Shuttle fuel gauge problems

Exactly how does one of these work anyway? How do you measure the amount of liquid hydrogen in a bazillion gallon thermos jug? Couldn't they just go out and give it a good hard whack on the side of the tank and see if that fixes it?
Reminds me of a cartoon my biker buddy Butch drew up:
Two bikers, New Biker and Old Biker are talking and OB is introducing the members of the club.
NB, "Who's that ugly bald guy over there?" OB, "That's the guy we call 'CB'." NB, "CB. What, he was into those cheesy radios back in the early 80's?" OB, "Nah, it stands for 'Crispy Bob'. He used to check how much gas was in the tank by standing over the filler with a lit match and lookin down the hole."

wrote:
  Quoted Text Here  
To hear it explained yesterday on NASA TV, the sensors measure if they are "wet" or "dry". Wet meaning there's still enough liquid fuel to run the main engines safely, Dry meaning there's not.
If the high pressure pumps in the mains try to run while dry, that would be a Very Bad Thing ("Uncontained failure of the main engine" in NASA parlance is one possibility).
When the sensor goes from "Wet" to "Dry" that is supposed to start the process of shutting down the main engines before they go dry. There is a reserve built in to the fuel supply and only once (in my memory) have I every heard that a main shut down early (it didn't affect the mission).
There are 4 sensors in the LH2 and and 4 in the LO2 tank. It is acceptable to launch with only 2 but, because of the type of intermittent failure they have been seeing (during the tanking test previously and then during yesterday's launch attempt) it was felt best to scrub and see what's going on, fix it and fly later.
  Quoted Text Here  
Well, here's the Terminal Countdown procedures...
http://www.hal-pc.org/~jsb/countdown.html
while tapping the tank is not specifically listed, at T-04M00S you will notice they do prime the pumps <g>
Steve (remove ATTITUDE to reply via e-mail)

  Quoted Text Here  
Dont know if the technology is the same, but where I work we have sensors in our tanks that only read 'wet' or 'dry'. They work like little tuning forks that extend into the (potential) liquid. They vibrate at ultrasonic frequencies if dry, but if they are submersed in liquid then the vibration is damped significantly... and read 'wet'.

  Quoted Text Here  
This makes sense. I was wondering how you could use a pressure-based sensor and make it work. As long as it was stationary, it would work fine (a container with a liquified gas will remain at a constant pressure until all the liquid has gone into gas [and correcting for temperature]). I heard that when they were testing the sensor, it was reading both "wet" and "dry".