I am building a custom 49 Cadillac. I am using a 2014 CTS Coupe for a donor car. It was equipped with a 3.6 V6 non turbo and a 6 speed automatic. At the moment I am working on the fuel system and have a question.



On the original CTS tank there are two sending units. One is a sending unit only and the other is a pump and sending unit. I can't use the CTS tank because of the shape so I need to install the pump and sending unit into a new tank. I would like to eliminate the second sending unit if I can. In the CTS, there was a raised tunnel in the tank where the drive shaft passed below it. This tunnel acts as a partial dam between the two sides so the other sending unit is also a suction pick up for that side of the tank. The new tank has no restrictions at all

While searching, I've found various styles of assemblies but don't know which one I have. Some systems have two completely separate tanks (Corvette) and the pump will suck one tank dry then switch to the other. The sending units appear to be the big question, though. Some systems appear to average the readings to get a signal for the gauge. Some appear to have one sending unit that measures ohms on the low end, then the second unit picks up from the lower range to higher range. I am not good with electronics and I'm guessing at a lot of this

Ultimately, I would like to end up with only the pump/sending unit assembly without having to go into the PCM. How can I trick the PCM into believing both sending units are there and transmitting the correct fuel level reading to the gauge? Any help would be greatly appreciated.

If the PCM is expecting two sending units and averaging the two, and the sending-only unit is removed, and you don't want to go into the PCM then you'll likely get readings of zero to half as the other unit will be gone therefore returning a value of zero.

If the PCM is expecting two sending units and one is reading the low half and one the upper half, and the sending-only unit is removed, and you don't want to go into the PCM then you'll likely get readings of half to full as the other unit will be gone therefore returning a value of zero.

First step I'd recommend is to confirm the behavior of your combination of sending units on the donor car. Disconnect the sending-only unit and see what you can read with the pump/sending unit.

Take the sending unit you want to remove and measure the resistance of it when the float is all the way up(full) then find a resistor that comes close to that reading and attach it to the sending unit wires. Your pcm will always think that tank is full.,

I never ran into this system.......so how does the fuel gauge get it'* signal? OBVIOUSLY FROM THE PCM..... But what does the PCM do? Average the two, like Cathedral says?

If that happens, then Mike'* idea won't work, because you could be empty but the gauge shows 1/2....

Also, I would think a code would be set if there is a large disparity in one sensor from the other....

I was thinking of this too, but it'* half of the tank'* reading. If the computer is watching both sending units then doing the math, doing this would make the fuel gauge read full when full, and half when it is empty, and that would be bad. I'm thinking that if it sees no second sender it might have code in there that just goes with the pump/sender only. Or maybe it will have flashing lights and bells.

If it does get mad about a large difference between one side and the other, I imagine it would've ended up being a recall for GM because you know that the every one of these that was a rental car would have had this condition after a hard corner with low fuel thanks to the currently-popular "bring it back empty" pre-paid-fuel option. Not to mention the other yahoos that rely on the low-fuel warning for their refill reminders.

From what Cadillac Jack said, there is a tube that sucks from the sender-only side of the tank, so I suspect that it sorts delivery out that way and averages the two sending units. I'm looking forward to Cadillac Jack reporting back after trying stuff.

Me either. Part of why I like it here: exposure to new stuff.

I've read about these (fuel tanks wrapped over the drive shaft etc.) and crawled under a few for other reasons, but never knew for sure how they measured and delivered reliably. I was imagining flop-tube of some sort like in aircraft and some other magic for the measurement.

A thought: Worse comes to worse you could get clever with some epoxy and kitbash just the sender onto the sender-pump assembly. In this hair-brained idea I imagine the sender-only sender would be cut out of the sender-only assembly along with its wires and plug. When it is added to the sender-pump assembly, some of the sender-only float arm would be used to interlock with the sender-pump float arm and all float movements would cause both senders to move at the same time. The plug would have to get epoxied into a new opening in the top of the pump-sender unit. In this solution, your PCM would simply never think that you go around corners based on your fuel-level readings while being satisfied with seeing both senders and doing the math.

Yes it'* a big of tilted-head junkyard engineering, however, it'* already a 300HP 3.6-liter VVT V6 1949 Cadillac so I figure it'* not toooOOooOOooo bad of a last-chance idea.