• Welcome to the new NAXJA Forum! If your password does not work, please use "Forgot your password?" link on the log-in page. Please feel free to reach out to [email protected] if we can provide any assistance.

Fuel Level Sender Concept And Modification

John D

NAXJA Member # 788
Location
Rockford, Mi
I haven't found what I am looking for so far. I am trying to use a 1998 XJ fuel level sender with an Autometer 5716 gauge. The output from the Jeep sender doesn't equal the Autometer's range but it seems like there must be a way to modify it so it is; I haven't been able to think it through.

Jeep sender, from FSM:
With float in up position, resistance
should be 20 ohms (+/- 5%). With float in down
position, resistance should be 270 ohms (+/- 5%)

Autometer Phantom 5716
240 ohms down
33 ohms up

If I had to choose I would rather have empty be the accurate reading of course, but do you see anything that can be done to make it closer?
 
Looks like you would need a 210 Ohms of resistor in parallel to get to the 240 Ohm down indication. Unfortuntely, it would then read 115 Ohms in the up position...

I do not see a good solution to using the stock sender. Anyone else have an idea short of using a different sender?
 
I don't really see a problem there...changing nothing, it'll indicate 'empty' early, and stay on 'full' a little longer.

I'd probably run it. It's like having a reserve tank.
 
Looks like you would need a 210 Ohms of resistor in parallel to get to the 240 Ohm down indication. Unfortuntely, it would then read 115 Ohms in the up position...

Can you show your math/equation. The resiter numbers you posted don't seem right; I'm not an expert though.

I don't really see a problem there...changing nothing, it'll indicate 'empty' early, and stay on 'full' a little longer.

I will hook it up again to check, but I think it went over full and didn't go to empty. I will be dropping the tank again this week.
 
I will hook it up again to check, but I think it went over full and didn't go to empty. I will be dropping the tank again this week.

Well, from what you posted, the numbers for the XJ sender exceed the range on both ends of the gauge limits, so I'd think it should go from pin to pin if that's the case...

And yeah, that resistor math isn't right. Putting two resistors in parallel yields a net resistance less than either of the two by themselves...
 
i thought every cherokee came with this option from the factory. i'll drive to and from work twice ~40 miles with the guage on E, then fill up and i'll only be able to stuff 18.5 gallons in.

I can tell you that my stock 23.5 gallon MJ tank had nearly 6 gallons left in it when the fuel light came on. I don't think AMC was too hip when they designed the senders.
 
The problem with trying to compensate by adding resistors in series or parallel is that the resulting variable resistor is not linear. I've looked into this as my MJ has a RENIX sender and an OBD-I gauge right now (since the only rust free sender I could get was a RENIX one, via gambit4000s :worship:.)
 
Well, from what you posted, the numbers for the XJ sender exceed the range on both ends of the gauge limits, so I'd think it should go from pin to pin if that's the case...

I thought so too. I will verify in a couple days but from what I remember it wasn't the case.

The problem with trying to compensate by adding resistors in series or parallel is that the resulting variable resistor is not linear. I've looked into this as my MJ has a RENIX sender and an OBD-I gauge right now (since the only rust free sender I could get was a RENIX one, via gambit4000s :worship:.)

Come on man, you always have the answers for this kind of stuff. I was counting on you...
 
I have an answer for it, but it involves a lot of custom electronics I haven't built yet... still have two jeeps that need this myself! The 96 has a 97+ tank and sender and thus reads all screwy, and the MJ has the setup I already described.

If you could tap into the sender resistance strip and add small resistors in series (to increase overall resistance) or large resistors in parallel (to decrease overall resistance) this would be possible but I can't think of a way to do that without causing real reliability problems. My intent is to build a converter that reads the sender and adjusts a digital potentiometer to the value the ECU/gauge cluster is expecting, then I can build a pair of them and tune one to convert my RENIX sender to an OBD-I sender and the other to convert the 97+ sender to a 96/bastard sender and forget about it, but I haven't determined how much power is dissipated by the stock sender so I'm unsure if the digital potentiometer I have in mind will be able to handle it. It also isn't something most people will want to try doing, it'll be easier to just get a sender that sends the right resistance range.
 
I have an answer for it, but it involves a lot of custom electronics I haven't built yet... still have two jeeps that need this myself! The 96 has a 97+ tank and sender and thus reads all screwy, and the MJ has the setup I already described.

If you could tap into the sender resistance strip and add small resistors in series (to increase overall resistance) or large resistors in parallel (to decrease overall resistance) this would be possible but I can't think of a way to do that without causing real reliability problems. My intent is to build a converter that reads the sender and adjusts a digital potentiometer to the value the ECU/gauge cluster is expecting, then I can build a pair of them and tune one to convert my RENIX sender to an OBD-I sender and the other to convert the 97+ sender to a 96/bastard sender and forget about it, but I haven't determined how much power is dissipated by the stock sender so I'm unsure if the digital potentiometer I have in mind will be able to handle it. It also isn't something most people will want to try doing, it'll be easier to just get a sender that sends the right resistance range.

The renix and the OBD-I gauges use the same resistance values, the OBD one is just reversed correct? What stops you from simply reversing the resistance at the gauge? Seems like a simple inverse circuit would fix that one.

Of course I may be wrong about the chryco gauge, but I htought they were both 0-90, just the chryco gauge was backwards.
 
The renix and the OBD-I gauges use the same resistance values, the OBD one is just reversed correct? What stops you from simply reversing the resistance at the gauge? Seems like a simple inverse circuit would fix that one.

Of course I may be wrong about the chryco gauge, but I htought they were both 0-90, just the chryco gauge was backwards.
This is what I thought too, but it is NOWHERE NEAR linear :roflmao: one expects a curve one direction, the other compensates for a curve in the other direction. So:
Reading / Actual
Empty / Full
1/4 / 5 gallons
1/2 / 4 gallons
3/4 / 3 gallons
Full / very, very empty. Get ready to use the jerry can or start walking

It literally reads 100% full (well, empty) till about 3/4 of the tank is gone, then plummets to empty (well, full) over the course of the next 50-70 miles.

I am not sure if this was the intended behavior / design, my sensor and gauge may be very inaccurate due to age, but that's what mine does.

From what another thread indicated, it is possible to retrofit a RENIX gauge into an OBD-I cluster and vice versa, but hacking that stuff up isn't something I'm really interested in doing. I'd rather build my own custom thing that goes in between that I can just build another of whenever I need it, and even use to calibrate my gauge in the 98 so it reads linearly instead of being a bit nonlinear like most gauges.

Wish I had a simple "toss some resistors in here and it will work" answer for you but I am afraid none exists, at least I'm not smart enough to figure it out.
 
most of the not being linear has to do with the geometry of the linkage rather than any electrical reason. The pot in the sender is a nice semi cricle that's all perfect. The arm on the float is not, and since the gas tank is also not perfectly geometric and consistent up it's sides that's what you get. This is a big reason why most cars seem to run forever on the first half of the tank, then the bottom falls out on the lower half.

I've seen european cars that have the gauge calibrated to adjst for that, with much smaller spacing on the upper half for full, 3/4, 1/2, and then much wider spacing for the bottom half, 1/4, E.

Personally if it were mine, I'd just make a new gauge face and glue it over the old one after I found some known levels.

I'm glad my renix junk just used a standard 0-90 GM sender, because my RCI fuel cell plugged right up and the gauge and my low fuel light work perfectly, more accurate than the stock one even.
 
good luck with it. Do they sell the mounting flange separately to be welded to the tank?

Others do, I found one from VDO with a mounting system; but not to be welded. Seems kinda hokey for 30 bucks though. Most just have screws and a gasket. You think it would be an issue with leaking? I would use bolts with nuts and a gasket plus RTV. I bet it wouldn't leak.
 
I'd probably just use a big sheet of nitrile rubber (buna-N) to make a custom gasket and call it done. You can get 1/8" nitrile sheeting in 1 square foot pieces from smallparts.com, btw, that's where I got mine for another gas tank gasket project.

Before using *any* rubber or RTV on a gas tank, check http://www.coleparmer.com/techinfo/ChemComp.asp for compatibility with gasoline. I seem to recall red RTV resisting it well but I'm not sure.
 
Before using *any* rubber or RTV on a gas tank, check http://www.coleparmer.com/techinfo/ChemComp.asp for compatibility with gasoline. I seem to recall red RTV resisting it well but I'm not sure.

Short of specialty stuff, the only gasoline-compatible one I've found is Permatex Form-A-Gasket #2:

http://www.permatex.com/products/au.../auto_Permatex_Form-A-Gasket_No_2_Sealant.htm

It doesn't harden and shouldn't be used as an adhesive, only a sealant. I used it around the sender/pump ring on my fuel cell.
 
I'm busy with supper and family but it seems like extending the arm the float is mounted on might lessen the angle and would more closely match the range.
 
Back
Top