Chevy Nova Forum banner

Fuel gauge woes

5K views 52 replies 7 participants last post by  Winch 
#1 ·
62 Nova vert, new gas tank and sending unit. Gauge shows 1/4 tank even after I filled it up and topped it off yesterday.
When I bought the car the gauge showed 3/4 or a little more when filled up but would run out of gas with gauge still showing 1/4 tank or a little less.
What should I check?
 
#2 ·
Hey Winch,

I have to clarify, did you just buy that new sender and tank or were they new on the car when you bought it?

Either way you likely have a grounding problem to chase but there's good details inside this Section on how to quickly troubleshoot the circuit - and verify your gauge as well, before dropping the tank.

Under this Section you'll find at least 3 different Fuel Gauge threads you can review, consider what applies and approach for your next move. It might help to first determine the operating resistance of your gauge so you know what to expect. I believe 63 vintage gauges use high resistance for empty and low Res. for full but that detail is available within the threads.

Custom Jim and others offer great troubleshooting advice and insight throughout these threads.

Let us know what you find, please.
 
#3 ·
The first thing I would check would be the ohm range of the new sender vs the ohm range of the gauge. They must match for the gauge read accurately. Second would be the float on the sender. It must be free and have a full range of motion.
 
#4 ·
To clarify I installed a new tank and sender when I restored the trunk and floor pans which included fabricating and welding on the last 20 inches or so of the frame rails. As best as I remember now I got the sending unit from Rock Auto. A stock replacement.

I'm not sure how to check the ohm range as suggested. Are you saying to check what the manufacturer says the range is or actually do a test with a meter? It's been a couple of years now since I put the tank in and yesterday was the first time I filled it up.
 
#10 ·
When I first started getting into knowing how a fuel gauge system worked, I did the two basic tests of disconnecting the sender wire in the trunk and turning on the ignition and the gauge should swing to full or above. The second test required leaving the key on and then on the wire end that went towards the gauge and NOT the tank, if this was grounded, then the gas gauges needle should go to empty or below.
These two tests were the only thing I though was needed to verify if I had a gauge issue and/or a sender issue but have since found out these two test can give you some false positives.
I never knew years ago that the fuel gauge has what they call a shunt across the back terminals of the gauge and if this shunt is bad or has issues, then the basic tests of disconnecting the sender wire and then grounding it could show the gauge working right but in actuality you could then put in a sender that does work and perform right but still have an issue of the gauge reading wrong.

SOOOOOO, with all that said, ideally get some resistors to mimic the sender values it can output and then try these and if the gauge performs as it should with these resistor tests then chances are it's from your test point back to and through the sender and where it is grounded at.

On a 0-30 ohm setup. shorting out or grounding out the sender wire to the gauge will be 0 ohms. Now if one put in a 30 ohm resistor between the sender wire going to the gauge and the other end of the resistor to ground the gauge will be seeing 30 ohms of resistance and should read Full. Now do the math and if we put in a 15 ohm resistor in place then the gauge should read 1/2 full. By doing more math a 7.5 ohm resistor would or should get the gauge to read 1/4 full and guess what,if we series the 7.5 with the 15, the total ohms would be 22.5 and then if inserted would get the gauge to read 3/4 full.

Gotta test things fully and do more than just guessing as to the issue and adding more tests than just the two of disconnecting the sender wire and then grounding it.

Jim
 
#11 ·
OK I'm going to show my total lack of understanding of all things electrical when I ask is there a current flow in that wire that runs from the gauge to the sending unit? How else would it be able to measure ohms of resistance? Isn't it the ohmmeter itself that provides the current via the probes when testing for resistance?
 
#12 · (Edited)
I'm just going to add that Custom Jim inserted a video within a thread in this section that helps understand the testing procedure he uses.
Find it on Page 4 of this Section in a thread by Ruperflood entitled "Another Fuel Gauge ??? Help Requested".
(Sorry, my browser will not allow linking within this forum)

I used Jim's method and saved dropping the tank and 'enjoying' (NOT!) a gas bath, as sanitizing as they are. :mad:

Yes, there is a very small current 'flow' in essence but in this type of circuitry, current can cause an electrical arc (spark) and not really what you want near gasoline vapours. Think of it more as 'information' the sender is providing to the gauge and your anxiety should subside enough to drive again.
And yes, an Ohmmeter has an internal battery to allow measurement of resistance but again, don't get hung on current. It's important, yes but not what causes these problems. A different 'current' value is a symptom of the problem but a change in resistance is a cause and an absence of sender voltage is another. Is that change caused by a high resistance grounding point (bad ground) or a bad wiper on a sending unit, a bad gauge, an open wire or a stuck float?

Consider there are two circuits involved:

Circuit #1 is Switched (Keyed) Bat+ to the gauge and a ground wire at the gauge (provides lighting (often) and gauge operation).

Circuit #2 is the Sender connection point on the gauge with a single wire to the Fuel Sender at the tank. Electrically speaking, the sender wire is the moveable 'pointer' on the Fuel Sender's wiper inside the tank. The float physically moves this pointer across the wiper.
There is also a ground wire from the Fuel Sender to an appropriate ground point at or near the tank. (BTW - That may have been damaged when you did your Trunk work)

The ground wire at the tank is, electrically speaking, connected to one end of the Fuel Senders wiper. Think of the wiper as a fixed point. As the gas level changes, the float moves up and down moving the pointer on the wiper lengthening or shortening the electrical path from the pointer through the wiper to ground (varying resistance). This "Sensor info" based on the position of the pointer on the wiper is sent to the gauge where that Sender value is interpreted and reflected as a visual representation of that varying resistance - like an Ohmmeter. Circuitry inside the gauge dampens the Sender info values preventing the needle from bouncing as the gas sloshes about inside the tank.

Note: In my limited experience, I have found a Fuel Gauge needle that doesn't change (stuck) even with key removed has been resolved by repairing the ground point connection at the Tank.

If anyone finds errors within the above, please feel free to point it/them-out for all to benefit.
 
#13 ·
Wow, thanks for pointing me to that video. I've been reading a lot of threads but hadn't got to that one yet.

I suspect you are right about a ground problem. I think I see how to check it out now. I hope I can find some resistors if needed. All our Radio Shacks went out of business around here.
 
#16 ·
I hope I can find some resistors if needed. All our Radio Shacks went out of business around here.
See if you have a local electronics repair place. If not, maybe mail order. You only need 3 (7.5 ohm,15 ohm, and 30 ohm). These values do not have to be exact to where if you find a 5 ohm, that would be fine. The same amount of tolerance on the others is fine.

Looking at service parts here at the shop we have resistors in ohm values of 7.5, 15, 27, and 33 and I would think a better electronics repair business would also stock these values.

Jim
 
#15 ·
Hey guys, I just dealt with this issue just 1 month ago.... not that I am expert/nor electrically adapt, but it is easy to go around and around on this, which of course, I did. In the end, I had a old sending unit/float, bad wires (or grounding), and a faulty gauge. I bought the new sending unit from Classic Industries to ensure the type was correct for car. Also, the 62/63 shop manual talks about the types of faults and that the gauge type is "floating". Therein, there are gauge 'floating' symptoms and gauge 'indicates full' symptoms. The incorrect sending unit type/oms is not talked about as its assumed 'you get the right one'. :) I recommend to check to ensure its compatible with your car. Here is the one, Sending Unit, to check against as I bought it and installed, it works.

Importantly, I experienced both 'symptoms' from 'floating' and 'indicates full'. I thought the same that 'the sending unit was of wrong type'... it was correct type. I also thought to test the oms or to manually move the float to judge full/empty. But the manual does not suggest this, it basically says: if its 'floating' its a wiring issue, if it 'indicates full' its a short, if reading is incorrect its the sending unit. (yes, just replace the old unit and it threw me as, isn't a short a wiring issue? :) )

I think here, we talking about a 'floating' symptom. I also started with a floating gauge and also that it somewhat indicated something. So, 'some' reading is not a good path to solving in my opinion, again, just ensure the part is correct (!). After that, the manual says: When the key is 'on' and the gauge 'floats', it seemed to say "this is a wiring issue" which I tackled.

Here is what I did,
A. Symptom 'floating' gauge with key 'on'.
1. Replaced the old tank sending unit
2. Replaced the wires
3. Cleaned grounding contact (near tank)
3.1. Tested ground by connecting to Neg battery terminal directly with a long wire. (test failed, but was just to ensure the non-functioning gauge was not result of bad sending unit grounding)

B Symptom changed to 'indicates full' with key 'on' (normally wired with new sending unit)
--- I was a bit confused as "what else is there?" ---- 1. Sending Unit (check), 2. Wiring (check). 3. The Gauge itself (?).
4. Pulled Instrument Cluster
5. Removed fuel gauge (very carefully - the resistor is wire coiled and exposed - here is the same type but not the same gauge as Nova). if broken, that is expensive, but my thinking was 'its broken already'
6. Disconnect the parts, very carefully, clean with copper brush. Clean very sensitive parts with only a small amount of spray cleaner. (I used electrical spray cleaner)
7. Reassemble Gauge into Instrument Cluster
8. Clean Instrument Cluster grounding
9. Lean the Cluster into place and test.

Cleaning the gauge itself worked for me, works perfectly. Old Gauges can be the issue. Most times guys replace the entire instrument cluster. The fuel gauge seems 'solid state' so the contacts (screws, nuts and copper contacts) were old and dirty. Brushed clean and spray cleaning the rest (very carefully!).

.02 Auto part Wire Engine Technology Vehicle
 
#17 ·
I fiddled with it a little this morning and my results are not what I expected. Here is what I did:

1. Disconnect sending wire at sending unit
2. Turn ignition on - needle still shows 1/ 4 tank and hasn't moved.
3. Unplugged double connector (pink and brown) on back of fuel gauge while still in car
4. Checked voltage on pink wire - shows 12v
5. Rigged up jumper to connect pink wire to it's spade on the fuel gauge so as to leave the brown sending wire off .
6. Turn ignition on - needle still shows 1/ 4 tank and hasn't moved.
7. Rigged up jumper to attach sending wire terminal to ground
8. Turn ignition on - needle still shows 1/ 4 tank and hasn't moved.
9. Tried various ways to make insure cluster is grounded but no difference. Looks like a 62 fuel gauge does not have a dedicated ground pole so it must ground through the cluster. All other lights and gauges in cluster work OK so I assume I have a good enough ground there.

What's next? Pull fuel gauge out of cluster? Maybe check continuity of sending wire to insure it isn't shorting out on it's way back to the tank?
 
#19 ·
I would not freak out yet. The shop manual, pictured, shows it could just be the electric connections. (Gauge does not register correctly, hoping its not just bent!) B. C. and D (high resistance, partial short, loose connection) seem to say to me, its dirty. That was it for me. Pull the gauge, clean it (I mean pull the gauge carefully, the coil of wire is exposed.)

I think Custom Jim is correct too. if the oms are wrong, then it could be reading 1/4 tank when empty (I don't know really). I think he is saying' if the gauge is getting oms but wrong amount, then its the oms. I'm tactical, I couldn't figure out which setting on the tool was oms! So, went with option B, just clean it.

I think many people just replace the entire instrument cluster for 600+ dollars (too much for me), there are guys on the site who will get a used one for much cheaper. Text Material property Font
 
#20 ·
This may not be an issue but these "rigged" up jumpers, could those be an issue ?. I know I have grabbed what I "thought" were good alligator clip ended jumpers only to find out they had an open between one end and the others or I've shoved a stripped wire end into a connector only to have it contact with rust, paint, or oxidation preventing a good connection.

If all else fails, get the gauge out, look over things real good, clean connection posts if needed, and then bench check things.

I know on an old test I did on a factory indash tachometer for my 74, there were push in connections between the case of the housing and the circuit board and maybe yours is assembled the same way ? and corrosion and poor connections are buried inside ?.

Below is the inside of the tachometer housing with the female connectors inside that then male connectors off of the circuit board push onto when things get assembled.





You also stated:
4. Checked voltage on pink wire - shows 12v
5. Rigged up jumper to connect pink wire to it's spade on the fuel gauge so as to leave the brown sending wire off .
9. Tried various ways to make insure cluster is grounded but no difference. Looks like a 62 fuel gauge does not have a dedicated ground pole so it must ground through the cluster. All other lights and gauges in cluster work OK so I assume I have a good enough ground there.

Now did you check to see if you had 12V on the pink wire still at the terminal of the gauge ?. Reason I'm asking is a meter needs a VERY small amount of current/amperage to measure voltage and if there is a poor power connection tow the gauge, when the gauge is then powered up and in the circuit and working, if the voltage goes away, then it's a power delivery issue. Sorta like checking a battery. It might show 12 volts while sitting there but will then not crank over a motor BUT if you measure the voltage WHILE trying to crank over the motor and the voltage drops to 6V or less, then it's a battery issue.

Jim
 
#21 ·
As for that jumper I attached a female spade connector on one end and a male on the other. Jumper is about 3 inches long and the ends are crimped on good. With it hooked up like that I was able to touch the male spade on the gauge with my meter probe when I had the jumper on it and it shows 12v.

Next I removed both wires from the sending unit and attached my ohmmeter. Red to sending terminal and black to ground post. Shows 25 ohms. Tank should be full. I filled it until it wanted to come back out on me. Then drove only 5 miles home. I might be getting 25 ohms instead of 30 because back end is jacked up pretty high so I can get under there??

I also noticed that I was not getting a good ohms reading when I touched the black probe to the frame. It's all been thoroughly cleaned and undercoated. I was able to get a good reading at some places such as the little bracket that holds the union of the brake lines. So I made a short jumper with 3/8 inch circle connectors and ran that from the brake bracket to the screw that attaches the ground wire to the gas tank bracket. Then I got back under the dash and connected ohmmeter to brown wire and to ground and it shows 25 ohms so I'm confident I'm getting a good connection from gauge to sending unit to ground.

I suppose I could siphon out 4 gallons to see if ohms drops to 18 or 19??

I guess I'll pull the gauge out next and clean it like suggested.
 
#24 ·
And now for a question on removing the gauge. The shop manual says loosen and lower the mast jacket from dash panel (See Steering, section 9, for complete procedure for lowering the mast). Do I have to do all that section 9 lists to remove the mast? All I really need to do is remove the clamp holding to the dash panel and let it swing down a little, right?
 
#25 ·
Found out I was wrong about having filled the tank completely. I put in about 3 more gallons from a small can. It is definitely full now. Ohms reads 30.9 so 25 ohms with 13 gallons was right. I also took volt reading on the sending wire post and it shows .5 volts. I assume that's about right?

I got the gauge out. Here are some pics. Also a pic of brushes and electrical cleaner I bought. How do I go about cleaning it?
Gauge Speedometer Tachometer Measuring instrument Auto part
Auto part
Gas
Scale
Material property Auto part Lubricant
 
#26 ·
That is exactly what mine looks like.
  1. My approach was anything that looks breakable - I sprayed just enough (yeah, what is 'just enough' - till it looks cleaner i guess)
  2. I gently moved the needle mechanism and sprayed.
  3. I removed the backplate and wire brushed contacts, nuts posts on both sides most till the copper was clear or no old residue remained.
  4. Good Luck!
here is a picture!
Auto part
 
#27 ·
Winch said: "Found out I was wrong about having filled the tank completely. I put in about 3 more gallons from a small can. It is definitely full now. Ohms reads 30.9 so 25 ohms with 13 gallons was right. I also took volt reading on the sending wire post and it shows .5 volts. I assume that's about right?"

Where did you measure the 30.9 Ohms? At the Fuel Sender across/between the Fuel Tank's Sender wire and ground wires? If so, yes. You're Sender is good.

How did you measure the .5Vdc at the gauge? Between which two points? And what was setting on your Voltmenter? Should have been "Volts DC" and "20" Volts range. (Always set your meter to a range value above what you expect to see). If not, try again using those settings. Doesn't matter which colour meter wire goes where.
Where I've never measured the value of the Sender lead, we'd need Custom Jim to jump in here to validate your method and reading.

Like Mburnette6 says.....go REAL EASY on that gauge with those brushes where any wires are located but lay into the spade connectors to get 'em good.
 
#28 ·
Where did you measure the 30.9 Ohms? At the Fuel Sender across/between the Fuel Tank's Sender wire and ground wires? If so, yes. You're Sender is good.

How did you measure the .5Vdc at the gauge? Between which two points? And what was setting on your Voltmenter?
I stuck the ohmmeter red probe into the brown wire plug removed from the gauge and then touched black probe against metal of dash underneath. Yesterday I also tested ohms by touching probes directly to posts on the sender but I got the exact same readings as testing under the dash. By testing at the gauge I think it confirms a good circuit all the way back and a good ground return.

I measured voltage coming out of gauge by touching sender wire spade with red probe and touching a ground under dash. I did have voltmeter set to 20 on DCV. I just did that to see what it shows. I thought if zero then maybe that would indicate a problem inside the gauge.
 
#29 ·
Well no luck. I cleaned it as good as I could following your instructions. It just doesn't respond at all.

Does anybody repair these things? I'm not going to spend what I see for a repro ($250 - $400)! I think I'll get back to making a 68 console with a 4 gauge pod work. I was going to plug the 4th hole with a clock but now I'll put a matching Autometer Phantom fuel gauge in there.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top