Everyone's comments are correct and valid within their context. But the original question was optimizing for the casual mobile operator in motion. So here are some additional considerations...
Second, any "vertical" antenna...will have a [null overhead] So, think about it: A GREAT pass of 90 degrees goes overhead your vertical is valueless during the middle of that great pass.
True, but when we put some actual numbers on that, the loss is quite insignificant. Using AO51, only one pass every 5 days goes above 82 degrees, and the total duration above 82 degrees is 24 seconds, out of all in view times. So we agree, the antenna is valueless, but for under 1% of the time. The tradeoff is that it is 3 dB better most of the rest of the time...
The measured elevation profile of a 3/4 wave vertical (a 19.5" whip on the 435 downlink) is shown 80% down the page http://aprs.org/astars.html. The overhead null is only 10 dB down at 82 degrees... BUT the satellite is 10 dB closer at that point, so you still hear it almost the same as when it came above 25 degrees. So you get full sky coverage above 25 degrees > 99% of the time with the vertical.
Ideally - in a car - ... stopping and parking and getting a vertical perpendicular to the orbit of the satellite would give best results. Then maybe a Larsen mag mount... And bend - er, I mean, "re-form" that whip right at the base so that it is about 20" bent. you'll be more successful with the FM birds!
Yes, good idea, but now there is some directionality and so the car needs to remain pointed towards the satellite during the pass. Otherwise there is still the null, it is just in a different part of the sky.
- If the satellite is circular polarization the 19.5 " whip looses 3 dB
of gain.
Yes, but most people have observed a polarization shift during overhead passes. And in that case, the loss of a circularized antenna is much worse than 3 dB... usually a complete fade. SO I agree, a circular antenna gets 3 dB better half the time, and much worse the other half the time. I prefer the simplicity of the whip that does not give up more than 3dB no matter what the polarization is. Mounting a quadrifilar on the roof of the mobile is also problematic.
- The common Quadrifilar Antenna is 1/2 wave 1/2 turn.
The 1 wavelength, 1 turn Quadrifilar Antenna has its highest gain near or at the horizon depending on the length to diameter ratio.
But it is still only a 3 dB gain antenna out there, and so it is not going to hear the satellite down on the horizon anyway because the satellite is 10 dB farther away and usually blocked for the mobile. So having gain on the horizon for a mobile omni antenna is wasted. It either is not enough or it creats additional nulls higher up. Better to move that gain higher up but smooth where the satellite is closer and then have good contacts, than waste gain where it is already insufficient or problematic.
Just different perspectives, but the devil is in the details...
Bob, Wb4APR