didn't bother to include in his original post.

All thermistors suffer from self-heating. It is "obvious", but a father and son combination published a paper in Review of Scientific Instruments that ignored the effect, and still got past Rev.Sci.Instrum. physicist referees.

You really go to a lot of trouble to avoid doing any quantitative thinking.

Why bother? If you had a better understanding of what you were doing, you could optimise from first principles.

Understanding what you are doing in quantitative detail isn't "overdesign". It allows you to relate what you are doing to a whole mass of scientific data, which you find intimidating, but does happen to be useful.

Sure there are. You don't want your house temperature to fluctuate outside of your comfort zone - which is probably somewhere between 10C and 30C (if one imagines that you are willing to put on a sweater and gloves indoors).

--
Bill Sloman, Sydney
Reply to
bill.sloman

Search on W1209 on AliExpress and or Ebay. May or may not be what you want, but worth checking. On Ali the price is $1.48 including shipping. Requires 12 volt power and provides relay closure on Temp rise or Temp fall, not both.

Dan

Reply to
dcaster

I have a couple of TI evaluation boards (extremely small) that are RF remote temperature sensors. They seem to be repeatable and use little current (based on the MSP430 devices). The part number is EZ430-RF2500.

Reply to
maury001

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.