I bought an 87 pickup a few months ago. When I fill up, the gauge reads full and by about 100 miles, it reads empty.
Well, I read a bunch of posts about fuel gauge problems and decided my sending unit was obviously bad.
The guy I bought it from said he had changed the sending unit and that it had worked for a while and went bad again, so I decided either he was confused or put in a used one or something.
I bought a new sending unit and today I dropped the tank and sure enough, the sending unit that was in it is brand new. Put a meter on it and it reads from 114 ohms to about 3-1/2, which is just what its supposed to do. Works smooth all the way.
So now what do I do?

 

test the guage, I assume there is a test in the FSM, if it's resistance is out of spec you'll need to source a new cluster or single guage. If you don't already have the sr5 guage with proper guages instead of lights now might be a good time to upgrade. A new fuel guage will be quite expensive compared to the full sr5 cluster swap.

It's also possible you have wiring problems between the tank and guage. You should test this while you have the dash apart to check the fuel guage resistence.

is the guage consistently off eg' always 100miles, or is it intermitten eg' hit a bump and it's full hit another and it's empty.

In theory it's possible to repair and refurbish the guages. Getting the needles off without making it worse is probably the worse part, or getting them on after you've gone to the trouble of replacing components or rewapping the coils.

 

Thanks. The test I saw in the fsm was to connect a light bulb in series with the gauge to see if the needle moves. I'll try reading more carefully and see if I can find anything more detailed.

The gauge doesn't change with bumps etc. It goes down correctly for a while, then jumps to about 1/4 then to E.
Doesn't seem like that would be caused by wiring, but this stuff is pretty mysterious to me.

 

The cluster contains a sort of 1960s-era mechanical voltage regulator to drive the fuel gauge. Since you've already ruled out a wonky sender (by checking it with a multimeter), I would suspect the voltage regulator. There is a test for it on page E-34 of the manual. http://personal.utulsa.edu/~nathan-b.../6combinat.pdf

Someone on this forum took that regulator apart and described it quite well; good luck finding the post, though.

Keep us posted.

 

Yep, the fuel gauge is way out of spec by that test.
IG-FU: 82.5 - OK
IG-E: 85.2 - Should be 239
FU-E: 95 - Should be 156
I'll hunt for the post about the voltage regulator.
Thanks

 

Scope103 - I looked for that thread you mentioned for about an hour with no luck so far.

Anybody have a hint how to find it?

 

Might be a 7908, not sure Kinda bored of googling. Just uhm pull it out and post some pictures. Regulator will be a three legged IC. Should be pretty apparent how to take it apart of it's not post some pictures and I'll try to walk you thru it.

 

Any reason to think that it's an Integrated Circuit regulator? I don't think silicon had been invented back then. ;^)

Note that the FSM test calls for: "check that the voltmeter needle vibrates near the 7 V position." Obviously, a modern regulator is not going to "vibrate" the voltmeter needle. Instead, it's like a thermostat, switching back and forth as the voltage rises and falls.

I too would like to find the post I think I recall. If I recall correctly, the poster considered replacing the mechanical "regulator" with a zener diode (which has a soft knee at 7v, but it has to be better than a relay). Unfortunately, we can't search on combinations of words, so searching for "fuel gauge regulator" retrieves every post that mentions fuel, or gauge, or regulator. That's not going to work.

But, who knows? Toyota might have gotten religion somewhere along the line, and your cluster might have a different current source for the gauge. I think you're going to have to take it apart and see if the voltmeter needle vibrates.

 

vibrates = switching on and off.

If it was switching slowly it wouldn't be a very good regulator

is THIS the thread you were thinking of?

Haven't read it yet but I'm about to. The other one I spoted that mentioned replacing the regulator with a diode and resistor specificly mentioned a 7908 regulator, I didn't bother to look up the specs on that, or even the FSM tests. Figured I'd get to that when the OP decided if he wanted to try and repair it or just throw a salvaged one in.

I uh actually have my old cluster sitting here in my office lemme poke at It i guess with the screw driver

 

Exactly. That's how we know it isn't a 7908.

Yes. It turns out "zener" was the magic word to find it.

 

I traced the 7V line out of the fuel gauge and into the temp gauge.

Yeah I didn't make it very far before I thought it best to go read that other post with a description of the circuit.

And you're perfectly right no since to use a 50 cent part when a 5 cent one will do, I forget to think large scale production some times.

 

I just got it back together so I can drive it tomorrow. Next time I have a couple of hours I'll pull the cluster apart and see what's inside. Hope I don't wreck something.

Scope103 said there was a *mechanical* voltage regulator though. Where does the IC fit into that?

Looks like a new replacement gauge would cost almost $200.
Maybe I can find another gauge cluster somewhere. I've seen a couple for sale.
Sounds kind of fun to try repairing the gauge though.

 

Sorry, I guess I need to learn to use the forum too.

 

An IC would/could replace the bimetal switch. The mechanical regulator is like a circuit breaker once the current flows it heats up and disconnects cools off and reconnects.

Source:
LM1117T-ADJ, 220 Ohm 1/4 watt, 1k Ohm 1/4 watt, 6+ inches of wire, 3 ring terminals, 1/8-1/4 inch (20guage?) ID shrink wrap, 1inch ID shrink wrap. And my favorite goto a bit of epoxy.
Construct:
Connect wire #1 to Vin solder and shrink wrap, this goes to the IGN terminal
Connect one end of the 220 Ohm to Vout, lay this on the plastic side of the IC and solder the leg resistor and wire #2(output wire) all at once.
Place shrink wrap to the Vout leg covering the output wire and solder junction, and shrink with heat.
Place small guage shrink wrap over the 220 Ohm resistor.
Bend the unconnected side of the 220 Ohm so it overlays the Vadj pin.
Bend the 1k so it overlays the 220 and overlays the Vadj pin.
Solder the two resistors to the Vadj pin.
Heat and shrink the 220 wrap
Solder ground wire to open end of 1k resistor.
Place and heat shrink wrap over 1k resistor and ground wire
Heat and shrink the wrap over the 1k resistor.
Connect the 3 ring terminals to the wires and heat shrink if needed(depends on your terminal choice)
Test:
Connect the output to your multimeter(set on ~12v), the ground wire and multimeter to Neg battery, and the Input to Pos battery. You should be reading around 6.9 volts.
Enclose:
(Optional)
Apply a heat sink if you want to to the back of the IC.
Using a piece of foil for a mold, you want to cover the front of the device with epoxy. eg, fill foil with epoxy and push device in plastic side down untill part of the IC is submerged. Trim after cured if desired, eg ugly not gonna fit in the Large ID wrap.
(End optional)
Place the device in the large ID wrap with 1/2-1 inch or more of extra at the top(side opposite of the wires), and heat shrink the portio around the device.
Prepare:
Punch a hole in the free section of the shrink wrap case, this is for mounting to the cluster.
Remove the Cluster bezel, the four(?) screws in the back of the fuel guage, two(?) in the front, and remove the gauge from the cluster.
(Now the tricky fun part I don't have pictures of mental or otherwise)
Locate the 7V marked screw hole on the back of the cluster and guage.
You want to identify the metal switch inside the gauge connected to the post/pin. It needs to be cut out and removed.
Reassemble cluster and guage.
Mount:
Attach Ground and 7v ring terminals to the back of the cluster. Attach the device with the IGN screw making sure to put the washer between the device and gauge. Another one on top if you desire to firm up the mount.


Now assuming I didn't mess anything up typing this all out in the little quick reply box You should have a new fangled regulator that won't wear out again in 20+ years *mashes "go advanced"* Wow thats a lot text for such a simple circuit.

I of course left out the output capacitor, tantalums suck ect.. If you don't have proper voltage in the test this might be the issue but it could be something else inspect the junctions and consult the directions. If the junctions are all correct and the regulator still won't start up and stay steady you'll need to add a >10uf low ESR (tantalum) capacitor between Vout and ground.

You can source parts from where ever is convienent, mouser.com digikey.com newark.com or even radio shack. Should cost more in shipping than parts. Shrink wrap is probably cheapest at your local harbor freight, but the above stores or your local auto store should have it.

 

Interesting. Did you actually try this, or is it just a "scratchpad" suggestion?

I've never done actual measurements, but this guy: https://www.yotatech.com/forums/f116...luster-259564/ did. He found that 7v wouldn't work, and he ended up with 9v based on actual measurements to make the gauge read more-or-less accurately.

Also, did you actually spec the power requirements? I calculate around 850mw, which is at the upper limit for the TO-220 package. http://www.ti.com/lit/ds/symlink/lm1117-n.pdf . The good news is that the LM1117 is self limiting, and will just reduce the current rather than set the cluster on fire. The bad news is that a sunny day with a full tank will cause the gas gauge to read irregularly low. Sometimes.

Or my calculations could be completely wrong.

It sounds like you have a spare cluster to work with. I'd be very interested to see what you come up with if you actually try your replacement regulator.

 

Just a quick response while I look into the current draw. The zener diode shunts 9v to ground. Leaving around 5v to power the guages.
http://hades.mech.northwestern.edu/i...s#Zener_Diodes

 

Uh, not quite. Take another look at the article you cited. A 9v zener diode "always" has 9v across it, regardless of the Vin. If it worked as you described ("shunts 9v to ground") it wouldn't be a voltage regulator, because the output voltage would be 5.1v with the engine running (Vin about 14.1v), and only 3.6v with the engine off (Vin about 12.6v). As the article describes, you have to size "R" so that it drops the "excess" voltage (whatever the voltage is above 9v) under all expected cases of load current and input voltage. And R must be have the wattage to handle the power dissipation (V squared / R).

It was the last sentence that caused amateurw7vp the most work.

 

I guess I need a standard internet disclaimer here. If you electrocute or burn something down you should know better than to follow advice of some random internet dweller already but PROCEED WITH CAUTION

I don't belive my yazaki (non sr5)gauge works as described in that other post.


12v flows in thru the bottom of the horseshoe on the left, and grounds at the top, this piece is firmly seated and does not oscillate. resistance is ~1Ohm.

7v flows from the bottom of the horseshoe thru the resistance wire and out the 7v marked pad. 7v-12v = 113 Ohm

F, power flows in thru F thru the 55 Ohm wire, energizes the other horseshoe and pulls the guage needle towards the right side(full).

The yakazi non sr5 temp guage has a resistence of 25 Ohm.

.. So there is no mechanical bimetal oscilator switch in this unit. Maybe in my sr5 cluster but not this one.

I'll not argue the zener voltage cause duh I'm wrong "Vout is the reverse breakdown voltage or zener voltage"

Now show me how you came up with such a high load (850ma) cause I'm not coming anywhere near that.

It's a blizzard out side and I'm not going to disassemble my dash in it
pictures of the cluster backside is here, A is +, C is -, B is input.

At 15v
Temp, 57 Ohm + Sensor reading, 263ma
Fuel, 123 + sensor reading(3 at full), 119ma

I lost my data sheet page But I'm thinking a ~5.7watt load might be a little warm for the to-220. Have to account for the ~200f cabin temp possibility.

So about the 9v vs 7v, he indicated he had a larger needle swing lowering voltage should make that more accurate.

 

P.S. I might have a failing grade in reading comprhension today, third or fourth time I read that response and I just picked up 850 milliwatt no amp duuur

PPS. Yeah definatly exceed my alotment for electrical reading. dump links and wander away..
Lm117 Datasheet
Random heatsink calc
Multi use calc, and charts

 

OK - I only understand the electronics part conceptually, but I'll give it a try.
I may even have a voltage regulator lying around from an old power supply project.
I'll pull the cluster apart and see what the gauge looks like as soon as I get a chance.
Thanks for all the interest in this.