Folks, I appreciate electrical items develop heat but I don't like devices that appear to get pretty hot, especially when voltage is low. Also aware it is the amperage or wattage that heats it up. The Flightaware tuners have a good reputation but around a year ago I sent mine back as I felt it was getting considerably hotter than the cheaper unit I was running. I have a Pi Zero W (WiFi on board) which, with the tuner plugged in, takes a total of 430 mA so even the combined total is a mere 2 watts. Why the heat? Geffers
Heat generation (and how to improve cooling) is one of the most debated topic about rtl-sdr dongles. Most of them use the R820T or R820T2 tuner chip by Rafael Micro, it is well known to heat noticeably while working. Maybe you can be interested to read this: https://www.rtl-sdr.com/rtl-sdr-heat-dissipation-as-seen-by-a-thermal-camera/
I've never tried to tune and decode TV channels on Flightaware stick, but I did it on a generic "no-brand" stick and I can tell you the heat was the same, so I think frequency doesn't really make any difference. If you google a little you can find a lot of comparisons between different brand and models of rtl-sdr sticks, and many images of the pcb boards. Some of them have plastic case, other have metallic case, but NO ONE of them mount any heatsink on chips, at least at time they leave the factory, so I suppose they were made to endure high temps. Anyway if you feel worried about heat, you can try to remove the case and leave the pcb exposed, that should help to dissipate heat and eventually blow fresh air by a small fan.
Inspired by your post about the heat of flightaware stick, I've successfully tested a cooling fan solution, Soon I will post pics of it.
Getting fun in a boring afternoon, I've spent some time to realize a cooling system for the FlightAware stick. I've just recycled what was in my drawer, it is cheapy and effective. The small fan is a 40x40x10mm 12V model and I've used a spare power supply to power it. A 5V unit would be much better, it could be powered directly by Pi via an usb plug or via the GPIO stripline connector, but I was interested just to proof the concept
Results?...well, I can tell you the stick is pretty fresh, a noticeable difference compared to how it used to be before. Theory suggests that less heat should reduce noise figure...anyway, I didn't see any improvement as for receiving performance, same msg/s rate, same max distance, etc, so I guess the only point is maybe a better expectation as for health and lifetime of the usb stick. I've read that somebody took advantage from cooling at higher frequencies, eg satellites reception, but it is not the case of 1090Mhz.
I know from my colleague that FA ProStick exposed to sun is loosing performance (look at FA website user gcugucio). I will suggest him to use your solution.
So with your experiment and cooling fan have you any idea how much the temperature was dropped by? Geffers
I haven't any thermal probe, I can't measure the exact temperature on chip surface, but I can't feel any heat touching the stick, so I suppose it dropped to a value very close to human body temperature.
I have an array of SDRs running various things (9 NESDR SMArTs and 3 Airspys). 2 of which are dedicated to 1090 and 978. They get hot enough you can't touch them for long without burning your finger... but they require very little effort to cool. I have a row of 4 40mm low-speed fans moving air across the top of all of the sticks (this is in my air-conditioned home), and it's quite sufficient to drop the sticks to ambient temperature.
NESDR SMARTS get hot ... the ones cased in aluminum use the case as a heatsink .. so theoretically hot is good since it's dissipating it You can use heat sink epoxy and glue a big fins to them.
Y Yes and no. Heat sinks are great - but they aren't a cooling solution. They are a heat conduction solution - which will eventually reach thermal saturation. You need some way to move that heat out of the sink - either via radiation into the air or via convection. In short - if your heat sinks are hot as hell, they aren't working properly. They are either undersized or need more air flow.
After reading this threat i want to add the following: I used for years Adsbreceiver image on an old first generation Raspberry Pi 1 (made in China) until the SD card corrupted. I decided to reinstall from scratch and used the full version op Raspbian Stretch and the Github version of Adsbreceiver. To be sure to have the required power for the WiFi and RTL2832 dongles I used the power supply delivered together with a new Raspberry Pi 3+ rated 5,2V, 3A. Installation was time consuming but all worked fine. But after some days of working I found the RTL2832 dongle very hot and I could not hold it in hand. This was not the case with the older setup. So I decided to control the power on the dongle and was surprised to read a voltage of 5.57 Volt and current of 0.26 Amp. I reused the old power supply rated 5V 2,1A and the readings where 4.81 Volt and also 0.26 Amp. Temperature of the dongle was normal again an I could hold it easily in hand. The system worked as expected. I suppose the RTL2832U is a 3,3v chip and in the dongle most of the heat is produced by the power regulator. So be careful if you mount a system on a new Raspberry Pi 3. See measures in pictures and excuse my pour english.
Thanks for your post. Problem is related more to power supply than RaPi 3+. Maybe you can ask supplier to replace power supply beacause of overvoltage (not according to specification).
