Perplexing "FUEL" light malfunction

[dmorales-bello]

1 hour ago, Patcon said:

Maybe ambient air temperature is contributing. Warm fuel and all...

I thought about that as well but it's been 94 degrees at most. The same thing has happened during a morning drive when the temperature was 84 degrees. Not what I would think of as extreme temps.

[dmorales-bello]

22 hours ago, Dave WM said:

Ok I have to ask, you are sure you have the mount wire correct and the thermistor is in the tank such that its near the bottom?

Next check the function of the plug by shorting out the wire harness, there is a common neg lead, connecting this to one lead will turn on the fuel light the other will peg the fuel.

The last thing you can test would be to actually take out the sending unit, hook it up and dip the thermistor in a glass of water, make sure everything works as it should while you can observe the effect of the thermistor in and out of the water.

 

I'm not sure what you mean by having "the mount wire correct". The factory connector only fits one way.

The whole unit fits only one way and straight into the tank which was full of fuel so yes, the thermistor must be in the fuel. As a matter of fact, with the unit plugged into the harness but not installed in the tank, the thermistor triggered the "FUEL" light on. It went out as soon as I introduced the sender into the tank as the thermistor hit the fuel.

I will run your suggested check by plugging in the connector with the sending unit out of the tank and putting the thermistor in water (alcohol might be better perhaps) and also shorting the leads to the common negative wire to check function.

Is there a chance that perhaps the rating on the thermistor is wrong and therefore triggers the light while fuel warms up a bit from the heat outside plus whatever other heat sources could be involved (engine?, exhaust? etc...)??

Edited July 24, 2020 by dmorales-bello

[Captain Obvious]

I got no silver bullet. The only explanation I have is what you already mentioned... Sounds like the thermistor is not quiiiiite right for the application. CLOSE, but not enough headroom for tolerances and environmental stressors like ambient temp and system voltage.

And as Zed mentioned, why it only happens when the engine is running is probably because the system voltage is higher then. That extra volt-n-a-half matters.

Thermistors come in all sorts of base resistances and with all shapes different curves. For this application, you want one with a "knee" in the curve as opposed to linear, and you want the knee to be at the correct temperature. I'm thinking that it's too close to the limit most of the time, and the combination of ambient temperature and system voltage (when the engine is running) puts it over the edge.

@Dave WM, Did you find any specs at all about the thermistor? Anything at all? I didn't look at any of the FSM's, but I bet you did.

I wonder what thermistor ZCD picked for their unit. And why.

[dmorales-bello]

Thanks for your input,@captain obvious. It's good to see we're all going in the same direction. It's also clear that if we're correct, there seems to be no easy fix unless a properly calibrated thermistor replacement can be sourced.

 

Sent from my SM-N950U using Tapatalk

 

 

 

 

[SteveJ]

8 hours ago, Captain Obvious said:

I got no silver bullet. The only explanation I have is what you already mentioned... Sounds like the thermistor is not quiiiiite right for the application. CLOSE, but not enough headroom for tolerances and environmental stressors like ambient temp and system voltage.

And as Zed mentioned, why it only happens when the engine is running is probably because the system voltage is higher then. That extra volt-n-a-half matters.

Thermistors come in all sorts of base resistances and with all shapes different curves. For this application, you want one with a "knee" in the curve as opposed to linear, and you want the knee to be at the correct temperature. I'm thinking that it's too close to the limit most of the time, and the combination of ambient temperature and system voltage (when the engine is running) puts it over the edge.

@Dave WM, Did you find any specs at all about the thermistor? Anything at all? I didn't look at any of the FSM's, but I bet you did.

I wonder what thermistor ZCD picked for their unit. And why.

See, now you made me go look. The FSM doesn't have much in the FE section, and the BE section refers to the FE section.

I asked my frenemy Google about this and found a couple of interesting threads in other forums.

https://www.element14.com/community/thread/56806/l/which-thermister?displayFullThread=true

https://advrider.com/f/threads/honda-thermistor-fuel-light-to-led.954021/

As was mentioned earlier, the wattage of the bulb and characteristics of the thermistor come in to play. (Thanks for the re-hash, Steve. Do you have anything to add to the conversation?)

I tried a search to see if one could buy a fuel light thermistor. Sure enough, there are some listings on ebay for some Chinesium thermistors. I'm not sure I would want to experiment with incorporating one into an existing fuel sending unit. 

https://www.ebay.com/itm/Automotive-Fuel-Level-Sensor-Fuel-Pump-Alarm-Sensor-NTC-Thermistor-/283606453547?_trksid=p2385738.m4383.l4275.c10

 

 

[Dave WM]

"The whole unit fits only one way and straight into the tank which was full of fuel so yes, the thermistor must be in the fuel. As a matter of fact, with the unit plugged into the harness but not installed in the tank, the thermistor triggered the "FUEL" light on. It went out as soon as I introduced the sender into the tank as the thermistor hit the fuel"

Since the above was done, it pretty much means everything is working electrically. I mentioned the mount cause on mine its a side entry on the tank, and the mount wire can be bent up or down to adjust when the sensor is in the fuel. I cant see any other explanation other than an intermittent short in the harness and/or sensor wiring  or a misplaced sensor. Maybe perform the test again with the engine running to confirm its not a voltage issue (higher voltage from alt running). Thing about testing is you want as few variables as possible.

Edited July 24, 2020 by Dave WM

[Captain Obvious]

2 hours ago, SteveJ said:

See, now you made me go look.

Ha!! Made you look!

So the only thing they can muster is "It's resistance is different". Not much help!

The whole thing is a balancing act... The thermistor in the tank is a NTC and the bulb filament is a PTC. You need the current through the thermistor high enough to achieve the desired power dissipation such that there is self heating when not submerged in fuel, but never enough power to damage it.

And when the thermistor does heat up and it's resistance drops, you need the resistance low enough to get the bulb filament to glow.

[Captain Obvious]

16 hours ago, dmorales-bello said:

Here's a pic of the bulb. It reads 12V 4W 3E. Does that seem right?

I had originally thought there was simply some dyslexia going on here, but now I'm not so sure... The original bulb spec is 3.4W, but the one in your pic (and part number) appears to be 4.3W instead. If that bulb really is 4.3 instead of 3.4, that might be part of the problem.

Do you have another (known to be correct) bulb you can toss in there just to see what happens?

My original bulbs were Toshiba A12V3.4.

[SteveJ]

29 minutes ago, Captain Obvious said:

I had originally thought there was simply some dyslexia going on here, but now I'm not so sure... The original bulb spec is 3.4W, but the one in your pic (and part number) appears to be 4.3W instead. If that bulb really is 4.3 instead of 3.4, that might be part of the problem.

Do you have another (known to be correct) bulb you can toss in there just to see what happens?

My original bulbs were Toshiba A12V3.4.

I was wondering about the bulb wattage, too, based upon one of the links I posted.

So @dmorales-bello, are you game to try a different bulb?

[dmorales-bello]

@Captain Obvious, @SteveJ , I'm definitely open to try any experiment that might clear this up. I must first admit I made a mistake with the bulb identification. I assumed the other bulb I had in my "bulb drawer" was the same as what was in the "FUEL" light housing because they looked the same. In fact, when I pulled the actual bulb form the housing it was a Toshiba branded 12V 3.4 amp bulb. In the "bulb drawer" I not only have the 4.3 watt bulb I photographed, I also have a 6 watt bulb of the same size. So I will now experiment with all 3 bulbs and report.

Question: In theory, is the wattage of the bulb reversely related to the amount of current needed to illuminate it?

 

Edited July 24, 2020 by dmorales-bello

[SteveJ]

2 minutes ago, dmorales-bello said:

@Captain Obvious, @SteveJ , I'm definitely open to try any experiment that might clear this up. I must first admit I made a mistake with the bulb identification. I assumed the other bulb I had in my "bulb drawer" was the same as what was in the "FUEL" light housing because they looked the same. In fact, when I pulled the actual bulb form the housing it is a Toshiba branded 12V 3.4 amp bulb. In the "bulb drawer" I not only have the 4.3 watt bulb I photographed, I also have a 6 watt bulb of the same size. So I will now experiment with all 3 bulbs and report.

Question: In theory, is the wattage of the bulb reversely related to the amount of current needed to illuminate it?

 

Wattage is proportional to current. It is also proportional to voltage.

[SteveJ]

If you want to experiment further, here is a site with different wattage bulbs: https://www.bulbtown.com/9mm_Diameter_Bayonet_Base_s/1421.htm