Thermal imaging while getting EM-measurements (full RF+antenna system)

[GTheis]

Hi,

I am very new with thermals. I've, however, to not only provide separate measurements from the chips I'm using (which have thermal sensors built-in) but also from the different parts and the assembly of the system. So what my supervisors (who know even less than me from thermal measurement) asked me is a concurrent measurement of EM characteristics from the whole system (Transceiver+antenna) and thermals, which I would need to chose a proper camera setup and justify why to buy it. Can anyone point me references about the measurement techniques I'd need to use?

Even further, I might need to provide the thermal measurement of the heatsink chosen (only the heatsink) to validate the measurement setup and datasheet. Can anyone point me what I would need for this?

Thank you in advance

[Vipitis]

you are probably better off with a thermocouple or two.

Looking at a heatsink with a thermal camera doesn't tell you much. There is reflectivity- and therefore emissivity issues. Heat links lose heat by conducting it to nearby air. Not radiation.

[bap2703]

To be more specific :

- heatsinks are typically made of high thermal conductivity material (usually metals) to move the heat from source to sink (usually air) with the least possible of thermal resistance
- high thermal conductivity usually have low emissivity => they radiate way less infrared energy than a blackbody. Worse, that emissivity is hard to precisely know = it's hard to relate the infrared flux you measure to the surface temperature.

The typical case is an aluminum heatsink : the emissivity can vary wildly because you have no idea of the surface finish. Aluminum oxide can range from low to very high emissivities depending on thickness and porosity (again typically, due to anodization).
Yes you can alter the surface locally to obtain a high emissivity (drilling a cavity or using a coating). But then you get a local measurement and lose most of the benefit of using a camera.

Hence Vipitis recommendation to use thermocouples.
If you are doing some heatsink only characterization and want precise absolute temperatures i'd go that way too.

Now it seems you might be doing some system level measurements (like a PCB in an enclosure). If you want to have an idea of where the heat is, yes you can use a thermal camera. You might even want to use the plastic wrap film to have a way to look trough a closed enclosure.

[Ultrapurple]

I agree with much of what has been said so far but would add the following. If you're working on a test system then it might be practical to spray everything with a (?water-soluble) black paint before attempting thermal measurements. That would mean everything had the same emissivity (so everything at the same temperature looks the same).

If you're only interested in the temperature of certain areas then a piece of black insulating tape provides a remarkably consistent, cheap and easily removable sensing surface. If you have a shiny thing (which is the very devil to measure with a thermal camera) then the black tape provides a reference point and it doesn't matter what the rest of the thing looks like.

From a practical standpoint, my (limited) experience of thermal cameras and high RF fields is that they don't mix very well. On the few occasions I've tied to look at antenna losses (in terms of 'what bits are getting hot?'), the moment I applied RF the thermal camera went screwy. This certainly happened to me at HF and VHF. That said, I was only using a cheap camera in an uncontrolled environment.

The bottom line is that if you plan to spend serious money on this (say >US$1000) then I'd recommend having a detailed description with the camera vendor. But if you are in the lucky position of being able to just throw say $500 at it there are some reasonable 320x240 or 384x288-class cameras from China that you'll learn a lot from. And possibly have some fun, too. And at the end of it all you can probably get most of your outlay back by selling the camera on eBay.

[coppercone2]

you need a shielded enclosure and a IR window for the camera, then have that window screened, and i don't know if it would be good enough with the screen.

Does anyone know how the translucent electrically conductive glass coatings do with RF?

[Bill W]

you need a shielded enclosure and a IR window for the camera, then have that window screened, and i don't know if it would be good enough with the screen.

Does anyone know how the translucent electrically conductive glass coatings do with RF?

Germanium is a metal - so solves that problem once earthed at the edge.

ITO should be fine, used for EMC on things like case windows to see LCD's through

Bill