Hi, I've been looking for weeks and its doing my head in!
All I'm after is a VERY simple circuit, just one led that lights up when the fuel gets low on my bike. (making a custom dash using the "Veypor")

I'm usually pretty good with electronics ect, but i need some help with this one!
I've been searching for ages and only come across crazy complex (for a noob) circuits with time delays and 100,000 led light bars!

Is this anywhere near a working cercuit?!

What i've "tried" to do is have the led switch on when the fuel drops below 90Ω this should be @ 6.2v that the zener is at. So when the fuel drops the zener has more voltage and switches the led on.

Dont realy need any sort of 'time delay' to combat fuel sloshing around and continualy ficking the light on and off. This is because as it stand the fuel guage sits at 'full' 90% of the time anyway, its not till the last 30km or so that needle moves down. (and when it does its a rapid decent!)

Also any feedback on the resistor ratings would be very apriciated.

Cheers Steve.


Click on pic below

View attachment 11998

I like your design. I designed a universal flashing LED in 1975 and has been used on various vehicles for 4 decades.
My design self-adjusts to various battery or alternator voltages. It can be used on different gas tank voltage wiring as shown.
Some vehicles use an empty tank signal at full voltage. Some vehicles use a full tank signal at full voltage. Just flip pins 2 & 3 as needed.
I don't have a flasher circuit. The LED flashes in real time when any gasoline wave forms in the gas tank and bumps the float up or it goes down, the LED changes its lit state.
If you drop below the variable voltage barrier or trip point voltage, the LED stays on without any waves, like if parked.

If I am on 'fumes' in a remote desert (yes it happened in 1986), I tap my brakes to create a wave and I see the always ON LED indicating low fuel, but a wave will turn it OFF as if it was 'flashing'. That ability to flash in real time on fumes indicates I have a quart or two in my gas tank. Anything that causes a wave peak and trough will make the LED flash in real time. To be clear, any wave peak makes the LED turn OFF. The trough or wave valley makes it turn ON.
This circuit can be adapted to any vehicle or motorcycle with a flaot that sends a voltage to any gas guage. I am not sure about series dual gas tank setups like a Subaru uses.

The input resistance of the 741 op-amp with a single voltage supply, is of the order of 2+ megohms. It will not load down any voltage from the gas tank circuit or float setting. I measured one 741 at 5+ megohms. It is essentially a vacuum tube voltmeter, a VTVM, high impedance perfect potentiometer knock off.

My Universal Real Time Low Fuel Indicator LED Circuit from 1975. The image is a bit dated but the information is there.

One problem was the puny 21 millicandles (mcd) of old LEDs. I had to over drive old LEDs with a short duty cycle to get it brighter. That shortens an LED lifetime.
This problem was fixed and changed recently by using Super Bright LEDs of the 28,000 mcd or 34,000 mcd type.
These super brights have a voltage drop of 3.1 - 3.2 volts. At 10 ma, you WILL see these flash in bright sunlight.
I use this circuit in duplicate to flash either turn signal direction inmy vehicle and my LED indicates 'an in-use reminder' because a newer small relay system like a relay system from Denso, et al, are just too quiet with a radio on and a window down.

So, the 220 ohm resistor should be about 1K to 4K. To run higher currents than the 741 can sink, you will need to install a 2n2222 type transistor at the 'bottom' of the LED and put the emitter on ground. The 741 output pin 6 should have enough voltage to turn on the NPN base with a positive voltage.
I use 4.5 K. Yes, the current is lower through the new LED but knocking it down from 30,000 mcd to 20,000 or 5,000 mcd is going to be seen in bright sunlight anyway.
The problem is the 741 datasheets just say the Short Circuit current is typically 25 ma. Well, how about the typical test circuit or recommended clearly stated sinking current? Be safe and add the NPN transistor. They are cheap and general replacement for almost any NPN transistor. ECG123A/NTE123A, 2N2222, PN2222, MSPA42, etc. all work. If you can't turn the NPN transistor totally OFF and put it into cutoff, add a 1N915 diode between the emitter and ground to cheaply raise ground. I think the 741 will sink 20-25 ma but I have never seen that stated as typical. Never short pin 6 directly to ground on a 741. I will be a goner.

Your circuit is just about the same but I wanted to show and explain how mine actually works for some readers. Most of this applies to how your circuit that works with the Zener.

HTH, and JMO,

Andre

 

Rather than using an old operational amplifier like the LM741 as a comparator I would likely go with a LM339 Comparator or similar chip. Not to say a LM741 won't work as it certainly will but I see a comparator as better suited for the application. Keep in mind the voltage on a bike is likely not a constant 12 volts DC but since the basic circuit is ratio metric I doubt it will matter. As to any fuel sloshing around consider adding some simple hysteresis in your comparator circuit. There are several ways to get where you want to go but a basic comparator with hysteresis is likely the simplest route to take.

Ron

 

Since you say that you have a existing working Fuel-Gauge,
there is already a built-in resistor in the Bikes wiring,
so you don't need to add a second one.

Treat the output of your Fuel-Level-Sender as a varying Voltage-Source which you will monitor.

You do not want, or need to, use a Zener-Diode, as the Bike's Electrical System is not very stable.

Since you may be measuring Voltages which are very close to Ground or Battery-Voltage,
you should use an Op-Amp that does not have any troubles with operating close to "the-Rails",
( the "Rails" meaning Ground or Supply Voltage ).

For the most consistent drive Current for your chosen LED, you should use a
Current-Limiter instead of a Resistor, which is specifically designed to regulate LED Current.

LM741 Op-Amps are really ancient and have lame performance,
I suggest using a more modern Op-Amp with "Rail-to Rail" Inputs and Outputs.
This will allow you to measure all the way down to zero-Volts without any weirdness going on.

 

@andrejarret
Welcome to AAC, but do you realise you responded to a 12-year-old thread? This is gnerally not good practice as the original contributors may well no longer be interested in the topic. It is best to start your own thread if you have a query or suggestion.

 

I would think that the age of the thread doesn't necessarily mean
that it's dead, and of no use anymore.
Sure, the original Thread Starter has probably moved on,
but this thread may come up in thousands of searches looking for a similar subject
and may be delivering service for years to come.

I found this Forum, and joined it, after doing a search years ago,
and it was an "out-dated" "dead" thread at that time.

He may be talking to a ghost, but he'll figure that out after a while.
( I never checked the Date either

 

Is Necro Posting ever a good thing is a good read on the subject. What generally happens is an old thread is used for spam. A kindly moderator removes the spam but the spam post moves the thread back up to the top. Sometimes there is merit to adding to an old post and sometimes not. Maybe if when moderation removes spam they copy and paste an annotation to that effect calling attention to the threads age. I have been sucked in several times including this thread.

Ron

 

Jumping to the end to make this statement: Use much higher value resistors. A 91Ω resistor at 12 volts will need to dissipate 1.6 watts of energy. You'll need at least 2 watt resistors at that level of power consumption. Using a 91KΩ resistor (no such specific standard values exist at 91Ω or any value 10 or more times higher), but using a 91KΩ resistor will dissipate 1.6mW of energy. MUCH more efficient.

When using a comparator, you set a specific value at one end (negative). When the positive value falls below that setting the OpAmp will switch the LED. Since you're not asking for a level sensing and indicating circuit - you only want to know when you reach some specified quantity (trial and error method) a simple circuit such as has been offered will suffice. Dual OpAmps - nice, but not necessary for your stated desire.

Necro-posts? It happens to us all.

 

Jumping to the end to make this statement: Use much higher value resistors. A 91Ω resistor at 12 volts will need to dissipate 1.6 watts of energy. You'll need at least 2 watt resistors at that level of power consumption. Using a 91KΩ resistor (no such specific standard values exist at 91Ω or any value 10 or more times higher), but using a 91KΩ resistor will dissipate 1.6mW of energy. MUCH more efficient.

When using a comparator, you set a specific value at one end (negative). When the positive value falls below that setting the OpAmp will switch the LED. Since you're not asking for a level sensing and indicating circuit - you only want to know when you reach some specified quantity (trial and error method) a simple circuit such as has been offered will suffice. Dual OpAmps - nice, but not necessary for your stated desire.

Necro-posts? It happens to us all.

The Bike already has it's own Resistor, because it already has a working Fuel-Level-Gauge.

 

The Bike already has it's own Resistor, because it already has a working Fuel-Level-Gauge.

So you just want something that will catch your attention when the tank is low. Is that a safe assumption?

 

It was 12 years ago ......
You gotta read the whole thread for context.

 

I would think that the age of the thread doesn't necessarily mean
that it's dead, and of no use anymore.
Sure, the original Thread Starter has probably moved on,
but this thread may come up in thousands of searches looking for a similar subject
and may be delivering service for years to come.

I found this Forum, and joined it, after doing a search years ago,
and it was an "out-dated" "dead" thread at that time.

He may be talking to a ghost, but he'll figure that out after a while.
( I never checked the Date either

I was looking around for some idea's on this subject today and found some starter points here. the tread is not dead. Thanks. Jv