E30 Fuel Gauge and Sending Units Bonanza of Doom

[E30Bastard]

I have reviewed all the electrical diagrams and verified the correct operation of both sending units. I have also verified the wiring between the sending units themselves and the wiring from both sending units to where the wires terminate in the harness connecting to the fuel gauge. I have also tested ohmage and resistance for all of the above.

I have even gone to the point of bridging the 64ohm fuel gauge resistor inside of the cluster with a 47ohm resistor to halve it. Then I tried a straight bridge to bring the resistance down to O.

The only thing that gets closest to representing what is actually in tank is when I disconnect driver side sending unit and short it, reverting to using only the passenger side sending unit which is not accurate or going to provide me desirable fuel gauge behavior. If I short the passenger side I get the same reading on the gauge as when I short the driver side.

Also, in the wiring diagrams I did notice that there is a T for Pin Number 4 (main passenger sending unit) over by the OBC where the OBC just reads the resistance to calculate fuel data. It does not interfere with anything and I'm not seeing any resistance with everything, with the OBC unplugged, or any combination thereof. I disconnected the OBC just to see if it made any difference. There was no difference.

The code for the cluster matches up to the year and make of my car. I even forked out 400 dollars for a new driver side sending unit and 200 for a spare/tester cluster because Im literally losing my shit trying to figure this out. I had a brand new passenger side sending unit in storage for 15 years or I would have bought one of those too. Both clusters are both exhibiting the exact same behavior no matter how I alter the resistance.

I'm about to run all new wiring which will bypass the OBC. Soldering all connections at the sending units and cutting the old wires from the back of the harness and crimping into the pigtails. I can think of no other option even though I've tested and verified the wires and sending units in the car.

This should be such an easy fix! There's really nothing to this system. I've got a long drive ahead of me tomorrow maybe it'll hit me on the road after some rest. Before I lose it completely I am going to bed. It's 2AM. Good night and thanks in advance for any advice that doesn't include setting this car on fire because that's what I want to do right about now! lol

[E30Bastard]

Man, I rolled out of bed hoping for the best...haha

So upon further thought tossing and turning before I passed out I forgot to mention that the red warning light functions perfectly so the float on the passenger side is traveling the entire length of the body and shorting to ground.

...I have two more ideas for testing as I'm typing this so I'm going to go scramble some eggs and see what I get before I hit the road to Sacramento to pick up some popout windows with unbroken hinges.

[roguetoaster]

Sometimes we have to accept that it's JFM, or ask jlevie.

[TobyB]

So when you hook up a 120 ohm variable resistor in place of the 2 sensors, can you
sweep it and get the gauge to indicate 'empty' through 'full'?

t

[Jean]

Hey Dimi, was nice to meet you. As we discussed I would wire a variable potentiometer (300ohm would work and Frys should have them ) I’m ensure that you are able to sweep the fuel gauge from empty to full.

You can do that test with cluster in the car , otherwise take cluster out and bench Test it . Supply it power and then wire resistor of known values to simulate fuel senders.

[E30Bastard]

Hi Jean and Toby,

Thanks so much for the responses. Sorry have been MIA past couple days pretty much bedridden with a terrible cold. I'll be headed to Fry's to get the suggested gear for testing as soon as my last conference call of the day is over. Will report back tonight or tomorrow!


Thanks,


Dimi

[E30Bastard]

Okay guys!

I had to drive to three different Fry's two last night and one today but I finally found a 250 ohm potentiometer and I ran a series of tests. But first I went back through and re-tested all the wiring again. Sorry if some of the information or tests seem redundant I just wanted to be thorough and go through all the steps of the complete circuit from gauge through sensors and back to cluster. Here's what I got:

I estimate 3/4 to full tank in it now.

Pretest notes: all wires show zero resistance from end to end cluster to driver side, driver to passenger, and passenger back to cluster. Both sensors tested 0-60+ohms outside of the car. I don't have exact numbers but they were within spec.

Installed sensor resistance:
Driver side resistance shows 21.6 ohms, passenger is 25.6 ohms.

Pin 5 cluster brown/grey wire to passenger side shows solid 11.9 volts.

Pin 4 cluster brown/green wire to driver side varies between 8 and 10 from available test clusters

Cluster 1 - came with car. blue backing all motometer.
Service lights: 5 green lights, 1 yellow, 1 red - 4 green lights light up. Provides 8 volts to brown/green wire pin 4

Cluster 2 - ebay. white backing labeled vdo but mix of vdo and motometer parts.
Service lights: 5 green lights, 1 yellow, 1 red. - 1 green light only.
Provides 10 volts to brown/green wire pin 4.

Test 1 - Just Driver Side Wiring (I used brown/green wire coming from pin 4 cluster to driver sensor for potentiometer and alligator clipped a ground):

Cluster 1: output 8v @ driver side. Twitchy gauge. potentiometer only takes a quarter turn to sweep entire gauge.
Cluster 2: output 10v @ driver side
Very smooth gauge. Potentiometer only takes a quarter turn to sweep entire gauge

Test 2 – Pass-through driver sensor to passenger harness:

Cluster 1: confirmed 8 volts on passenger side brown/green wire.
Same result as test 1.
Cluster 2: confirmed 10 volts on passenger side brown/green. Same result as test 1.

Test 3 – No sensors, using both harnesses as per diagram – Pass-through wiring from driver to passenger. Potentiometer connected to passenger harness via brown/green from driver, harness ground, and harness brown/white which goes back to pin 5 cluster:

Cluster 1: Successful sweep through gauge all the way until red warning light.
Cluster 2: Successful sweep through gauge all the way until red warning light

Test 4 - Plugged in the driver side sensor and used only the passenger side harness with potentiometer:

Cluster 1: shows half tank max and sweeps down to empty until warning light on.
Cluster 2: shows just above half tank max and sweeps down to empty until warning light on.

Test 5 - Both sensors plugged in:

Cluster 1: 1/4 tank
Cluster 2: 1/4 tank


The glaring variable is the different voltages coming from the two different clusters concerns me. Is it supposed to be 12v from pin 4 at the cluster to the driver side?

Also, I know we touched on this during our conversation, Jean, but what are the defining marks for SI and backing color (blue vs. white) in regards to whether the clusters I have are for early model or late models. The fuel gauge resistors in the clusters themselves are 68ohm which matches the 1988 electrical diagrams I have found online at armchair and in my Bentley.


*edit - I swapped the fuel gauges just to see if the twitchy/smooth behavior was because of the cluster or the gauge itself. Ended up being the gauges themselves as the smooth gauge from cluster 2 is now smooth in cluster 1 and the twitchy gauge from cluster 1 is now twitchy in cluster 2. They still both read exactly the same. I guess it is time to desolder the 68 ohm resistor from the frankenstein cluster with higher voltage, test to see if it was good or bad, and then solder in a new one (I bought a pack of 1w 68ohm 2% resistors when I was at Fry's for the potentiometer).

**edit part deux - Old resistor from cluster tested good at 69 ohms. Not going to solder another resistor back in yet as I'll just get the same behavior.

[2mAn]

bump :)

please help this guy before he lights the car on fire

I think we figured it out... both fuel gauges are from an early model. Lets see

[E30Bastard]

Popeye, My Friend, I have a proposition for you...

I'll gladly give you TWO early model fuel gauges tomorrow for one late model fuel gauge today. ;D


Look what I just found on the web from 2012 by jlevie:

https://www.bimmerforums.com/forum/showthread.php?1174914-Late-model-fuel-tank-into-early-325is

[E30Bastard]

Dude...

[roguetoaster]

Can we just download Jim's brain in to some sort of epic FAQ thread already?

[2mAn]

Can we just download Jim's brain in to some sort of epic FAQ thread already?

I wish

[AndrewBird]

Is it possible the float is not floating all the way to the surface of the fuel? Check the resistance with it installed and again out of the tank to see where it sits.

[TobyB]

Potentiometer only takes a quarter turn to sweep entire gauge

Oh, you beat me to it- you have 2 early gauges. And a late car. Early gauge- 60 to zero. Late gauge- 120 to zero via 2 60 ohm senders in parallel.

Because sucking jet valve.

You could hook gauge one up to each sender and have a meter for each half of the tank.

Jim helped me with this one years back.

t
late to the par-tay.

[Jean]

Haha nice! Sorry I should have asked you if this was an original cluster and/or tank and didn't realized you swapped it :) I had to do exact opposite, I am using a LATE model cluster and I swapped my early gauge years ago.


But to answer your question, late cluster is 5 green 1 red and 1 yellow. Early cluster has 5 green, 1 yellow and 3 red lights.