Click here to go to our main page on temperature effects
Click here to go to our main page on transmission lines
Click here to go to our page on coax temperature modeling
New for June 2020: this is an often ignored topic. Cables, microstrip, waveguide, they all change with temperature. And they always get lossier when they are hotter, unless you go out of your way to make a transmission line with some metal that has a negative temperature coefficient of resistance (TCR). There are surely some composite thick-film pastes where this would be possible (see our page on temperature variable attenuators) but it would be like a cheap bar-room trick like "how do I remove the dollar bill from under the upside down beer bottle without making it fall over?"
Cheap bar tricks
Note that in many aspects of engineering, we talk about temperature coefficients such as expansion (TCE) and resistivity (TCR). These are merely the slope of a nonlinear behavior at a given point (usually 25C). If you need a higher accuracy prediction over a wide temperature range, you may have to fit the behavior to a function, like a fifty-order polynomial, and use that function in your analyses. Excel can do that model fit using a trend line, while you eat donuts and pretend like you did real math!
Length change
Length affects transmission phase.... of you were making series fed antenna feed, you might see the antenna scan slightly when you heat it up! You can model transmission line length change using a simple temperature coefficient of expansion (TCE). On a copper board, this might be 17ppm/C. For the military airborne temperature range of -54 to +85C, you should expect a 0.24% change. This might not have a big effect on your design, except if you were designing a time delay unit and you wanted to maintain accuracy over temperature. In that case you might want to install a heater on your longer delay lines to keep them at constant temperature, or design them on something that has low TCE, like quartz.
Resistivity change
Here you can model the loss change by varying the resistivity of a transmission line with temperature, using the material's TCR.
New for June 2020, we updated our coax spreadsheet to include the ability to account for resistivity change with temperature and make loss versus frequency plots at three temperatures. Here is a new page that describes it.
Dielectric properties change
This is an effect that is not well studied, and is usually ignored. Loss properties of dielectric may change, but that change is probably swamped out by metal loss effects. The permittivity of the material may slightly change as well. Unless you are working at NIST you probably won't need to think about dielectric change over temperature.
One effect that is common is that dielectric expands more than metal when it gets hot. In the case of coax cables, there may be permanent changes when dielectric squeezes out the ends. If you are looking for high temperature cables, consider silicon-dioxide-filled cables. Get your wallet out, they start at $1000!