Finally got around to playing with the temperature probe on my cheapo DT33C multimeter.
It does seem to be taking a while to settle down. It's been in the chilly bin against the cooling plate for at least 5 minutes and is still making its mind up.
I intent to check the freezer in a bit (just defrosted, cleaned and getting back to temperature) then I may go wild and check the oven.
At the moment it is saying that the chilly bin is at 9C which I don't believe as it must e at 4C or lower judging by the temperature I feel when touching the bottles, and also because there is frost on the chilling plate.
Anyway, it seems to be stable(ish). Just wondering how fast it should change.
Hmmmm......double checking with my Thermapen 3 food thermometer. Perhaps it is more accurate than I thought. What I do note is that as soon as I open the lid the temperature climbs quite quickly (according to the Thermapen).
Five minutes sounds a long time but it depends on a few things. Presumably the sensor itself is inside a stainless can that may be 5mm diameter? That will slow things down. Also, will be faster if actually in contact with cold stuff (like ice in the freezer).
Maybe your thermapen is thinner. I have bare thermocouples in the ovens, their response is almost instantaneous.
Probe is very thin and wrapped in black tape (or summat). Looks to be a piece of thin wire sticking out of the leads which are wrapped in heat shrink. Almost like just the bare wires from the leads twisted together and blobbed with solder. I would have expected almost instantaneous.
Another fun thing. The probe was in the freezer which is undercover outside. It was registering -11C, but when I stepped closed and my shadow fell on the multimeter the temperature dropped to -15C. I checked by moving away and closer again and when diffuse sunlight fell on the body of the multimeter the temperature reading changed each time.
-11C in sun, -15C in shade. The probe was well down inside the freezer with the lid shut on the leads.
Sounds like a simple thermistor probe. Plastic is a good insulator so the heat shrink will cause a bit of a delay in response, but I'd expect seconds rather than minutes. The response of the meter to sunlight (or is it radiant heat?) is a bit odd. Martin Brown will be along soon with a comment.
The thermocouple will generate a voltage, and so will your meter terminals (the Reference junction). The reading is just the difference between their two temperatures, so when the sun heats the meter by 4C the reading will rise by 4C. The meter probably assumes itself is at a normal temperature like 20C.
If it is a thermocouple (which I doubt, it is much more likely to be a thermistor), it will be amplified by a circuit which includes temperature compensation.
NTC thermistor say 322K ohms at -50C, 10K ohms at 25C, 758 ohms at 110C There is an equation to fit this line, with two fitting params.
diode/junction Inject constant current (ohms scale), measure developed voltage, use maths to print a final F or C on the screen.
thermocouple bimetallic, potential difference
...
If a 1N914 was used, with a transparent glass envelope around the active element (diode), the diode junction could be light sensitive.
The observation about light, hints at what you've got. There is a semiconductor in there, with a transparent glass shell.
The diode Vf voltage changes little with temperature, say
0.7000V at 25C and 0.6000V at 50C. You have to measure small voltage changes, do the maths, and compute the temperature from that. This might make the measurement seem "a bit skittish" as these things go.
A good thermistor costs 20x what a cheap diode costs, especially a 1N914 (glass) or a 1N4148 (glass) versus a 1N4004 (black epoxy, no light effect). To meet a need to "measure a temperature in a roundabout way", using a diode saves on parts cost. The ohmmeter on the instrument, provides the measurement stimulus. In addition to a diode, you can also use a bipolar transistor junction. Perhaps tie base to collector, run a measurement current from collector to emitter.
formatting link
(
formatting link
)
Thermocouples are great, because they can have tremendous measurement ranges. But unnecessary for ice box work. Engineers who use those, love them, and I've seen guys hook up 20-30 sensors on projects at work (and then need a multichannel recorder while they go drink coffee). They're kind of fanatical about it.
Since the multimeter has a processor and firmware in it, it does not cost any extra to be doing maths. It can't be a powerful processor, because a multimeter might only draw 8mA from the 9V battery supply while running. If you have a second multimeter, you can even measure that :-)
And there are lots of sensor types, that are sensitive to having their legs bent, or receiving a sharp blow. Some sensors only have the one calibration point (if any), and could be thrown off by being crushed in a refrigerator door.
And some of the sensors, the packaging (the glass or epoxy or kapton film), it gets damaged at a lower temperature than the active bits. The diode might easily be able to take
135C, but the glass might crack if you tried to do that. The table I have for my thermistor, stops at 110C, which says to stop at 110C, and that one is a hard-shelled dipped epoxy with shiny outer body. Presumably this was intended to make it "moisture resistant" but not designed for a steam bath :-) Research the sensor limits, before torturing the probe.
I assume that the active element would be at the end of the probe (not in the multimeter body)?
If so, please note that the probe end is bare wire as described above.
If it is in the multimeter body then that is odd, as it suggests that the temperature measurement inside a device (freezer or oven) depends on the ambient temperature (or incident light) of the multimeter body.
I don't have the manual, I don't even have the hardware. I have a couple of thermometers that *do* use thermocouples, several others that use thermistors. I have a multimeter with a temperature probe, and that is definitely a thermistor. Although there is no reason why a multimeter should not have a thermocouple input for measuring temperature,if it did it would certainly include a temperature compensation circuit.
OK so it has a thermocouple. I said that it could be. I can assure you that commercial portable thermocouple instruments give pretty accurate results whatever their ambient temperature. Accurate (0.1 deg C) portable instruments were around at least 50 years ago. I guess they may have used semiconductors for the compensation.
You do of course use a temperature controlled "cold junction" for compensation in accurate scientific work.
I recently bought a new Brymen DMM and have also been checking the fridge and freezer.
DMMs have a temperature compensation circuit close to the cold junction, but it is critical that the DMM itself is acclimatised to the room before making measurements and not in a draught or direct sunlight. Perhaps the settlement you saw was the DMM acclimatising?
Have you tried shorting the DMM terminals without the probe plugged in? Does it show the correct room temperature? Is it stable?
formatting link
If it is like the Brymen probe (a type K thermocouple) it will be very fast to respond -- a second or less for a bare thermocouple.
I got best results measuring the temperature of a cup of water that had been in the fridge for some hours, or with the probe sandwiched between two packets of frozen veg in the freezer.
There is a good Linear Technology application note, "Thermocouple Measurement", Jim Williams, Linear Technology application note number
HomeOwnersHub 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.