I purchased an used LimeSDR from Ben NO5K and he had added a nice improvement by adding a small exhaust fan as seen below. My initial tests show the fan is doing a great job. I want to share my initial impression and how this can be easily replicated at low cost:
The fan and heat sink seem to make LimeSDR run cool, reading 33.7C after 10 minutes running the self test:
Here is the fan mod done by Ben NO5K:
Here is a low-cost candidate fan:
The card get warm even just plug in i have put tiny heat sink on the main chip and going to put more when i have done all the necessary wiring and put all in my custom case.
The enclosure ($5) and panels ($15) are sold on eBay:
http://www.ebay.com/itm/112088499793
http://www.ebay.com/itm/371873005786
@robalenper A question, is the fan blowing or… eh… sucking? (no pun intended)
I understand entering cold air is the way to go, but it might not the case.
BTW I hate making square holes 
Looks like goes in side then blow it on the chip then out side those connector.
@Ismas The fan blows out of the enclosure. I can barely notice the air flow since this is a low-rpm, slow-flow fan.
The cold air enters from the USB3 panel as highlighted below. The entire enclosure warms up evenly and stay in warm temperature. The Altera FPGA has a heat sink and no heat sink is recommended to the Limemicro transceiver chip to avoid coupling and noise floor increase.
Recommended? Can you please provide the source of that?
BRG
Ben NO5K added a small heat sink to the transceiver chip and observed an increase in the Noise Floor at 2456 MHz.
For me, this was enough evidence to leave the transceiver alone.
Maybe a non-metallic (ceramic) heat sink would work better on the transceiver chip, if it gets warm at all.
Tnx…
Transceiver chip goes on 65 C on 20 MHz bandwith RX, no TX.
I have on mine heat sink too but used 0,5 mm thermal pad.in between
thermal conductive 6W/mK and electric non conductive.
No any change in noise level at last on HF and up to ADS B 1090 MHz
So must check on 2 GHz 
too…
A heat sink then would help at 65C on the transceiver chip.
I’m wondering if a ceramic heat sink would be a better option to avoid any coupling?
for sure…
I might be being a bit thick, here (it wouldn’t be the first time), but ‘ceramics’ generally (in my experience) don’t tend to be the best thermal conductors - how effective are these ceramic heatsinks, given that they’ve not got that much ‘finning’ either?
Has anyone got a link to thermal resistance figures for these types if heatsink, because I’m curious, now
Cu thermal conduction is 401 W/mK electrical resistivity is 16.78 nΩ·m
Al thermal conduction is 237 W/mK electrical resistivity is 28.2 nΩ·m (at 20 °C)
AlN thermal conduction is ~200W/mK electrical resistivity is ~10^14 Ω·m
Al2O3 thermal conduction is ~24 W/mK electrical resistivity is ~10^17 Ω·m
Ok, I’ll buy those figures for bulk thermal conductivity ;). In mitigation, I think I’d been confused by the ‘ceramic’ term. The materials you described are what I’d call alumina - and that has much poorer conduction - the AlN on the other hand …
So, it’s clear that AlN is well capable of shifting heat from one place to another, but the heatsink in the eBay link has very little in the way of ‘finning’ - ie not much more surface area at the ‘cool’ end than at the ‘hot’ end… So what’s the overall thermal resistance between the surface of the ic package and the air, compared to an aluminium or copper finned heatsink ?
Footnote: I’m only looking at the link on my phone so I might not be able to see all the info, but as far as I can tell, the eBay listing doesn’t specify what the material is (apart from ‘ceramic’, and doesn’t provide a K/W thermal resistance figure.
The Ceramic Heat Sink is probably going to be Al2O3, just because it is much much cheaper to make, and if you are looking on ebay it is usually the cheapest, but even Al2O3 can perform well when compared to similar Al heatsinks.
e.g. http://www.ohmite.com/cat/sink_wc.pdf
Not sure about the figures but passive air versus active air is the key with most any material for efficient cooling. There’s debate on blowing air onto the heat sink versus drawing air away from the heat sink. I read that drawing air in and blowing it away is most efficient.
There are heat pipes too.
