Anyone aware of such a thing?
Clearly, one can achieve that just by putting a thermistor in series with an ordinary resistor, but my UPS seems to contain a single component that behaves that way - perhaps less than 1% at 20 Celcius - it's in the circuit that controls the battery charging float voltage.
Did they perhaps exist 20 years ago (when my UPS was made), for some reason?
Sylvia
Yes, I have purchased custom positive temperature coefficient resistors in relatively small lots (10K, if memory serves).
They're usually in around the tempco of metals and alloys. (a few thousand ppm per degree C, positive and fairly linear.
You can find a few values of nickel thin film resistors on Digikey that have a similar tempco.
Putting a resistor in series with a conventional NTC does not work so well over a wide temperature range because of the nonlinearity of the thermistor.
My impression was that nickel thin film resistors were the poor man's platinum resistance sensor. They had around the same temperature coefficient - the resistance was more or less proportional to absolute temperature around room temperature. They weren't as good as platinum resistance sensors - for one thing they oxidise if they get too hot. Wikipedia says that they go non-linear about about 300 Celcius.
fredag den 10. december 2021 kl. 12.50.41 UTC+1 skrev Sylvia Else:
PT1000/PT100 is ~0.4%/C
Propietary nickel alloy base metal sensors are used in HVAC applications by folks like Honeywell.
Ni120 sensors (wound and now thin film) are traditionally used in some applications like bearing and motor stator winding temperature (sometimes they use Cu RTDs for the latter).
The SMT 0603 etc. Vishay ones I was mentioning are pretty awful for sensors, almost a 10% tolerance on Tempco and a max temperature of
150°C minus self-heating so their applications would be limited. The ones I sourced from a Japanese factory had much tighter tolerance on tempco.
The ZNI1000 nickel RTD is a cool part. It can be linearized with one resistor.
Of course, ic temp sensors are cheap and can be had analog or SPI.
The only thing that comes to mind is a diode.
Cheers
Phil Hobbs
Carbon comp resistor, roughly -1200 ppm/c.
lead acid batteries are temperature sensitive when recharging
Yes, this is presumably the reason for the NTC resistor in the charging circuit. But if it has a typical thermistor characteristic, then in this particular implementation, the compensation will be out by a factor of ten, which is probably worse than no compensation at all.
I may have to remove it from the circuit to better characterise it.
Sylvia.
Stupider than Anybody Else wrote: ========================== >>
** WTF are you on about ????Themistors ( NTC or PTC) do not look like regular resistors.
Are you claiming KNOW there is a resistor ( banded or not ) that has a very large neg tempco in your UPS? Or just making wild, idiot guesses as bloody usual.
...... Phil
If it is a carbon composition or carbon film resistor it won't look much like a thermistor
It would still be a pretty horrible temperature sensor. The original designers may have been a bit optimistic about the reproducibility and stability of the part they put in.
Carbon film resistors might not have looked quite as cheap and nasty twenty years ago - they did to me back then, but then there were engineers around who saw them as cheap, rather than nasty, and there may still be some who think that way.
Where did I say it looked like an ordinary resistor? It doesn't.
Amongst other things, it's labelled on the board as NTC001.
It also clearly has a negative temperature coefficient, as demonstrated by the use of a multimeter and a hair dryer (yes, of course I have a heat gun, but the hair dryer seemed less likely to damage anything).
It's in series with a fixed resistor and a trimmer, which together with a resistor from the output to the adjustment pin of an LM317T, define the voltage on that pin, and thus sets the charging voltage for the batteries.
Is more evidence required?
Sylvia.
Stupider than Anybody Else wrote: ==========================
** In the heading - f****it. ** What other crucial facts have you omitted ??** OK, so that is one of them. ** Why did you LIE and hide info in your first post ? ** Charging and float voltages are not the same. ** None more needed to prove what a time wasting troll you are.
..... Phil
Oh, that's just about true for all classical metals; getting a non-PTAT (proportional to absolute temperature) resistor is a notable materials-science achievement.
At 20 C, proportional to absolute temperature means about 0.0034 per kelvin; for copper, resistance goes 0.00386, for tungsten 0.0045, for aluminum 0.00429, for platinum 0.0039 ... The impure (alloyed) metals used for low tempco have very non-ideal metallic nature.
The biggest Fuckwit on Usenet posts as whit3rd ====================================
** " very non- ideal metallic nature" ???????Could you please be bit more ambiguous ?? Cos some fuckheads here will imagine it makes sense.
...... Phil
Back when I was in paid work we got a marketing talk from Vishay on their very low temperature coefficient resistors which could get below +/-5ppm per degree Celcius, if you were willing to buy their most expensive parts (which we were, if not all that often).
They not only mucked about with the composition of the materials they laid down in their thin films, but they also figured in temperature generated strain by laying down material with one cofficient of thermal expansion on a substrate with a a different one. This is a longer way of saying that they got non-ideal behaviour. I can see whit3rd might have wanted to avoid testing your attention span (which isn't all that long).
The simple ideal metal doesn't expand or contract with temperature changes, hasn't any crystal defects, no bandgap changes, no magnetism, Einstein model phonon spectrum, and all resistance is the consequence of electron-phonon scattering. Then, if you know the thermal phonon population, THAT determines the resistivity. Alloys for low tempco have various deviations from these behaviors, so the thermal phonon population statistic is not dominant.
Phil doesn't understand what is being talked about, and feels hurt about it . He''s the witless moron in this context.
<snipped his characteristic statement of his bewilderment>
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.