My double active cooling solution now keeps the board on 35 C (Rx1_W use up to 20 MHz)
without outside active cooling goes 43-48C, without any active cooling even on idle
can reach 55 C
All C readings are with Lime Suite
Just starting LimeSuiteGUI (idle?) with out running anything I see temperature in the high 40’s to mid 50’s after 10-minutes.
While I was running the loop back test procedure I saw temp of 63C.
We provide 6 SMA female connectors with cables, 5 for mounting on the Fan Panel and 1 for mounting on the USB/Power Panel. In anticipated operation, the SMA connector on the USB/Power Panel serves the Ref Clk board connector(s), while the 5 connectors on the Fan Panel serve RF connections. Our provided 4" cables don’t reach the Ref Clock from the Fan Panel, so we may add an additional SMA connector/cable to the next run of Kits to provide more options.
Remember to install some finger stock along the enclosure lids perimeter for best RF shielding, it will help to make an even ground on the lid. When it comes to RF, particularly at the SHF freqs, 4 corner screws aren’t really a good ground for shielding purposes. Also screening across the fan/air flow openings otherwise you aren’t really shielding it from anything except something getting dropped on it.
Another 10 hour run 2Msps. This time instead of reading from an angle I pointed directly from above the chips ~15cm away.
Temps: 55C,55C,51C
And LimeSuiteGUI reports 55C.
The device seems stable in regards to heat. I was curious since I had a usb/midi device that had some serious heat issues due to a poorly written usb firmware/driver.
May not come as a surprise to many of you but some may find it interesting. I read a paper on the RTLSDR dongle and temperature was a big factor in the work they were doing. I can’t find the paper at the moment but it was some radio astronomy work and they required stability. Without cooling the performance was erratic (at best). Once they cooled it off the signal/temp profile was very good (i.e. flat) and stable.
After Cebokli solder the “cooling finger” we abandon digging deeper into the problem.
My idea was that something is going wrong with the inner tracking filter as sensitivity drop suddenly after 10-15min. down for 40-50db. The PLL lock is present but we did not investigate if the LO signal is down too.
Anyhow, $8 device is working more than good for that “investment”