| There are 2 limiting factors taken into account during design - regulations and peak current that the battery can supply. The battery thing (usually a coin cell CR2032 that can supply ~13mA per spec) can be a bit mitigated with a beefy capacitor (though RF packet length and space in-between have to be appropriate for the capacitor recovery, not to brown out the system). Usually the bigger bottleneck is the regulation, so you tune for best performance in a clean environment. Hand and body during operation are really unpredictable to model in/retune, but luckily usually adds for the better.
Anyways the range of the classical RF fobs (@3xx-4xx MHz and 8xx-9xx MHZ bands) is well beyond the real world needs that the 10-20% drop of performance due to certain fob positions will make a difference. If it does, you probably have a low battery. Next gen fobs (if they hit the mass market) are BLE and there the range performance is more critical. Dunno what will be the approach there. What can cause problems are ornament and the mechanical key insert. It has to be extensively tested that there is no coupling with those metal parts that are close to the antenna. Problem comes sometimes when they decide to redesign a bit the keyfob without going through the tests internally (e.g. oh it's just a minor tweak of the radius of the metal ring around the fob housing...) and then the certification fails and has to be repeated (not just money, but getting the slot for it...).
Some OEM might even be inclined to think they don't need a certification for such a minor change and push it to the market, making the problem worse if there is an issue. Regarding the ornament things on fobs, there were even instances of them magnetically coupling with the wireless chargers or even immobilizers and that lead to some melting... |