light led at specific resistance?

Simply changing the value of the pot is going to do nothing to make it easier to calibrate.  It will work with a 1K or a 100K pot as well.  It will still be just as hard to get it to come on at the precise level you want it to come on.  If I could ever figure out the voltage that you want it fo come on at, we can add resistors above and below the pot so that the pot instead of having a range of 9.6 down to 0, we could make the effective range of the pot say from 2 to 3 volts.  That would make it much easier for you to get it exactly where you want it.   First you said it was hard to set it exactly, then I asked for a voltage.  You gave a range of voltages, then your desired voltage did not fall into that range.  I asked for clarification, now all of a sudden we are changing to 3K pots.  If it needs to be easier to set, I can help you with that when I get a desired voltage number.   If it is too hard to calibrate, I need a number. 

I put those values last night. PLease don't think I'm trying to change the circuit...that's way wrong. I'm simply using the components I have here trying to figure out exactly what info I need to give you man. I am literate with this but not greatly. I've been trying to get you the figures you need but I get diff values each time I try. So with that said...I believe the data I gave last night are the definitive figures I have to work with. It's possible that my mimicking the sender with a pot is also giving me a little hassle. However I have to because the gauge is not in the car. Here's my situation.... If I bridge the gauge input (slosh) wire with 15 ohms across GND the light comes on in the circuit. If I then switch to 16 ohms across GND the light stays off. This gives me the point that I want the light to operate at. The voltage output at the slosh wire when the sender reads 15 ohms. That is around the 1/4-E level. I am also getting weird readings on my DMM. I'm thinking I need to look at that too possibly batteries as it reads higher ohm readings than it should. Like 1 ohm is reading as 16 across all the resistors in the strip and other values are reading higher too. My original value range was apparently wrong and sorry for any confusion. I'm trying to give as accurate information as I can. Thanks

Yeah the voltage drops as fuel burns from the tank. And I believe the resistance actually goes to 0 at empty. I'll try the extra resistors. I was excited when it actually worked. You guys are some smart dudes I tell ya. nice little demo there. Makes it easier to understand what's actually going on. Thanks for all the help.

If you're having problems with the indicator flickering as opposed to your just being unable to easily set the threshold, you need to add a small resistance between the comparator output and the positive input of the comparator. The positive feedback will introduce hysteresis, which is needed when the comparison range is too noisy or touchy. The bottom line on that neat swf demo mentions it. When you do this, what happens is that the fuel level will have to rise to a higher level beyond the normal threshold before the LED turns off, and once it's off, it won't turn back on until it drops to a level lower than the normal threshold. This effect keeps it from flickering when it's sitting right at the setpoint. Unless you have very clean signals, it's always a good idea to have the hysteresis to keep it from toggling back and forth erratically. If your concern is not a matter of it flickering, but just that you can't twiddle the pot to the right setpoint, maybe because the pot is not smooth: determine the approximate range you're setting it to, and then substitute a fixed resistance in that range but lower in value, and then put a new pot in series with it, with a smaller range, such as 500 ohms or so.    Then a single turn of the pot will be much finer in adjustment, and you should have no problems adjusting it. It's simply a matter of picking a fixed value close to where you want to adjust it, and then adding in a low range pot to do the fine tuning.