Temperature dependencies of electronic components

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: luketeaford, Joe., lisa, Kent

Post Reply
User avatar
Grumble
Super Deluxe Wiggler
Posts: 1665
Joined: Thu Dec 17, 2015 5:24 am
Location: U aint much if U aint Dutch
Contact:

Temperature dependencies of electronic components

Post by Grumble » Thu Sep 12, 2019 7:55 am

I am asked by a colleague to say something about temperature dependencies of electronics.
Is there a pdf or something like that floating around with some basic knowledge about this subject?
Googling gives me a lot of hits but no "textbook" type of article.

Can somebody point me in the right direction plzz?

User avatar
guest
Super Deluxe Wiggler
Posts: 4330
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Thu Sep 12, 2019 12:49 pm

was anything more specific asked for? if i had to give a short presentation on temperature dependency, i would probably look at a just a few fundamental components: resistors, capacitors, transistors. you can pull up some datasheets for these, and present the graphs inside that show how key parameters vary with temperature. i would show data for 3 different types of capacitors: plasitc film, ceramic, and electrolytic. for resistors, the cost has come down so much that there is almost no need to use anything other than metal film resistors (at least for SMT), which are quite good. and for transistors, there are a number of parameters that vary, but current increases for increasing temperatures, which can lead to thermal runaway. the temperature effects of ICs are then just a result of the temperature effects of the transistors they are made out of.

i wish i had a single reference, but most of this is from a few paragraphs in random textbooks.
openmusiclabs.com

User avatar
Grumble
Super Deluxe Wiggler
Posts: 1665
Joined: Thu Dec 17, 2015 5:24 am
Location: U aint much if U aint Dutch
Contact:

Post by Grumble » Thu Sep 12, 2019 1:51 pm

Thank you for your input, but reading datasheets would work for me, but not in this case.
The colleague is a phd doing research with a device designed for measuring the circumference of the torso.She discovered that there was a drift in the output signal of the device and she thought (correctly) that the drift was temperature dependent.
But now she has to explain why she thought it was drift caused by temperature change and not by some other phenomenon while to me it is quite obvious.
It’s like that there is a difference between a man and a woman obvious to everyone, but explain that scientifically is a problem.

User avatar
guest
Super Deluxe Wiggler
Posts: 4330
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Thu Sep 12, 2019 2:10 pm

well, you could collect data with a temperature probe and show that its related, although that wont give insight on how to fix it.

do you believe the drift is due to the circuitry or the sensor? if its in the circuitry, you can add up the drifts of all the components (from datasheets) and then give an overall drift of the circuit. the sensor would be more difficult if its a custom device.
openmusiclabs.com

User avatar
Grumble
Super Deluxe Wiggler
Posts: 1665
Joined: Thu Dec 17, 2015 5:24 am
Location: U aint much if U aint Dutch
Contact:

Post by Grumble » Thu Sep 12, 2019 2:54 pm

both the sensor and the measuring device (which is a black box to me) are new in methode and design.
in my lab there is a temperature controlled oven, and she already discovered that there is a temp. dependancy in the drift when the measuring device is in the oven when there is a temp. change while the output stays the same with no drift if just the sensor is in the oven with a changing temp.
So that is not realy the issue, I was more looking for a pdf or something like that with some examples of temp. dependence in electronics.

User avatar
guest
Super Deluxe Wiggler
Posts: 4330
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Thu Sep 12, 2019 4:57 pm

the "precision circuit techniques" chapter of horowitz and hill is pretty good at covering basic opamp circuits, and where temperature drift comes from. if you learn anything more about the "black box" i might be able to point you toward more specific resources. also, if its an offset drift, its typically the powersupply or opamps, if its a scale drift its typically the resistors.
openmusiclabs.com

ixtern
Common Wiggler
Posts: 228
Joined: Mon Jan 22, 2018 4:34 am

Post by ixtern » Fri Sep 13, 2019 1:02 am

It's not so easy. Particular parts may be temperature dependent but whole circuit may be or may be not.
Example: if you take two resistors of the same type - both are temperature dependent but in classic voltage divider scheme absolute values are not so important as relative ones. If both resistors have the same temp coefficients, circuit may be temperature independent.
Another one: capacitors. Good mica capacitor may have positive (e.g.+70 ppm/deg) and styroflex capacitor negative (e.g.-160 ppm/deg) coefficients. Combining two of them with proper capacitance relation may result in zero temperature coefficient.

Transistors and diodes have exponential temp relation but there are methods to compensate it (we all know transistor pairs and tempco or oven in VCOs).

What I am trying to say is that all depends on circuit design, not on components only.

User avatar
Grumble
Super Deluxe Wiggler
Posts: 1665
Joined: Thu Dec 17, 2015 5:24 am
Location: U aint much if U aint Dutch
Contact:

Post by Grumble » Fri Sep 13, 2019 1:21 am

Thanks,
@guest: I have found a free pdf version of that book and sent it to my colleague.
Since we do not have the schematics of the measuring device she has to stay with common behavior of electronic components.
Thanks for thinking about this!

jorg
Veteran Wiggler
Posts: 738
Joined: Fri Apr 03, 2015 9:38 am
Location: East Coast USA

Post by jorg » Fri Sep 13, 2019 9:21 am

Your colleague should be using an instrumentation-grade amplifier. These are commonly used in medical equipment.

User avatar
chuckg
Common Wiggler
Posts: 69
Joined: Wed Dec 05, 2018 11:01 pm
Location: Florida Space coast

Post by chuckg » Tue Sep 17, 2019 1:43 pm

I would think that proving that the measurement device is temperature-dependent is good enough for the work, without need to explain why.

More important is to take steps to reduce the dependency, or account for it in some way.
….working on it.

Post Reply

Return to “Music Tech DIY”