Jump to content
dmorales-bello

Perplexing "FUEL" light malfunction

Recommended Posts

Posted (edited)

I bet that's just a slosh quickly dousing and cooling the sender. My DD does the same thing.

7 minutes ago, Zed Head said:

(sometimes I like to live on the edge). 

I always pictured you as that kind of guy.    :cool:   Zed "Danger" Head

Edited by Captain Obvious

Share this post


Link to post
Share on other sites

11 minutes ago, Captain Obvious said:

I always pictured you as that kind of guy.    :cool:   Zed "Danger" Head

Thanks.  But I have to confess that there was some calculating behind it.  I knew my numbers and what should be left.  Still, it was thrilling...

Share this post


Link to post
Share on other sites

No!   Own It!    LOL

But I know exactly what you mean... Off quick, and slowly comes back. I'm sure you nailed it and it's just fuel sloshing onto the sensor.

Share this post


Link to post
Share on other sites

Oh, and BTW, the discussion of bulb filaments changing their resistance dramatically as they heat up reminds me... Sorry for the diversion, but since we're so close...

The higher inrush current from a big-ish incandescent load (like headlights) isn't the most gentle thing that can occur to switch contacts. Between that, and the lack of arc extinguishing when you turn stuff off, you can eat up the contacts in the switch even though you aren't switching an inductive load.

@ETI4K  We were having a discussion some time ago and the million dollar question you didn't ask me is "OK, smart guy... Then if the headlights aren't an inductive load than why do the switches burn up like that? Huh? Why's that?" 

Share this post


Link to post
Share on other sites

So much of the early electrical systems on cars were(was?) developed around a deep understanding of the fundamental properties of the materials.  Today, it seems like everything is handled with computer code.  So, the ballast resistor works on the same principle as the filament in the bulb of the fuel light, I think, right?  Using that increase in resistance with current for a control purpose.

Here's a stretch, another cross-reference.  Would a condenser/capacitor on the headlight circuit damp the spark across the headlight switch contacts, like the condenser on a set of points in a Kettering type ignition system? 

Just testing my brain, the switch will still get hot and break the solder joints just from the current, over time, I'd guess, so a relay is the best solution.  But the pitting might be reduced, which contributes to the heating effect.

Share this post


Link to post
Share on other sites

did some more test after the video, the engine on (14v vs about 12.6 engine off battery voltage) resulted in about 8v at the back probe when the same test as the video was done. I filled up the tank this morn just to make sure I did not goof up my gas gauge. The light was coming on and off while driving to the station, gas slosh I presume, this was after adding about 2 gallons to the tank. I am pretty sure I was nearly empty as it took over 14 gallons to fill (so 2 I put in at home plus 14.5, 16.5 total, I must have been on fumes). So with lets guess 2.5 gallons in the tank it would come on and off, Gauge indicated "E" even with the 2.5 est gallons. Filled up it goes right to Full. I may have cocked the sensor a bit on the install. I think it should read a bit higher on empty., I will drive it around to the 1/2 mark, to make sure my gas is below the sensor hole on the side of the tank, then release the hold fast ring, and see if I can tweek the position of the sensor in the hole. Try to get a max reading on the gauge. fyi, the cold resistance of the thermistor reading it at the plug was about 3k which seems very high to me I=E/R 12/3000 or .004 amps, seems like it should be a lower resistance to allow a faster ramp up of the heat once its clear of liquid. I think there maybe other causes for the high resistance other than the thermistor. I should have soldered on lead from the thermistor can to the mounting clip, as noted before there was a fine white residue build up that I had to sand off. So my only fix was to clean that residue and then solder the loose wire that I presume was from the thermistor to the bottom of the can (it was just bent over making a poor contact).

Share this post


Link to post
Share on other sites

Dr Dave, what the cold reading resistance of the thermistor in your new sending unit?

Share this post


Link to post
Share on other sites

Dave, that's a good question about the cold resistance. That would be good to know. I guess I could check mine too without too much trouble. It's a little bit of a pain to take the stuff out of the hatch area to get to the connector, but if I get that curious, I'll do that. I bet there's a connector up under the dash I could use too if I dug into the wiring diagram.

I'm assuming your 3K is not correct. At 13V supply, that will dissipate just over 50mW. I don't think there could possibly enough internal heating at that low power level to change it's resistance. It would have to be uber sensitive, and have an almost vertical knee in the curve. I'm with you... I think you've got connection issues elsewhere.

Share this post


Link to post
Share on other sites

22 hours ago, Zed Head said:

Today, it seems like everything is handled with computer code. 

So, the ballast resistor works on the same principle as the filament in the bulb of the fuel light, I think, right?  Using that increase in resistance with current for a control purpose.

Yeah, that's very true. If I were solving this problem now, I would throw a couple hundred thousand transistors and a one time programmable chicklet at it and call it a day. In fact, that's what I did with my temperature/oil pressure warning lamp. Microcontroller reading a D/A and using digital filtering to weed out spurious false positives.

As for the ballast resistor, it has a completely different purpose. They don't want the ballast resistor to change value, they just want it to limit the current to protect the coil and whatever is pulling it to ground. The reason it's necessary is because of the way inductors work.

When you first apply voltage to an inductor, it acts like an open circuit. Infinite resistance. Won't allow any current to pass.

Then as time goes by, the resistance of the inductor goes down, and it allows current to pass. More and more and more current as time goes on. And in fact, with a theoretical perfect inductor, after enough time, it appears as a dead short. Zero resistance.

So the problem is, if there isn't something somewhere in the circuit that limits that current at the upper end, it can reach the point where it gets so high that it will damage something. The ballast resistor limits that upper end of the coil current.

You see, when you have points, you really have no idea where the points were when you shut off the motor last time. Points may be open, or points may be closed. If you turn the key on and don't crank the engine, you will quickly cook your coil or your points if you don't limit the current.

And the early electronic ignitions systems sit in a stable default state with their output transistors turned on. There is no current limit built into it at the upper end. So the same thing goes there... You could cook the coil if you don't have a ballast resistor. It wasn't until 78 that they started making the ignition modules good enough such that you didn't need the ballast resistor. I haven't played with one, but I assume they did that by building in an upper end current limit into the output stage.

Share this post


Link to post
Share on other sites
Dr Dave, what the cold reading resistance of the thermistor in your new sending unit?
I didn't check that because I didn't perform any readings with a voltmeter. I simply checked if the "warm" thermistor triggered the light and that once the thermistor was dunked in fuel the light turned off.

Sent from my SM-N950U using Tapatalk

Share this post


Link to post
Share on other sites
Posted (edited)
2 hours ago, Captain Obvious said:

Dave, that's a good question about the cold resistance. That would be good to know. I guess I could check mine too without too much trouble. It's a little bit of a pain to take the stuff out of the hatch area to get to the connector, but if I get that curious, I'll do that. I bet there's a connector up under the dash I could use too if I dug into the wiring diagram.

I'm assuming your 3K is not correct. At 13V supply, that will dissipate just over 50mW. I don't think there could possibly enough internal heating at that low power level to change it's resistance. It would have to be uber sensitive, and have an almost vertical knee in the curve. I'm with you... I think you've got connection issues elsewhere.

its is high but it does go into run away in about 5 min. I agree it does not seem like enough internal heat to boot strap it but it did work. I suppose the upside to it taking this long is you will never get a warning until the fuel level is solidly below the thermistor. My starting voltage at the lamp was about 30mV, it take a good 4-5 min to get to about 200mv with a uncooled thermistor,  from there things go pretty quick, maybe a minute to get to 2-3v  which is when the lamp becomes visible.

I am pretty sure my rear harness to lamp wiring is ok, a jumper at the harness results in full voltage at the lamp. That leaves only the thermistor itself and the connections from it to the various points on the sending unit. I should have made a ground wire to get around that clip in connection. I cant account for the connections on the hot side but they are soldered. The only other place for resistance to creep in could be the ground terminal on the sensor, not soldered but riveted to the metal shell. Perhaps there is corrosion under that rivet, creating a high resistance, It does not look bad so I assumed it was ok.

 

 

Edited by Dave WM

Share this post


Link to post
Share on other sites
On 7/25/2020 at 10:39 PM, Captain Obvious said:

@ETI4K  We were having a discussion some time ago and the million dollar question you didn't ask me is "OK, smart guy... Then if the headlights aren't an inductive load than why do the switches burn up like that? Huh? Why's that?" 

Yep, had my chance and I blew it.  So, arc suppression diodes all around!  Did you see the Combination switch teardown info I posted in Knowledge Base Electrical.  Good pics of eroded contacts.

Sorry for off topic commentary ?

 

Share this post


Link to post
Share on other sites

But that's the problem with DC non-inductive loads. I'm saying that I don't think suppression diodes are the answer. I started typing a whole bunch of stuff about inrush currents, arcs, plasma, and ionization, but worry that we're getting too far off topic and stopped. Let me know if you really want a bunch of details.

Transient suppression and mitigation is not an area that I ever specialized in (it's one of the mysterious black arts of engineering), but I just can't picture diodes being an answer for DC loads unless you are trying to deal with the flyback from an inductive load.

Sorry OP for the diversion as well. I'll try to keep my mouth shut now!   :tapemouth:

Share this post


Link to post
Share on other sites

2 hours ago, ETI4K said:

Did you see the Combination switch teardown info I posted in Knowledge Base Electrical.  Good pics of eroded contacts.

Oh, and about this. Great pics and write-up.

I went through all of this a bunch years ago and only took a couple pics. I replaced my loose turn signal pivot with a threaded in version I made on the lathe and I cleaned out all the switches and put in headlight relays. Hopefully both of us are done with issues on any of that stuff!!

Share this post


Link to post
Share on other sites

I have no problem with the temporary digression from the original topic. (I do wish I had a better understanding of electronics). I consider it an "intermezzo" while I wait on my new bulbs to arrive and I can proceed with my relatively simple experiment.

Sent from my SM-N950U using Tapatalk

Share this post


Link to post
Share on other sites

CO how about doing that back probe, I am curious what the lamp off voltage reading is on yours. Hopefully you cable connection is as easy to get at as is mine (the lamp harness to harness that goes along the back of the firewall) for mine they are just kind of stuff under the back of the center console. I do need to work on that area since a couple of the connectors have lost their locking tab (very easy for that wire to come unplugped).

Share this post


Link to post
Share on other sites
5 minutes ago, Dave WM said:

CO how about doing that back probe, I am curious what the lamp off voltage reading is on yours. Hopefully you cable connection is as easy to get at as is mine (the lamp harness to harness that goes along the back of the firewall) for mine they are just kind of stuff under the back of the center console. I do need to work on that area since a couple of the connectors have lost their locking tab (very easy for that wire to come unplugped).

My bulb plug has a 5 inch long harness that plugs into another harness above the HVAC vent. Is that where you want me to check? I have a duplicate sender out on my workbench. Is there something I can do with that sender that will provide the info you need instead?

Share this post


Link to post
Share on other sites
Posted (edited)

it can be kinda hard to hold the dvm probes on the sending unit. some jump wires (alligator clips on wires) are a handy way to grab the pins on that sending unit. Just clip onto the black wire and the yellow/blue wire, set dvm to ohms and if auto range you should get the reading. If its not auto ranging start out at the 1k setting and go up or down from there until you get a reading.

Test the setup of the meter by shorting the two leads, should be 0 ohms. always test... test equipment before relying on results.

this is to test the resistance of the sending unit on the bench. My sending unit was about 3kohms (3000 ohm).

all the above test are done with the sending unit out, no voltage required, you are just testing resistance.

 

The back probing I was talking about was aimed at capt obvious, the assumption is his setup is working so I was wondering what the key on/lamp off voltage was. Mine is about 30-50mV as long as the thermistor is in the gasoline. This test requires everything hooked up and the key in the (run) position.

(.03-.05v), starts there and takes about 5 min to get to .250v (250mV) once there it quickly goes to 6-8v  that is after its not immersed in gas (low fuel, I presume about less than 2 gallons usable fuel left in tank).

 

Edited by Dave WM
clarity
  • Like 1

Share this post


Link to post
Share on other sites

I bench tested the resistance of my sending unit with 2 different DVMs and got very similar results: 2.99k ohms and 2.98k ohms. eddb10013cbf8f8bbd999dd996e601ff.jpg&key=15ebabf9ef958a115265d953be662247d96815629904e026fb93cb99e8b28e107663dd179e65a0c868ade6277014d14f.jpg&key=f08d879fab5bdf8581eee5a7a04ee1f00e316db891325f341d48fabc86722247

 

Sent from my SM-N950U using Tapatalk

 

 

 

 

Share this post


Link to post
Share on other sites
Posted (edited)

hmmm well there you go, very close to what I got. for cold resistance.

So just that I am clear, right now with all hooked up and gas in tank say at least 1/2 way full, the light will come on after about 25 min and stay on?

If that is the case, seems like the next logical step will be to back probe, and check the lamp voltage from off to on over the 25 min period. will talk about that later, but its what I was doing in the video. You just have to find that plug, set the voltmeter to volts, and insert in the back so the probes contract the inside of the plug. I am assuming starting with the "2000m" to see if you get a reading if not switch down to "200m"

Edited by Dave WM

Share this post


Link to post
Share on other sites

Well OK. I stand corrected, 50mW IS enough power to heat that thing up!  LOL

Dave, I'll check my cold resistance too when I get a chance just for another data point.

  • Like 1

Share this post


Link to post
Share on other sites
21 hours ago, Captain Obvious said:

Let me know if you really want a bunch of details.

Absolutely!

Share this post


Link to post
Share on other sites
hmmm well there you go, very close to what I got. for cold resistance.
So just that I am clear, right now with all hooked up and gas in tank say at least 1/2 way full, the light will come on after about 25 min and stay on?
If that is the case, seems like the next logical step will be to back probe, and check the lamp voltage from off to on over the 25 min period. will talk about that later, but its what I was doing in the video. You just have to find that plug, set the voltmeter to volts, and insert in the back so the probes contract the inside of the plug. I am assuming starting with the "2000m" to see if you get a reading if not switch down to "200m"
The malfunction is exactly as you describe it.
As far as testing voltage at the lamp plug, sounds good and I think I understand it.
Do you still want me to try all the different higher resistance bulbs with simple 25 minute runs to see which ones go on with a full tank?

Sent from my SM-N950U using Tapatalk

Share this post


Link to post
Share on other sites

14 minutes ago, dmorales-bello said:

The malfunction is exactly as you describe it.
As far as testing voltage at the lamp plug, sounds good and I think I understand it.
Do you still want me to try all the different higher resistance bulbs with simple 25 minute runs to see which ones go on with a full tank?

Sent from my SM-N950U using Tapatalk
 

I think it would be a good idea to try the back probe the connector at the center console. This is the white connector that is on the lamp harness wire, and the connector that is part of the wire harness that runs along the firewall. Its the one you have to undo (along with a bunch of other) that when you are pulling the center console off. You know you have the right connector when if you UNPLUG it then back prob it should be 12v with the key in the "run" position. Its 12v since there is no lamp load plugged in. The 3k ohm resistance will not drop any voltage since the circuit is open (no lamp, and the meter is not enough of a load to effect it). As soon as you plug the lamp in with the probs still in place the lamp load will cause a current to flow and the voltage will drop due to the high (3k) resistance. The point to all this is to try and duplicate a known working system (mine). so far the resistance of the thermistor checks out the same, so I am curious to see if you have the same starting voltage (lamp load and battery will effect that), and then watch it to see how fast it ramps up. I will pm you my phone number if you want to talk thru it when you are ready to test.

Share this post


Link to post
Share on other sites

here is a thought

is it possible the thermistor in the can is not actually coming into contact with the gas? on my thermistor the can has holes in it to allow fuel to enter and exit the can. IF for some reason this was not the case in the new sender, then the cooling effect would be lost. You should see bubbles come from the can when inserted into a fluid, and likewise fluid should drain out when removed. Not a lot mind you, but something showing the gasoline is actually coming into contact with the thermistor.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use. We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.