PDA

View Full Version : Sending unit not sending...



qwik1320
08-28-2004, 09:55 AM
:banghead: Help!

New gas tank and sending unit installed in my 73 Camaro and I added enough gas to get to station to fill it up, the gauge read "E" but I figured it was because I only put a splash in the tank. Unfortunately now that it's filled (well at least 3/4 full) it's still reading "E"

Here's what I see -
1. there is power at the plug on the rear wiring harness
2. when both the power and ground are connected the gauge reads "E"
3. when only the power is connected (ground disconnected) the gauge reads "E"
4. when I have nothing connected (no power or ground) or just the ground connected the needle goes WAY past "F" to about 3:00

Anyone have any ideas how to fix this???

next69
08-28-2004, 05:04 PM
Todd,
Did the gauge ever work? Here is a link that will explain exactly how this cicuit works.

http://auto.howstuffworks.com/fuel-gauge.htm

qwik1320
08-28-2004, 05:17 PM
Yes the gauge worked before the accident... I was rear ended last year and pushed into another car but I think the gauge still worked after the accident, honestly I can't remember for sure though.

Is there a way to check the gauge to verify it works and then work my way back to the sending unit? The last thing I want to do is syphon the gas again and drop the tank.

Thanks for the link to that site, there's a lot of neat stuff on there!

next69
08-29-2004, 03:14 AM
Todd,

If you look at the illustration on the link I posted you see with a full tank of gas the resistance is very high thus a low amount of voltage is sent to the gauge and the lower the fuel gets, the lower the resistance gets, thus a higher level of voltage makes it to the gauge. So based on the illustration whith nothing connected the gauge should read full. If you wanted to test the gauge you could take 12 volts and a potentiometer and hook it up to the gauge and reproduce variable levels of voltage and see if the gauge responds accordingly.

derekf
08-29-2004, 07:07 AM
The HSW site oversimplifies some of it, I think.

The gauge receives +12v when the ignition is on. The wiring diagram for my '65 says this is a 20g B/P (blue/purple?) wire - but your colors may vary, if you've got a wiring diagram for yours you can verify pretty easily.

The gauge is grounded through a tan wire (at least in 65 it was, your colors may vary. I'll be describing this wire as tan, anyway), which is connected to the sender in the tank. The sender, as described on the HSW site, is a variable resistor. When the tank is full, the resistance of the sender should be 88 ohms and when it's empty it should be 0 ohms. The other side of the sender connects to a chassis ground.

From what your troubleshooting has already determined, we can figure this:

1. When you've got both the tan line and the chassis ground connected, the gauge sees a direct connection to ground (close to 0 ohms) and reads E - if your tank was empty, this would be correct, but it's not.

2. When the tan line is connected and chassis ground is not, the gauge still sees the direct connection to ground (again, close to 0 ohms) and still reads E. This isn't correct with any amount of fuel.

3. When you disconnect the tan line, the gauge no longer sees the ground (open circuit) and reads way past F. This is expected behavior but confirms that it's not your gauge stuck on E.

Sounds like you've got a short to ground somewhere after the tan wire you're disconnecting (I'm assuming you're disconnecting back by the tank and not at the gauge). Verify by unhooking the chassis ground from the sensor, unhooking the tan wire, and measure the resistance between the tan wire going to the sensor (not the one going back to the gauge) and the chassis ground. I imagine you'll see a very small resistance.

If you'd rather check the sensor, give it 12v in at the blue/purple wire (or whatever color it is in your car) and hook a ~100 ohm variable resistor in the ground.. That should get you pretty much the full sweep of the gauge.

Hope this helps, if I over- or under- simplified things please let me know.

qwik1320
08-29-2004, 12:29 PM
Derek,
OK, I checked the tan wire that goes to the sending unit. I put a meter between the ground wire off the sending unit (disconnected from the chassis) and the tan wire (unplugged from the harness). I get about 5.6 ohms resistance. The tan wire is free from where it enter the trunk back to the sending unit and I visually inspected every bit I can see - I thought maybe it got pinched between the tank and trunk floor. I'm trying to see if the actual connection point for the tan wire to the sending unit (the brass stud with the rubber boot) is pinched...maybe that's the problem? Do you think it could be something that simple?

derekf
08-29-2004, 01:03 PM
I do.

To my mind, the next step is to try and isolate where the short to ground is.. either it'll be a wire pinched/touching, or it'll be something inside the sender.

I don't know which way I like better... at least if it's a bad sender, once it's fixed you know that it's really fixed... there's nothing worse than an intermittent problem.

qwik1320
08-29-2004, 01:29 PM
Derek,
Well there's room between where the tan wire connects to the sending unit and the trunk floor. I took a piece of plastic milk jug and slid it in between pretty easily. I have my old mangled sending unit here and I measured between the 2 contacts and that reads open, which I guess is what this should read.

I watched the gauge again as I had only the ground connected, only the tan wire connected, etc. and now it's not moving at all no matter what's connected.

Maybe it is the gauge but the 5.6 ohms between the tan wire and ground isn't right either. My gauges have that plastic printed circuit board on the back. Any other thoughts?

I think I'll get the guys on the 2GCOG site thinking about it too with a post there.

derekf
08-29-2004, 04:07 PM
The contacts should not read "open". There should be a resistance (as described above).

The gauge should move if the tan isn't connected - it should move way past F as you were describing before. (Silly question - verify you're testing with the ignition on)

5.6 ohms between the tan and the ground is correct if the tank is almost empty.

qwik1320
08-29-2004, 04:30 PM
I turn the ignition to "ON" to check, the engine isn't running but everything's hot. Don't know what happened but the needle on the gauge isn't moving off E now.

If 5.6 is about right for an empty tank then it sounds like my float isn't floating or something else internal to the tank. Damn, I don't want to drop the tank again, what a PITA!

My old sending unit does read open through the whole range of travel for the float. I held one lead to the brass pin and the other to the ground lug.

derekf
08-29-2004, 06:09 PM
5.6 is pretty close to the bottom.

If your gauge isn't moving past F when the tan is disconnected, we've got even more problems now. Verify that power's coming to the tan wire (from the gauge) - you may have popped a fuse or something?

qwik1320
08-29-2004, 06:35 PM
Derek,
I'll have to check later this week - back to work tomorrow and the kids have a lot going on. The "idiot" light for the GEN light works but I'll have to look and see what fuse the gauge is hooked up to. Thanks for all your help, I'll let you know what I find.

68LSS1
08-29-2004, 09:55 PM
I agree with derekf that it appears you have a short and now your short has possibly burnt a fuse. I disagree with some other things mentioned. In this type of indication circuit if you remove a wire aft of the indicator you should get a full positive indication on the gauge or in this application a full reading. The resistance with a full tank of gas should be close to zero and increase to approx 90 ohms when empty for most GM sending units. With a open circuit our gauge pegs out (no resistance) so with a little resistance it will be on the full mark. Also he said he had 5.6 ohms with about 3/4 tank of gas. This is about right. Tank is close to full and has very little resistance. Current is higher to the gauge when the tank is full due to the decreased resistance and current will decrease with a empty tank as the resistance increases. (Remember Ohms Law E=IXR). Remove both wires from sending unit and from gauge (by dash connector) and check both wires to ground. Possible I'm wrong but I troubleshoot these type circuits all the time on aircraft. They use the same type system (circuit) to monitor temperatures. Good luck.

qwik1320
08-30-2004, 07:08 PM
Brian,
What you said make me feel a WHOLE lot better. I get about 5.6 ohms and I figure there's about 12-14 gallons in the tank. Not having to drop the tank would make my day. This week sucks with work so I may not get to check until the weekend, I'll post what I find. Keep the good ideas coming guys!
:icon996:

derekf
08-31-2004, 05:31 AM
Brian,

I'm afraid I have to respectfully disagree.

Quote taken from Team Chevelle (Link (http://www.chevelles.com/cgi-bin/forum/ultimatebb.cgi/topic/22/2399.html?))


John_Muha
Moderator
Team Member
TC #623


posted 28 September, 2002 09:28
--------------------------------------------------------------------------------

quote:
--------------------------------------------------------------------------------
Originally posted by MOABDADC22:
Maybe I am wrong but wouldn't an empty tank read 90 ohms and a full tank be close to 0 ohms? (Assuming the sending unit is 0-90 ohms.)


--------------------------------------------------------------------

Empty is 0 ohms. The tank float is attached to an arm of a rheostat. The rheostat arm is the sender wire. The bottom of the rheostat is attached to ground. The top of the rheostat goes nowhere.
As the arm moves up and down resistance changes. If the sender wire is open it drives the gage super full because resistance is greater than 90 ohms.


or even from Autometer's (http://hp.autometer.com/products/pro_comp/full_sweep_electric/full_sweep_electric.html) page:



Fuel Level
(0 O Empty/90 O Full)
For most GM vehicles from 1965 to present.
5415

If full was 0 ohms, gauges with extra resistance would read extra-low - it'd require backward current flow for fuel gauges to read past F.

qwik1320
08-31-2004, 04:03 PM
Well I got a few minutes to check and power is getting back to the sending unit through the tan wire (about 8 volts). The gauge doesn't move at all now no matter what wires I play with. I have a few 67 Camaro instrument clusters in the shed, do you think I can hook one up back at the tank to see if it gets a reading? I just need a 12V power source (reverse lights) the tan wire to the sending unit and a ground for the gauge to work right?

derekf
09-02-2004, 05:33 AM
Sounds about right, but I'd think you'd have more like 12 volts. What's the gauge in the car getting?

qwik1320
09-04-2004, 12:56 PM
Derek,
OK here's what I see now:

1. The tan sending unit wire from front of car (where wiring plugs into gauge) to back of car (at last plug before sending unit) is good. No shorts
2. I get 14+ volts going into the fuel gauge, it's about 9 volts back at the plug for the sending unit.
3. When all the wires are connected (and power turned on), the fuel gauge reads E.
4. When the tan sending unit wire is disconnected and the power is on, the fuel gauge reads past F (3:00).

I have the dash apart and I want to verify the gauge is working. I'll try your suggestion of hooking up 12V and some resistors to simulate fuel levels. There are 3 lugs on the back of the gauge that I can use, I just have the see which is which. If that works then I guess I'll be dropping the tank...

Any other ideas?

derekf
09-05-2004, 06:14 PM
Other than the 9v at the plug everything sounds pretty much as I'd expect, and I've never checked at the plug to know if that's proper or not.

Keep us posted as to the gauge-test (although it sounds like the gauge is fine)

qwik1320
09-07-2004, 05:52 PM
Well I checked the gauge using a few different resistors - everything checks out. Guess I'll be dropping the tank at some point and seeing what's up, sounds like a winter project. :hammer:

68LSS1
09-07-2004, 06:53 PM
Sorry for the bad info qwik1320. Derekf, do older Fords read the same way because I assumed that GM read the same as those do (assuming the way I was saying is correct on Fords)?

derekf
09-08-2004, 04:27 AM
Brian, I'm not so sure about Fords, I've never dealt with them.