Why is our fuel gauge so horribly nonlinear?

chrispitude · **Posted:** Sat Dec 29, 2007 2:23 pm

Hi folks,

I've owned and driven a variety of vehicles, and I know that no fuel gauge is perfect. 1/2 is never 1/2, 1/4 is never 1/4, and so on. Even so, I think the CRD has about the most inaccurate fuel gauge I've ever seen on a vehicle. On a recent trip, I made the following observations:

The needle didn't even come down to "full" until 78.7 miles.
The needle hit 3/4 at 144.3 (65.6 miles later, but this was just after an overnight cold start and I think the cold fuel changed the level indication).
The needle hit 1/2 at 266.2 (121.9 miles later).
The needle hit 1/4 at 376.3 (110.1 miles later).
The needle hit empty at 432.9 (a mere 56.6 miles later).

The last 1/4 is what really kills me. Boy, once you hit the 1/4 mark, you'd better have an idea on where you want to refuel. You can glance down every minute or two and actually see the needle dropping. This is in sharp contrast to the beginning of the tank, when the needle seems to be epoxied just to the right of full!

I had toyed with the idea of figuring out an op-amp circuit which would normalize the fuel sender signal to provide a more linear representation of fuel level. However, if the EVIC uses the sender level for its range calculations, then attempting to "fix" the fuel level gauge in this way will mess up with EVIC. (I don't have an EVIC, but I think I'd like to add one thanks to the recent writeup on how to do it.)

Share your stories! Has the CRD's bipolar fuel gauge ever surprised you or left you in a pinch?

- Chris

Sir Sam · **Posted:** Sat Dec 29, 2007 2:28 pm

chrispitude wrote:

Hi folks,

I've owned and driven a variety of vehicles, and I know that no fuel gauge is perfect. 1/2 is never 1/2, 1/4 is never 1/4, and so on. Even so, I think the CRD has about the most inaccurate fuel gauge I've ever seen on a vehicle. On a recent trip, I made the following observations:

The needle didn't even come down to "full" until 78.7 miles.
The needle hit 3/4 at 144.3 (65.6 miles later, but this was just after an overnight cold start and I think the cold fuel changed the level indication).
The needle hit 1/2 at 266.2 (121.9 miles later).
The needle hit 1/4 at 376.3 (110.1 miles later).
The needle hit empty at 432.9 (a mere 56.6 miles later).

The last 1/4 is what really kills me. Boy, once you hit the 1/4 mark, you'd better have an idea on where you want to refuel. You can glance down every minute or two and actually see the needle dropping. This is in sharp contrast to the beginning of the tank, when the needle seems to be epoxied just to the right of full!

I had toyed with the idea of figuring out an op-amp circuit which would normalize the fuel sender signal to provide a more linear representation of fuel level. However, if the EVIC uses the sender level for its range calculations, then attempting to "fix" the fuel level gauge in this way will mess up with EVIC. (I don't have an EVIC, but I think I'd like to add one thanks to the recent writeup on how to do it.)

Share your stories! Has the CRD's bipolar fuel gauge ever surprised you or left you in a pinch?

- Chris

either the PCM or the BCM will take the sensor feed from the tank float and convert that into information on the PCI bus, the gauge cluster, and EVIC then use that information, along with other data from the PCI bus, to calculate trip mileage, DTE, and such.

Its a jeep thing though, my cherokees were all the same way.

chrispitude · **Posted:** Sat Dec 29, 2007 2:30 pm

Sir Sam wrote:

either the PCM or the BCM will take the sensor feed from the tank float and convert that into information on the PCI bus, the gauge cluster, and EVIC then use that information, along with other data from the PCI bus, to calculate trip mileage, DTE, and such.

Thanks Sir Sam. Based on that, adding an EVIC will be the best "solution" to get a more reliable range indication. It's funny, it seems like on every tank I always think "hey I'm getting pretty good economy on this tank" until I hit 1/4 tank, then I spend the rest of that tank drafting trucks, cursing Jeep engineers, and looking for diesel fuel signs.

- Chris

Sir Sam · **Posted:** Sat Dec 29, 2007 2:41 pm

chrispitude wrote:

Sir Sam wrote:

either the PCM or the BCM will take the sensor feed from the tank float and convert that into information on the PCI bus, the gauge cluster, and EVIC then use that information, along with other data from the PCI bus, to calculate trip mileage, DTE, and such.

Thanks Sir Sam. Based on that, adding an EVIC will be the best "solution" to get a more reliable range indication. It's funny, it seems like on every tank I always think "hey I'm getting pretty good economy on this tank" until I hit 1/4 tank, then I spend the rest of that tank drafting trucks, cursing Jeep engineers, and looking for diesel fuel signs.

- Chris

It looks like its the BCM that converts the sensor signal, seems like pin 17 is the sensor lead.

glaspak845 · **Posted:** Sat Dec 29, 2007 10:23 pm

chrispitude wrote:

Sir Sam wrote:

either the PCM or the BCM will take the sensor feed from the tank float and convert that into information on the PCI bus, the gauge cluster, and EVIC then use that information, along with other data from the PCI bus, to calculate trip mileage, DTE, and such.

Thanks Sir Sam. Based on that, adding an EVIC will be the best "solution" to get a more reliable range indication. It's funny, it seems like on every tank I always think "hey I'm getting pretty good economy on this tank" until I hit 1/4 tank, then I spend the rest of that tank drafting trucks, cursing Jeep engineers, and looking for diesel fuel signs.

- Chris

Not so fast on that idea, my EVIC reads +/- 2 MPG, usually ~2 MPG to the good. I have the same occurrence on a normal basis and usually the EVIC is no better.

Big A · **Posted:** Sun Dec 30, 2007 3:05 am

Interesting topic...

I ran a test the other day, And waited till the EVIC said i had 8 miles to empty, THen filled up. I squeezed in 16 gallons out of 19.5.

Another interesting thing, Looking at the build sheet, It says "YGEA- 5 Additional Gallons Of Gas" and "NGAS- 19.5 Gallon Fuel Tank" Weird eh?

Uffe · **Posted:** Sun Dec 30, 2007 8:34 am

Don't be scared of 1/4 mark. That is not where it is really stuffed, that is one of the normal spots compared to this.

The place it is really stuffed is at the empty mark. When my evic tells me I got 3km left on my tank I can go at least 130kms before the tank will REALLY empty.

I have measured this so far:

1/2 means there has been ~30 liters consumed.
Empty means there has been ~60 liters consumed. Not when the chime comes and the low fuel warning starts. At that point you have only consumed approximately 55 liters.

The tank will hold 78 liters of fuel + the tube. Possibly 80 liters in total. For safety we stick to about 75 liters to make sure.

So that means when my needle hits Empty I have about 15 liters of fuel left. That will take me at least 125km or 78miles.

I experimented with this and tried to keep it running while the needle was on Empty. Obviously because I got scared of letting it dry out (which is bad for the pump and also for you because it is a witch to prime the system afterwards) I pulled in at a gas station and filled up. The most I ever put into the tank is 65 liters.

tulsa · **Posted:** Mon Dec 31, 2007 11:12 am

I have a 1978 300D and the fuel gauge is very similar to the CRD, both are quite non linear and almost identical.

RTStabler51 · **Posted:** Mon Dec 31, 2007 4:21 pm

I usually wait to the orange light to fill up because I know I have about another 30-50 miles worth of range. I don't worry about it once I learn the vehicles quirks.

Pote · **Posted:** Mon Dec 31, 2007 5:01 pm

while not a CRD, I know when MY KJ's fuel light comes on and the needle is at empty, I still have another 3 full gallons of gas in the tank. Not sure how many fillups over the last year of owning my KJ (alot tho) but EVERYTIME I've filled up right as the light came on, it would take 15.5 gallons of gas to top it off....everytime. I have a 18.5 gallon tank on my '03. Couple with my MPG, that equates to another 65 miles of driving after my fuel light comes on. But I do agree, the fuel gauge is not linear at all. I sometimes go 120 +/- miles on the first 1/4 tank of gas, and then only get to 185 miles by the time I get to half a tank...120 on the first quarter, 60 on the second quarter....either way I get roughly 21 MPG across the board on a whole tank (taking into account the additional 65 miles I could go if I went down to fumes), so one can't rely on the bi-polar fuel gauge.

Timmaah! · **Posted:** Tue Jan 01, 2008 7:26 pm

The float arm and float top out before the tank is full and it takes another 3 gallons to fill the rest of the tank to the neck, so you have to burn off that fuel before the float starts to drop. It takes (roughly) 7 gallons to fill my tank from the 3/4 mark, 11 gallons to fill from the 1/2 mark, and 16 gallons to fill from the 1/4 mark. Variations in gallons to fill the tank seem to come from the slope of the concrete by the pump affecting where full actually occurs. My gauge seems to be fairly linear with respect to fuel quantity below the full mark once the first 3 gallons is burned off.

Piker · **Posted:** Wed Jan 02, 2008 9:54 am

The most I've even filled up was 75 liters.

danoid · **Posted:** Wed Jan 02, 2008 2:14 pm

chrispitude wrote:

It's funny, it seems like on every tank I always think "hey I'm getting pretty good economy on this tank"

- Chris

I believe that is exactly why the gauge is calibrated so. That, along with the removal of the 'instant fuel economy' from the EVIC (after 2002) were probably done to enhance the 'perception' of fuel economy, without actually increasing fuel economy

Uffe · **Posted:** Wed Jan 02, 2008 2:40 pm

danoid wrote:

chrispitude wrote:

It's funny, it seems like on every tank I always think "hey I'm getting pretty good economy on this tank"

- Chris

I believe that is exactly why the gauge is calibrated so. That, along with the removal of the 'instant fuel economy' from the EVIC (after 2002) were probably done to enhance the 'perception' of fuel economy, without actually increasing fuel economy

Nah I got one of those EVICs with instant fuel consumption. It's taken out to prevent people like me from doing highway accelerations and then crashing into a bridge when looking in anticipation of 99liters/100km on the EVIC

Nearly gave a poor guy a shoulder tackle on the highway when I first got the EVIC because I wasn't paying proper attention. Lesson learned, nothing happend, but that was shear luck and good orientation by my fellow driver.

The EVIC does also go a bit on the "good side" of things unless you fill up before you reach the middle point on the fuel gauge. Somehow the EVIC was more precise when I did slow driving and kept the fuel gauge on the positive side of the middle mark, compared with doing faster cruising speeds with the needle going below 1/4 mark.

Reflex · **Joined:** Mon Jan 29, 2007 8:48 pm **Posts:** 567

danoid wrote:

chrispitude wrote:

It's funny, it seems like on every tank I always think "hey I'm getting pretty good economy on this tank"

- Chris

I believe that is exactly why the gauge is calibrated so. That, along with the removal of the 'instant fuel economy' from the EVIC (after 2002) were probably done to enhance the 'perception' of fuel economy, without actually increasing fuel economy

Seems unlikely since the EPA fuel economy is posted right on the window when you buy one...

chrispitude · **Posted:** Wed Jan 02, 2008 3:04 pm

danoid wrote:

I believe that is exactly why the gauge is calibrated so. That, along with the removal of the 'instant fuel economy' from the EVIC (after 2002) were probably done to enhance the 'perception' of fuel economy, without actually increasing fuel economy

You know, I was thinking this exact thing... If the user thinks he's ahead of the game 75% of the time, and annoyed at the rapid consumption only 25% of the time, it's a net psychological win! There's definitely an element of truth to this because many people will remark, "Wow, I've gone 120 miles and it's barely below F!" regardless of the number of gallons it actually took to get there.

- Chris

KJ79 · **Posted:** Fri Jan 04, 2008 1:25 am

I recall reading on a forum, that years ago, in the sixties, Cadillac once decided to provide a perfectly linear fuel gauge. This did not go over well with Joe Public. Customer relations was flooded with complaint calls and Cadillac ended up changing back to the non-linear fuel gauges = less complaints.
Most folks prefer to get fooled I guess.
I understand there are folks who are un-effected by the high fuel prices of today. They always get for $30 worth, duh. 'Nuff said. :roll:

Reflex · **Joined:** Mon Jan 29, 2007 8:48 pm **Posts:** 567

Sounds like an urban legend to me. Short of pressurizing the tank, I don't know of any realistic method of creating such a gauge, especially with how people top off and do other such inconsistant methods of filling up.

chadhargis · **Posted:** Fri Jan 04, 2008 1:03 pm

I've never seen a accurate fuel gauge on any vehicle of any kind. It's not just a Jeep thing.

bentrod · **Posted:** Fri Jan 04, 2008 1:18 pm

Pote wrote:

while not a CRD, I know when MY KJ's fuel light comes on and the needle is at empty, I still have another 3 full gallons of gas in the tank. Not sure how many fillups over the last year of owning my KJ (alot tho) but EVERYTIME I've filled up right as the light came on, it would take 15.5 gallons of gas to top it off....everytime. I have a 18.5 gallon tank on my '03. Couple with my MPG, that equates to another 65 miles of driving after my fuel light comes on. But I do agree, the fuel gauge is not linear at all. I sometimes go 120 +/- miles on the first 1/4 tank of gas, and then only get to 185 miles by the time I get to half a tank...120 on the first quarter, 60 on the second quarter....either way I get roughly 21 MPG across the board on a whole tank (taking into account the additional 65 miles I could go if I went down to fumes), so one can't rely on the bi-polar fuel gauge.

LOL, bi-polar fuel gauge! How true.

And I was thinking is was just me! It's true, the last 1/4 tank goes too fast to zero!
I ran out of diesel on one of my cars years ago. Just once. Trying for never again, not fun.

Why is our fuel gauge so horribly nonlinear?

Who is online