1 pip

More than likely there is a lot of gasoline left in your tank if you are typically filling with only 6-6.5 gallons at 2 pips. (If not, something is wrong with the gauge or tank and it would be good to know that so you can get it fixed under warranty.) Dropping another pip might give you 3/4 gallon or more. I've finally managed to put in a whopping (<sarcasm alert>) 9 gallons by waiting for the flashing pip. I did have some ~6-1/2 gallon fills at 2 pips and quickly decided that was ridiculous with a 12 gallon tank. When I bought the car I anticipated ~10 gallon range without pushing it, instead 9 gallons is sending me to the end of the display.

A simple analog display/gauge or something with several times more incremental resolution would be a big improvement.

 

Wait for the weather to warm up, it will get better.
After you fill to the first click gently rock your Prius from side to side for 20 to 30 seconds and fill to the next click, it helps the air to get out (burp it) but you shouldn't be over filling. Also the fuel in the ground is most likely warm and will soften the bladder a little if you let it allowing it to hold a little more.

Of course I was getting about 500km from a 60 litre tank with my 2 litre Camry before I started hypermiling (even though I didn't know what it was called) so your 514km from a tank isn't bad really is it?

 

*Burp*

 

I routinely go 10-15 miles past the point where the final pip starts to flash and the car beeps and pops the "Add Fuel" message on the display. I've gone further before refueling, but that's where I start to get nervous.

 

No it would not. Resolution is not the problem; accuracy is. People make this mistake all the time, so don't think I'm going off on you personally. I am attempting to correct this misconception, but it feels like I am playing "Whack-a-mole".

Here is the point: Let's say you replace the 10 pip digital display with a great big analog display with 100 divisions. What do you have? You have a lot more divisions, but no more accuracy. The problem is the data coming from the sender unit in the tank. It has a high component of uncertainty. Like most automobile fuel senders, the Prius uses a float sender, which is hardly a precision instrument. On top of that we have the uncertainty of the fuel tank bladder, and the calibration of the inclinometers. It is not a precise system.

Having more resolution on the display is like putting a second hand on a sun dial.

Tom

 

I disagree because as long as there is any fuel in the tank, precision wins, hands down. Also, from what I've seen of the tank it is consistent over the course of using the tank itself--this may be some trickery in the display programming, nevertheless, that is what is indicated. Until the display resolution exceeds the actual gauge precision, extra resolution is useful. Rather than a "second hand" what is missing is any sort of half hour demarcation...as well as several of the most important hours.

My other vehicles have shown hysteresis in the gauge reading (perhaps intentional to prevent incline/cornering effects.) There is a sort of "averaging" that appears to occur. It is most noticeable at start up/shutdown. With my truck the indication is a little higher when I shut down than when I start up. The level indication rises slightly over the first several minutes of operation. When fuel is low in the tank this is quite noticeable, but I can see it throughout the range of the tank.

I'm not even convinced about the uncertainty of the Prius sender. The uncertainty I've experienced is in the fill portion above what the sender displays. This appears to be the result of the bladder vapor/recovery system. In fact, if I could see the output of the sender it would not surprise me to find that its output correlated with the actual fill volume. If I had a traditional display of the senders output (adequate resolution) then I could ascertain whether the hypothesis was correct or not.

Evan summarized the difference well. He can see the actual output. Because that is still changing he can tell there is still level. If that was flat lining or rising, then one would be concerned that the vehicle was either out of gas or the gauge wasn't registering.

Evan, what is the resolution of the output? Is it limited by your reader or by the signal to it? If you get a "good" fill (low calculated mileage compared to the MFD) can you see the diffrence from a normal/average fill? Or does it max out on any decent fill?

 

Very useful. That suggests that a good fill is probably above the top of the sender. In that case it is the sender's span/range, rather than accuracy that is preventing a higher reading.

What percentage does the blinking pip correspond to? I've wondered if it is an independent indication (common for redundancy in process vessels) or if it is a low level indicator at a given sender signal.

The other thing I've wondered about the output is the range of the displayed pips. Are they 0-10, 10-20, etc. or perhaps 5-15, 15-25, or perhaps a non-linear variant?

 

Shawn, I agree with your post. My only quibble is with the precision of conventional fuel gauge displays. I've never owned or driven a car or airplane where the analog fuel gauge had any more useful precision than the digital gauge on the Prius. Because of that we agree on your comment stating: " Until the display resolution exceeds the actual gauge precision, extra resolution is useful". I haven't found one yet that does any better. You could design one for sure, but it doesn't seem to be a priority.

The fact that Evan can read the fuel level in smaller increments does not prove that the sender has more accuracy than the the precision of the display. We would need a lot more testing on several units before I could make that claim.

Repeatability is more important than accuracy in a fuel gauge. Do you find that the repeatability of your Prius gauge is questionable?

Tom

 

I can't speak to the piloting side of it. I do recognize that running out of fuel in an aircraft is not a peril one wants to risk.

However, I can say that other than actual defective senders I've had little problem with their precision or accuracy as seen on the gauge of most cars. And when they were defective, both precision and accuracy were affected simultanreouly in a major way. The only defective ones I've experienced were domestic (and not my personal vehicle.) None of my Japanese vehicles have had a problem with them (knocks on sheet metal--or is that resin?)

I did note the hysteresis of course. It introduces a bias, but one that I have no trouble mentally filtering out since it has been repeatable. When I fill up my truck I typically predict the fill within a gallon by eyeballing (not getting out the calculator.) That's on a tank that typically takes 22-24 gallons, and once 25.6. It's a 26.4 gallon tank IIRC, so the latter was well outside my normal margin for error. The angle of the truck during fill is the strongest variable I've noticed--but again it is predictable and not a reflection of the sender.

No...and yes. The resolution is so chunky that I can't really tell what the Prius level is doing except when it changes pip indication. So the only thing I can tell is how wildly variable the fills are, but that's in the final pip/off scale. That's more of a bladder/vapor system issue.

And I don't doubt others who have gross sender/display error either. The liquid/vapor balance & vent system is such that seal legs and other problems can be added to the mix. If someone is having serious filling trouble, then there is probably a 50/50 chance that the level indication will be erratic or grossly off as well.

EDIT: Meant to add my experience in a rented Mexican VW bug. It had a riding lawnmower like gas tank and gauge. I had a scare with it appearing to run low on gas very quickly...turns out it was the luggage pressing on the sender. Moved the luggage to the backseat and the level returned to where I anticipated it being.

 

My family owned one of the original VW bugs. You would have no problem with the fuel gauge in those, as they didn't have one. All they had was a reserve lever that you flipped when the engine started to sputter. On the plus side you knew exactly how much fuel was left when it started to sputter, unless you forgot that you had already flipped the lever.

My 1994 Ford van has a digital fuel gauge showing pips. The difference with the van is that it also has a digital readout of remaining gallons down to one-tenth of a gallon and a reasonably accurate "miles to empty" figure based on current mpg. When the "add fuel" light comes on, you know pretty much exactly how far you can go before running out of fuel.

It's a little trickier with the Prius, since the mileage can swing a lot, but allowing for that, there is no reason that the Prius can't do as well with the fuel gauge as a 1994 Ford. I think this is an area that slipped through the cracks.

All modern fuel gauges have some sort of averaging, otherwise they would swing all over the place around corners and when braking or accelerating. The Prius is no different in this regard. That shouldn't be an excuse for an inaccurate fuel gauge.

Tom

 

I think since Toyota has gotten rid of the bladder system for the 2010, they ought to do a voluntary recall for owners of Gen 2 Prius' who wish to have their fuel storage system converted to the same system the 2010 is using. Does anyone know how to submit a potential recall to the US Govt Agency that oversees this?