Rob Beschizza at 1:54 pm Sun, May 1, 2011
ADVERTISE AT BOING BOING!
For my job, I’ve often had to accurately measure the temperature of photo chemicals. I’ve found that the inexpensive digital oral thermometers sold in Canadian pharmacies are surprisingly corrrect and accurate – human body temperatures match colour photo chemical temperatures. I’ve compared lots of them against known-good thermometers.
For three to five bucks, you can’t go wrong.
/as far as ice vs. boiling water: that’d have to be distilled water and at sea level, eh.
I notice something similar with small compasses sold in bins. Either they all point in different directions, or if you pick up one and turn it, the needle turns with the rest of the compass.
These are not the thermometers you are looking for….
I have a similar problem with Lowe’s hammers. You would expect them to drive the nail in, but the one I got just ends up bending the nails or hitting my fingers. Chinese crap.
You know, I clicked this expecting an article about devices made specifically for measuring the temperature of feces. >_> No, I don’t know what’s wrong with me…
Anyone else notice that at the end of the videos one of the thermomitors seems to have leaked all over the place. Big red splodges…
Reminds me or rulers. While measuring a preprinted form I found that the rulers in our office differed by up to a quarter of an inch from each other.
If they’re inaccurate, are they at least precise?
Good test; unsurprising results. I’ll note that I’ve seen plenty of oral thermometers that weren’t much better. I have a modern thermometer from a reputable manufacturer that never changes its reading. It’s always at 99.5 F.
To my sorrow, my antique glass-and-mercury oral thermometer broke about a month ago, victim of a freak accident. You can’t buy those anymore on account of mercury being a lot more toxic than they understood at the time; but they were far better instruments.
Theresa, try here:
Sorry about the long link. It’s a search on Fisher Scientific that returns mercury thermometers. Another place to buy is VWR Scientific:
“You can’t buy a mercury thermometer” is such a commonplace meme that we forget to check it. Sort of like “they banned DDT”. I often wonder how many of these I carry around in my mind!
It’s always at 99.5 F.
That must have been Suzanne Vega’s thermometer.
I’d give you +1, but I’m giving you +.4 instead. well played overall.
Damn you are right. Should have checked before posting.
god yes!!! almost as unhelpful as the people who work there.
and don’t be fooled by digital ones either, unless you get a reliable brand
Most of the “clockface” type dial thermometers have an adjustment in the back so that you can calibrate it to what the actual temperature is.
And how do you know what the actual temperature is?
There’s an app for that.
You stick the thermometer in a bowl of slightly thawed ice, wait some time until the reading settles and calibrate it to 0ÂºC (or, for those of you in retrograde countries, 32ÂºF).
Why slightly thawed? Because ice out of a cooler will be at a bit less than 0ÂºC (somewhere around -5ÂºC), while water/ice at the phase transition is at precisely* 0ÂºC throughout the thawing.
* Yes, not precisely due to a number of factors, but precisely enough, which is the point here.
I’m sure someone with more thermodynamics knowledge could explain this more clearly, but it’s possible to calibrate relative to the ice point or boiling point of water (a task that can be performed in any household). I’m sure these values vary with altitude, but it should be accurate enough for placing the thermometer in your garden. This is the type of concept assumed to be understood in undergraduate chemistry lessons (most of which I’ve forgotten).
If the accuracy of the reading is actually important, two useful calibration points exist (one of which is practical for outdoor thermometers): a bath of ice and water is going to be around 32F/0C and boiling water is going to be around 212F/100C (near sea level).
The again, I prefer to wave my hand around outside the window and say “yeah, it’s around 17 degrees C today” and pretend that I’ve got mercury for blood.
The shape shifters are everywhere!
I prefer to open the window, wave my arm around and say “who cares what temperature it is, is it warm enough for tshirts, or do I need a jacket?”
Thing is, a temperature that feels freezing at one time of the year will feel super warm at a different time of the year.
If your buying one of those thermometer’s, accuracy isn’t your most important objective. But I’m sure you could get it close by using or borrowing a different, more accurate, thermometer.
Might these thermometers include instructions to calibrate before first use? That said, if it requires frequent calibration it might be crap to begin with.
A few of them just brought in from a non-air-conditioned storage room? But that 100+ one. Wow! Unless returned from someone’s trunk, really bad.
These thermometers were not intended to make factual measurements.
The glass thermometer reading 118F had a broken or ruptured bulb.
My grandfather has an electronic home weather stations. There is one wireless temperature sensor outside and there are two remote display units inside. Each display unit indicates a different outside temperature. For some reason each interprets the signal from the temperature sensor differently. In all fairness to the manufacturer the difference is not nearly as significant as the difference in the video above.
However, I would really like to see two or more of these systems side by side.
Are you sure the two indoor display units aren’t receiving signals from different outdoor sensors (maybe a neighbor has an outdoor transmitter that one of them is picking up). That would be my guess, rather than the two indoor ones receiving the same signal, and displaying a different temperature. You can test by bringing the outdoor sensor indoors and seeing if both receivers change as expected.
An interesting experiential would be to add them up and average them out. The theory goes that with enough wrong answers you come up with the correct results. Maybe they just intend you to buy ten of the cheap ones rather than just one good one. That’s why we have so many dim-bulbs in congress.
That could work, but if there was any sort of systematic bias in the design of the thermometers it would fail to produce an accurate reading.
If they’re all ten degrees too low what you’d get from an accurate reading with random error, then even if you average them out you’ll be off.
i’ve been through cheap and expensive, they are mostly junk. my fave is old metal ones with glass and mercury, form “general store” in North East Texas, it worked well for years. had it nailed to a post outside my window.
All thermometers need to be calibrated. One of the first things you learn in food safety is how to calibrate your cooking thermometer. Not sure how you do it with those big thermometers but it’s not surprising or unusual to have a stack of uncalibrated instruments show totally different results.
Even a broken thermometer is right twice a day.
also watching it again the majority of them see to be +/- a few degrees of each about the accuracy you expect from a cheap outdoor thermometer.
For some reason, instead of thermometers, I see the rolling stones on the main page…anyone else?
oh yeah? and what temperature are THEY?
but seriously, folks; this reminds me of the last one of these I bought. My reliable thermometer is a cheapie like in the video but it’s under the overhang outside my front door. I wanted another–one of those suction cup jobs–for my bedroom window so I knew how to dress. Brought it home and thought it was great until the first sunny day, as this thermometer was hit by direct sunlight.
I opened my curtain that day to find it was 120F. Insult to injury: the suction cup gave out a week or so later. Six bucks just layin’ on the ground. We should get an anti-CoolTools list going: Fucking Wack Tools.
Me too. 2,000 light years… I probably enjoyed it as much or more than I would have the thermometer video (had I viewed it).
Buy 23 of them and then figure the average…
this test was unfair: the frontmost thermometer was probably in the sun, where it would have been solar-heated to some arbitrary (high) value. (My homemade solar panel can easily exceed the boiling temperature of water, and all it is in black metal in a glass-fronted box.) the thermometers in the back show more consistency because they are further from the source of heating. it’s not fair to test thermometers except in the shade.
The video appears to be taken inside. There’s no sunlight falling anywhere near the thermometers.
rob, the windup key is on the opposing shelf. these thermometers are not intended to be accurate if they’re not properly wound up.
It would be hard to immerse one of those thermometers in water in order to calibrate them to freezing and boiling points.
I grew up thumping thermometers. These need a good thumping too,
Probably manufactured by a company called Heisenberg.
What about if it’s in the sun? Of course that would change the temps, and not be the true temperature.
You would expect these things to be accurate to within 5 degrees, yet some of them are off by 30 degrees.
They may not survive the shipping process. Thermometers are delicate, at least the bimetal strip and bulb varieties, which is what these two models are.
But note the price – $4 each doesn’t cover a lot of QC, factory testing/calibration or even the return-for-repair cost.
You get what you pay for.
Heh, this is why I have two therometers. They are usually off by a few degrees (most five). :P
thanks, Made In China.
I work translating products for Lowes: these thermometers include instructions for calibration.
Clock face ones are meant to be adjusted before use.
The liquid thermometers probably have a gap in the liquid that formed during shipping. You can just attach a string to the top and spin them fast for a minute, that will force all the liquid to the bottom and they should be fine after that.
These unstylish, inaccurate crap thermometers are raping America’s sense of temperature. If you want a thermometer you can trust, and one you can be proud to be seen in public with, get a Donald Trump Thermometerâ„¢.
A country needs to have standards. At a certain point an orchestra needs a conductor / a baseball team a captain.
Think of how many manufacturing jobs could be created if some of these
products were junked.
Consider that thousands of these poor quality thermometers are being sold daily to a population a 1/2 a billion people.
Be outraged by how this reflects on scientists/meteorologist/climate change all because middle manager with an economics degree gave a real scientist the finger to sell these to the public.
If you care about science be insulted.
Well, I think you are right to be outraged, but there is a larger issue here. The U.S.A. is really becoming a service economy. And most folks donâ€™t have practical need for stuff like that. Itâ€™s almost a novelty for folks who really donâ€™t need to know what the temperature is but somehow think that is â€œcoolâ€ to have an object like this. I mean I understand the concept of calibrating scales and thermometers but do most people? I scored an awesome vintage 1950s scale at an antique store because the person who saw it didnâ€™t understand how to adjust a scale to zero. I am actually going to use it to weigh items I ship and I didnâ€™t get the â€œvibeâ€ that the other person would put it to practical use.
Instead of getting angry at stuff like this, I have learned to reap the benefits of a technically ignorant world. Just keep on throwing out your perfectly good stuff that really needs either a tweak or a fix here or there. Iâ€™ll gladly take it and make it as good as new! And maybe then resell it on eBay so someone else can enjoy.
Consider that thousands of these poor quality thermometers are being sold daily to a population a 1/2 a billion people.
The cattle are Lowe’ing.
You can do a similar test with spirit levels at a big box hardware store, and get similarly poor results.
Use the “Rule of Three” and compare the flat part of the level to determine which of the levels have a perfectly flat measuring surface. Cheap cast metal levels and levels with carrying holes often have a warped bottom surface. Thus, your reading for level will vary based upon how your level sits on the surface it’s measuring.
Next try every level on the same spot and angle on the floor, which is usually smooth flat concrete at Lowe’s and Home Depot. You’ll notice that most of the levels show a different inclination. The cheapo plastic ones tend to be off by quite a bit and have no calibration.
I went through this whole process with every level in a Home Depot one time and it took me 10 minutes to find a reasonably good tool, a Johnson solid body aluminum level. It was the last one in the store, and was apparently a discontinued model.
In the South you really only need to know the humidity.
It’s either going to be hot, or sweltering and so humid you can barely breathe.
In the South you really only need to know the humidity. It’s either going to be hot, or sweltering and so humid you can barely breathe.
Same here in Palm Springs. The humidity is 5% and you bleed from your nose or it’s 2% and you bleed from your pores.
There’s a reason why laboratory thermometers cost an arm and a leg.
You could make use of your electronic skills. Arduino + lm335? I think the specs for the lm335 is about +/- 1C. Although, in practice, my experience, it’s more like +/- 3C. The LM335’s about $3 bucks, or you could grab the F version if you want.
RE: Anon. :”My grandfather has an electronic home weather stations….”
They’re generally designed to be time delayed. Samples at different times would give you slightly different measurements. You can’t have them talk together all at once. You could however get each of them to grab measurement samples at a much closure time.
Now, I’m not saying these thermometers aren’t junk, but they appear to be bubble packed, and sitting under a bright light: they’re being heated in a non uniform way by the lights in the room. If you’re going to criticise a product, you really need to perform proper tests.
It’s obviously due to local weather variations within the box.
Assuming they came in bins of a gross, and most buyers did similar comparisons when buying, the only ones left are the really f’d up ones.
Or perhaps they were precision culled at the factory. For example if you buy a thousand resistors at 5% tolerance they will all be between 5% and 2% accurate. The factory pulled all of the more accurate ones out of the batch and sold them at a higher since they’re more precise, though they were all made at the same factory the same way. (See also CPU’s priced by speed.)
Mail (will not be published) (required)