Hi - I want to place the kraken SDR I have on a drone for a summer project I want to run with some students. I want to remove the heatsink to save some weight.
Does anyone have any experience with this? Is the SDR likely to work after I do this? Is there anything I should know before I attempt this?
It will still work, but the board will certainly get very hot without any sort of airflow over it. I would still suggest adding a small heatsink to the underside of each of the tuner chips, and power LDO’s and try to incorporate some airflow from the rotors or drone movement.
However, you might expect to see poorer DoA results.
One of the reasons for the heavy heatsink is to add a big thermal mass. This helps stabilize the temperatures between tuners. Rapidly fluctuating temperatures can cause the tuners to phase drift which results in poorer DoA results.
It’s left over from another summer project, it’s fairly large but I’ll probably have to stick to higher frequencies and accept lower DOA accuracy. I think I’ll use a HackRF to transmit something in the 915 MHz ISM band. For the purposes of the project, the drone is good enough I think.
Thank you! I have some left over aluminum heatsinks from a raspberry pi project a few months ago.
I’ll try with that and do a before/after test. I’ll report my results here so if anyone else wants to remove the heatsink, the performance differential is recorded before and after.
The chips in between 9 and 10. It might look a little different on newer boards, but they’re in the same position. You can heat sink them on the bottom of the board.
I got it working for around 15 minutes with a mini desk fan and some mismatched aluminum heatsinks. I reused the thermal pads that were already installed. I’m sure I can make it more stable with a more professional solution.
The fan was necessary, without it the board disconnected/shut down a few times (I’m assuming due to overheating) and had to cool down then be restarted. I’ll have my students make sure the board has good airflow during operations. I may also tell them to cool the USB hub, since my temperature sensor saw it getting a bit hot.
I noticed no change in DOA accuracy against the 915 MHz LoRa signal I was using to test, but I don’t know how stable that will be on the drone or in hot weather. I guess my students will have to solve those problems .
I did struggle a little with reassembly, I accidentally twisted off 4 of the smaller screws that hold the board to the smaller plate. If anyone is trying this themselves, I recommend being very careful with them. In my case, I guess this doesn’t really matter since my students will have to come up with their own cooling solution anyway.