I've never experienced carb icing on a cold day but strangely I did experience it on a stinking hot, very humid day in mid July* on the M1.
Running at a bit more than 70mph, I gradually lost all power, declutched and the engine stopped, managed to pull over to the hard shoulder. Checked all the usual things, it hadn't seized, good spark, fuel ok to the carb. Cranked it over, and it still didn't start, left it 10 mins and it restarted and ran OK, set off again and the same happened a few miles up the road, again left it for a while and it restarted ok. Set off again and when it started to slow I backed off the throttle before the power was lost completely and it gradually came back to producing full power. I was able to repeat this process a few times.
Thoughts went to fuel vaporisation but the pipe ran well away from any heat source, or maybe a failing fuel pump or a blockage in the tank.
My passenger had a few hours training as a pilot and recognised what might be happening, so when I reached a quiet stretch I kept the throttle wide open and induced the fault again. When it stopped I quickly whipped off the inlet pipe. The carb throttle plate was solid with ice.
The pipe from the hotbox to the airfilter had dropped off completely so I was only getting 'cool' air. What was quite amusing was that all the manufacturers manuals regardless of make seem to infer the hot feed was only for winter conditions (with such things as pivoting air filters pickups with winter/ summer settings etc). They clearly aren't!
- How I recall it was mid July is that it was the evening after the British Grand Prix. So hot a weekend that I was wearing shorts and my skin was peeling from the sun.