Fuel Gauge driving me crazy

[mpierce9]

ok, several months ago I had started this thread regarding my fuel sending unit.

Original Fuel Sending Unit conversation

Tunrs out the sending unit works great. The gauge also works great. The brown wire between the sending unit and the fusebox also reads and works great. The problem is that everything only works great with the car powered completely off. I checked with a multi meter. i confirmed the multi meter batteries were brand new.

I've checked the gauge directly to the battery and sending unit and it reads PERFECT. Unless I turn the key to acc, start, or the truck is running. Then everything reads at 100% resistance on a multi meter and 3:00pm on the gauge. This is the part that confuses me. The gauge is laying next to the battery on a piece of wood and is only connected to three different wires at this point: directly to ground on the battery, directly to hot on the battery, and a splice to the sending unit. in this scenario it still goes to 100% resistance when the power is turned on.


I'm not sure where to begin.

• I read 64 OHM resistance on the tank currently (about 3/4 a tank) with the power completely off.
• I read it everywhere on the brown sending unit line, on the gauge and everywhere along the line, with the power completely off.
• I can trick the gauge into any reading I want using resistors.
• I can read the exact fuel level if I power the car off and hook the gauge to the power and ground on the battery and the fuel reading post directly to the sending unit wire

What would cause the entire system to read 100% resistance to ground when the power is turned on to the car but not when it isn't?

[imjeff]

I didn't know you could read resistance in a powered circuit. Am I missing something? When there is power in the circuit doesn't that create resistance?

[mpierce9]

I believe in this case it is the resistance to ground.
Shamelessly stolen from this thread

http://67-72chevytrucks.com/vboard/s...d.php?t=555432

http://img131.imagevenue.com/img.php...122_513lo.JPG#

[ray_mcavoy]

Originally Posted by imjeff

I didn't know you could read resistance in a powered circuit. Am I missing something? When there is power in the circuit doesn't that create resistance?

You're right, you can't get an accurate resistance measurement in a powered circuit. Power flowing through the circuit doesn't create resistance though ... instead, it results in a voltage drop across whatever resistance is there.

When measuring resistance, the multimeter's internal battery is used to supply a small amount of current through the circuit/component being tested. Adding in current from an external source will result in a false reading and can potentially damage the multimeter.

Originally Posted by mpierce9

I believe in this case it is the resistance to ground.

Yes, the sending unit is on the ground leg of the circuit. But that still doesn't mean you can accurately measure it's resistance when the circuit is powered up.

Originally Posted by mpierce9

I'm not sure where to begin.

• I read 64 OHM resistance on the tank currently (about 3/4 a tank) with the power completely off.
• I read it everywhere on the brown sending unit line, on the gauge and everywhere along the line, with the power completely off.

Okay, that sounds like your sending unit & sending unit wire are probably okay.

• I can trick the gauge into any reading I want using resistors.

What resistor values are you using? Does the gauge read empty with the sending unit wire connected directly to ground (zero ohms)? And does it read full with a 90Ω resistor substituted in place of the sending unit?

• I can read the exact fuel level if I power the car off and hook the gauge to the power and ground on the battery and the fuel reading post directly to the sending unit wire

That tells me you might have a problem with the power or ground connections to the gauge since jumpering them directly to the battery makes the gauge work okay.

To narrow it down, try leaving the gauge hooked up to the stock wiring and only jumper one connection at a time. For example, if the problem goes away when you jumper the gauge's ground terminal directly to the negative battery post, you know there's a problem with the stock ground connection.

[mpierce9]

Lots of information here to digest but quickly I'll mention two things.

1. Regarding resistors. My original post referenced at the top has many pictures included walking you through my resistor exercise. I essentially used increments of 10 OHM resistors to measure the impact to the gauge. If I directly ground the sending unit post the gauge will read at an empty tank.

2. Let me quickly present a thought about your final approach which I will defiantly do. This is to help me clarify.

Originally Posted by ray_mcavoy

To narrow it down, try leaving the gauge hooked up to the stock wiring and only jumper one connection at a time. For example, if the problem goes away when you jumper the gauge's ground terminal directly to the negative battery post, you know there's a problem with the stock ground connection.

I can and have taken the gauge out of the original wiring and hooked it directly to the power on the battery, ground on the battery and fuel sender unit wire. While the truck is powered off the gauge reads accurately, once the truck powered on the gauge reads at 3:00. Since the state of the powered line and the ground line would not change regardless of whether the truck is powered on or not since we have a direct connection to the battery wouldn't that indicate the fuel sending line has an issue? Or some other problem?

Would that indicate that the issue would be in the sending unit wiring? Since I have bypassed the dash cluster power and ground?

I'm just trying to understand.

[ray_mcavoy]

Originally Posted by mpierce9

I can and have taken the gauge out of the original wiring and hooked it directly to the power on the battery, ground on the battery and fuel sender unit wire. While the truck is powered off the gauge reads accurately, once the truck powered on the gauge reads at 3:00. Since the state of the powered line and the ground line would not change regardless of whether the truck is powered on or not since we have a direct connection to the battery wouldn't that indicate the fuel sending line has an issue? Or some other problem?

Okay, I see what you're dealing with now. I missed the part about the gauge still going to 3 O'Clock when the ignition switch is turned on even with the gauge's power and ground connections jumpered directly to the battery.

With that in mind, I agree ... that does point towards a problem with the sending unit leg of the gauge wiring. And since it only happens when you turn on the ignition switch, it sounds like some other circuit is backfeeding power to the gauge's sending unit wiring.

Try this:

Temporarily remove the fuel gauge from the cluster. Plug everything else back in as normal. Set your multimeter to "DC Volts" (use the 20V scale if it's not autoranging). Connect the meter's black lead to a good ground. Connect the meter's red lead to the fuel gauge's sending unit terminal in the cluster (should be the one off to the right when viewing the cluster from the front if I remember correctly). Or connect the meter's red lead to somewhere convenient along the brown sending unit wire.

Now turn on the ignition switch to see if you get a voltage reading on the meter. With the fuel gauge temporarily removed, you should not see any voltage on the sending unit wire. If you do, there is definitely something wrong. Try repeating the test again but unplug the cluster this time. That'll let you know if the problem is in the truck's wiring or the cluster.

[mpierce9]

I originally typed the statement below. Then reread your reply. Just to clarify are saying if I cut the fuel sending line say somewhere in the middle I shouldn't get a resistance from ground reading on that line? I left my original reply below in case it has insight.

I'm sorry to say I have done this as well with the same result. Clutter wiring reads perfect until the truck key is in any position other than off. The wiring reads accurate in any scenario as long as the truck is powered off. In any scenario the fuel sending wire reads 100% resistance to ground once the truck key is in any position other than off. Every inch of the brown fuel sending wire reads accurate with the key turned off. Any other thoughts?

[ray_mcavoy]

Originally Posted by mpierce9

I originally typed the statement below. Then reread your reply. Just to clarify are saying if I cut the fuel sending line say somewhere in the middle I shouldn't get a resistance from ground reading on that line? I left my original reply below in case it has insight.

No, don't cut the sending unit wire. Just unplug it at some convenient location (such as unplugging the cluster connector) so you can hook your meter up to it and take a measurement.

Originally Posted by mpierce9

I'm sorry to say I have done this as well with the same result. Clutter wiring reads perfect until the truck key is in any position other than off. The wiring reads accurate in any scenario as long as the truck is powered off. In any scenario the fuel sending wire reads 100% resistance to ground once the truck key is in any position other than off. Every inch of the brown fuel sending wire reads accurate with the key turned off. Any other thoughts?

I'm not quite sure what you mean by "100% resistance" ... Resistance is measured in Ohms (Ω). And as mentioned earlier, you don't want to measure resistance in a live circuit ... that's why I specified setting the meter to "DC Volts".

Are you saying that your meter is showing full battery voltage (around 12.6V) on the sending unit wire with the key on and the gauge temporarily removed?

[mpierce9]

I don't mean to sound out of line but the brown wire that runs from the fuel sending unit measures the resistance to ground. Is that not correct? Is the sending unit that's located inside the gas tank not just a variable resistor adjusting the resistance to the ground line coming off the sending unit attached to the frame? There is no power on that line. Am I mistaken here?

[ray_mcavoy]

Originally Posted by mpierce9

I don't mean to sound out of line but the brown wire that runs from the fuel sending unit measures the resistance to ground. Is that not correct? Is the sending unit that's located inside the gas tank not just a variable resistor adjusting the resistance to the ground line coming off the sending unit attached to the frame?

Yes, you are correct.

Originally Posted by mpierce9

There is no power on that line. Am I mistaken here?

Here's where things get a bit more complicated. In a correctly functioning fuel gauge system, current flows through the coils inside the meter movement. Some of the current goes to ground through the sending unit. The remaining current goes to ground through the gauge's ground terminal. So yes, there is normally some voltage present on the sending unit wire when the system is operational. It's not full battery voltage but it is enough to cause a false reading if you try to take a measurement with your multimeter set to the resistance (Ohms) scale.

However, if you take the fuel gauge out of the circuit, there shouldn't be any voltage present on the sending unit line. But based on your description of how the gauge is malfunctioning, it sounds like there might be some other circuit that has somehow become shorted to the sending unit line and is feeding power to this line when it shouldn't be. That's why I suggested temporarily removing the fuel gauge and using your meter to see if there is any voltage present on the sending unit wire with the key on.

Ideally, you will not see any voltage on the sending unit wire with the gauge temporarily removed. But if you do, that's a sign that some switched ignition circuit is somehow managing to feed power to that line when it shouldn't be. And if that's happening, it will definitely cause the gauge to read incorrectly.

[mpierce9]

Sounds good. I'll check this tomorrow around 6pm and report back. Thanks for the extremely detailed responses. I hope this uncovers a next step.

[mpierce9]

ok, I'm pretty sure the result was null. . . I did attach a video of what i did. Hopfully it will be able to be viewed. Regardless. . .

iCloud Fuel Sending units tester

• I removed the gauge from the cluster.
• I attached a black lead to a confirmed good ground
• I attached the red lead to the confirmed sending unit post in the cluster
• I set multi meter to 20V
• I plugged mulit meter sensors into the DC measurement sockets

While the ignition was completely off there was 0.00 volts on the line
While the ignition was completely on there was 0.00 volts on the line

Entertain me on this thought: So there are pretty much three parts to this system.

1. The Fuel sending unit inside the fuel tank
2. The gauge inside the dash cluster
3. All the wiring in between

Sending unit:
If I unhook the brown wire on the sending unit from all contact to the dash (say cut it in half for drama sake). If it works over time I should see readins between 0 - 90OHM correct? I could essentially check it now, note the reading, go fill up and note that reading. if all is well then the sending unit works correctly. . . is this accurate?


Gauge:
I should be able to wire it directly to the battery and use various resistors to the sending unit post (the post on the right side if looking at the back of the gauge) in order to prove the gauge reads accurately at different levels of resistance.


The wiring between:
This is where I lose myself because of the complicated nature of how the gauge reads voltage and resistance. How do can i test this?

[ray_mcavoy]

Originally Posted by mpierce9

ok, I'm pretty sure the result was null. . . I did attach a video of what i did. Hopfully it will be able to be viewed. Regardless. . .

iCloud Fuel Sending units tester

The video link is telling me my web browser isn't supported ... it's been a while since I've done any software updates on this computer so that's not surprising.

• I removed the gauge from the cluster.
• I attached a black lead to a confirmed good ground
• I attached the red lead to the confirmed sending unit post in the cluster
• I set multi meter to 20V
• I plugged mulit meter sensors into the DC measurement sockets

While the ignition was completely off there was 0.00 volts on the line
While the ignition was completely on there was 0.00 volts on the line

Okay, that's good news. It doesn't mean we've found the problem yet, but it does mean we can rule out any strange short circuit or backfeed problem feeding power to the sending unit wire.

Entertain me on this thought: So there are pretty much three parts to this system.

1. The Fuel sending unit inside the fuel tank
2. The gauge inside the dash cluster
3. All the wiring in between

Yes, that pretty much covers it for a single tank truck. Trucks with dual tanks have some additional wiring and a switch to select between the two sending units.

Sending unit:
If I unhook the brown wire on the sending unit from all contact to the dash (say cut it in half for drama sake). If it works over time I should see readins between 0 - 90OHM correct? I could essentially check it now, note the reading, go fill up and note that reading. if all is well then the sending unit works correctly. . . is this accurate?

Yes, that is true. I have seen some defective (worn out) sending units that have "dead spots" ... they might give a normal reading in some float positions but go completely open circuit (causing the gauge to peg over full) when the float level reaches one of those dead spots. But that doesn't sound like the behavior you're describing where you are getting normal gauge readings only with the key off and the power & ground jumpered to the battery.

Gauge:
I should be able to wire it directly to the battery and use various resistors to the sending unit post (the post on the right side if looking at the back of the gauge) in order to prove the gauge reads accurately at different levels of resistance.

Yes, but double check your connections. I just took a look at a spare cluster to refresh my memory on the fuel gauge terminal post positions.

Looking at the back of the gauge:

• Left Post = Sending Unit
• Bottom/Middle Post = Ground
• Right Post = Power

The wiring between:
This is where I lose myself because of the complicated nature of how the gauge reads voltage and resistance. How do can i test this?

In reading back through this thread, it sounds like you've already done quite a bit of testing on the wiring already and so far the results seem normal.

But I do have a couple more questions and things to try:

(1) Was the sending unit wire still attached to the gauge when you took those resistance measurements along the sending unit wire and obtained the 64Ω readings?

The reason I ask is because the wire needs to be disconnected from the gauge in order to get an accurate measurement. When the wire is still attached to the gauge, you end up reading the combined resistance of both the sending unit and the gauge's internal coils.

(2) I might have missed this in reading through everything, but have you tried the following test?

• With the gauge installed & connected normally to the truck's wiring, temporarily unplug the sending unit wire on the gas tank end. Turn the key on and note the gauge reading. It should peg to the 3 O'Clock position.
• Now connect a jumper wire from a good ground to the socket on the unplugged end of the sending unit wire. Turn the key on and note the gauge reading. It should read empty.

If so, what are the results?

[mpierce9]

Originally Posted by ray_mcavoy

But I do have a couple more questions and things to try:

(1) Was the sending unit wire still attached to the gauge when you took those resistance measurements along the sending unit wire and obtained the 64Ω readings?

The reason I ask is because the wire needs to be disconnected from the gauge in order to get an accurate measurement. When the wire is still attached to the gauge, you end up reading the combined resistance of both the sending unit and the gauge's internal coils.

(2) I might have missed this in reading through everything, but have you tried the following test?

• With the gauge installed & connected normally to the truck's wiring, temporarily unplug the sending unit wire on the gas tank end. Turn the key on and note the gauge reading. It should peg to the 3 O'Clock position.
• Now connect a jumper wire from a good ground to the socket on the unplugged end of the sending unit wire. Turn the key on and note the gauge reading. It should read empty.

If so, what are the results?

The sending wire was still attached to the gauge when originally tested.
I'm going to redo both of these test as described above. I don't want to use old reading. I do remember doing this in the past but the gauge was installed at the time. Seems like the gauge causes some odd readings. When i started doing this I thought it was a basic resistance lesson, now I know better.

[mpierce9]

Here we go. . .
I removed the gauge and started testing the wire from the sending unit. I tested all up and down it. From at the fuel tank connection to the post inside the gauge cluster. All reading were infinitely resistant or 100% resistance (my multi meter just shows a '1' on the left hand side of the screen, I'm not sure how to articulate resistance in the form of 1 bazillion Ω ).

I plugged in the gauge as it should be. I unplugged the sending unit line from the sending unit inside the fuel tank. I powered the car on and the gauge pegged at the 3:00 position as expected. I then touched the wire to the ground on the battery and powered the car on and the gauge swung to the empty position. I also had a few 10 Ω resistors that I chained together and each incremental 10 Ω resistor caused the gauge to move closer to full, as expected. This was all using the full length of the brown wire.

At this point I feel the wiring and gauge are working properly. I do have two fuel gauges and I used both for all the tests one is original (1975) the other is from a 1976 Blazer. The 1976 was a bit more snappy to characterize how quick it would change from one position to another but both landed in the same location on all tests.


To better understand the reading that have caused me such grief I checked the resistance to ground between the gauge sending unit post and the battery ground with the truck powered ON. I received a reading of 64 Ω resistance(sound familiar?). now realizing that each reading I had taken in the past was delivering information about the resistance on the gauge, not the sending unit.


I just ordered a new sending unit. I'll replace mine on Saturday and hopefully reply back with stories of how messed up the one i currently have in the tank is.

[ray_mcavoy]

Originally Posted by mpierce9

Here we go. . .
I removed the gauge and started testing the wire from the sending unit. I tested all up and down it. From at the fuel tank connection to the post inside the gauge cluster. All reading were infinitely resistant or 100% resistance (my multi meter just shows a '1' on the left hand side of the screen, I'm not sure how to articulate resistance in the form of 1 bazillion Ω ).

That would be an open circuit!

I plugged in the gauge as it should be. I unplugged the sending unit line from the sending unit inside the fuel tank. I powered the car on and the gauge pegged at the 3:00 position as expected. I then touched the wire to the ground on the battery and powered the car on and the gauge swung to the empty position. I also had a few 10 Ω resistors that I chained together and each incremental 10 Ω resistor caused the gauge to move closer to full, as expected. This was all using the full length of the brown wire.

At this point I feel the wiring and gauge are working properly. I do have two fuel gauges and I used both for all the tests one is original (1975) the other is from a 1976 Blazer. The 1976 was a bit more snappy to characterize how quick it would change from one position to another but both landed in the same location on all tests.

I agree ... that does sound like the gauge and wiring are okay.

To better understand the reading that have caused me such grief I checked the resistance to ground between the gauge sending unit post and the battery ground with the truck powered ON. I received a reading of 64 Ω resistance(sound familiar?). now realizing that each reading I had taken in the past was delivering information about the resistance on the gauge, not the sending unit.

Yup, that'll do it That's one of those things you have to watch out for when taking resistance measurements. If the component being tested isn't isolated from the rest of the circuit you can end up measuring the combined resistance of other connected components.

I guess one thing that made this particular case a little more confusing is that the 64Ω you measured just happened to correspond with roughly what would be expected of a good sending unit with 3/4 tank of gas. I have to admit, I didn't think much of it at first either. But then after you did all of the other testing that was confirming everything else being okay, I got to thinking about how you were taking those resistance measurements and if the gauge was still connected.

I just ordered a new sending unit. I'll replace mine on Saturday and hopefully reply back with stories of how messed up the one i currently have in the tank is.

Sounds like you're probably on the right track with replacing the sending unit.

Maybe you've already checked it and I missed it in reading through the posts, but the sending unit ground is about the only other thing that could potentially be a problem. The ground wire runs from the body of the sending unit to a bolt on the frame. A poor ground connection between the frame and the rest of the truck could do it too but that'd likely result in problems with more than just the fuel gauge.

[mpierce9]

Ray!!!!! I haven't replaced the fuel sender yet. I did get it in the mail and hooked it up to test. I watched the fuel gauge while moving the arm up and down and it worked PERFECT! Thank you so very much for your time and patience explaining the gauge to me. I hope and pray I never have to take my dash apart. I'm planning on dropping the tank later today and changing out the sending unit.

[ray_mcavoy]

You're welcome!

I'm glad to hear that the new sending unit did the trick and the fuel gauge will no longer be driving you crazy!!