Glow Fuel Sensor

[erwabo]

So, Im wondering if anyone remembers their basic electronics, cuz im a little rusty. Here is what Im doing. I have a voltage sensor for my D-series FrSKY receiver. I have 2 goals here :

1. Light up an LED when the fuel tank is full (so I don't have to count turns or get fuel everywhere)
2. Send me an alert on the Transmitter when fuel is low.

I built everything the way I wanted and when I went to test....goal 1 failed miserably and goal 2 worked( barely) When I checked the readings I realized that I didnt take into consideration the resistance of the glow fuel itself. Im using a 7.4v 200mA Lipo to power the LED and I have 270ohm resistor inline (so I don't blow LED)...which I found out I don't need due to the massive about of resistance going through the fuel. My Voltmeter has a single Ohm setting with 1K multiplier readout, and from what I could tell, there was about 2000 ohms and probably much more resistance in the fuel...but Id rather rely on the voltage sensor as I know very little about reading ohm meters. As I drained the fuel out, I noticed my voltage stepped down as I drained the fuel which was a pleasant surprise as I can use the data for a gauge with LUA scripts.

That being said...I need to know what Voltage and Amperage to send through this tank of fuel to generate 1.5v @ 20ma to light my LED. Right now Im seeing .5v passing through the nearly full tank...give or take .05v Some simple formulas Im sure, but I have had enough trial and error for 24 hours. Not sure if they have a battery that meets my needs, Any help would be appreciated it

Thanks so much

 

[mjmccarron]

Hi erwabo,

I don't think that glow fuel is conductive enough to light an LED through it without an amplifier circuit. I googled some images of touch switch circuits http://talkingelectronics.com/projects/TouchSwitch/TouchSwitch-1.html The first one is a simple darlington pair that should work. You could use a single MPSA29 transistor in place of the two or a pair of 2N2222's Then adjust the 100K resistor for the sensitivity (100K pot). Also, change the 470 ohm to 180 ohm to drive your LED from 7.4v at 20mA. I=E/R (Current = voltage/resistance)

Good luck,
Mike

 

[erwabo]

Hmm, not sure exactly how I would put that inline with my setup, you would probably need to see a diagram of what I have. Also I was wondering if a 7.4v 1000ma battery would work or even a 7.4v 2000ma. My voltage sensor is looking for a 2s lipo so that's the reason I ask. I don't care if it passes more than 1.5v, I can add a resistor if need be. Like I said, im getting about .5-.6v from the 20ma, I would think 1000ma or more would give me 1.5v...but all this is a bit above my level of knowledge.

 

[erwabo]

Here is my initial design (when for some reason did not even consider the fuel's resistance)

 

[Craftydan]

erwabo,

Unfortunately, it doesn't work quite that way. Increasing the mAh of the battery pack only increases the capacity of energy stored, but doesn't begin to change how that energy is delivered. Your circuit of fuel + bias resistor +unlit LED will draw just as much current from a 1000mA battery as a 10mAh battery . . . it will just draw it longer from the bigger pack.

̶R̶u̶n̶n̶i̶n̶g̶ ̶a̶ ̶q̶u̶i̶c̶k̶ ̶a̶n̶a̶l̶y̶s̶i̶s̶,̶ ̶s̶o̶m̶e̶ ̶o̶f̶ ̶t̶h̶e̶ ̶n̶u̶m̶b̶e̶r̶s̶ ̶a̶r̶e̶n̶'̶t̶ ̶l̶i̶n̶i̶n̶g̶ ̶u̶p̶ ̶.̶ ̶.̶ ̶.̶ ̶Y̶o̶u̶ ̶s̶a̶y̶ ̶t̶h̶e̶ ̶f̶u̶e̶l̶ ̶m̶e̶a̶s̶u̶r̶e̶d̶ ̶r̶o̶u̶g̶h̶l̶y̶ ̶2̶0̶0̶ ̶o̶h̶m̶ ̶(̶I̶'̶m̶ ̶a̶s̶s̶u̶m̶i̶n̶g̶ ̶t̶h̶a̶t̶'̶s̶ ̶w̶i̶t̶h̶ ̶n̶o̶t̶h̶i̶n̶g̶ ̶e̶l̶s̶e̶ ̶c̶o̶n̶n̶e̶c̶t̶e̶d̶ ̶a̶n̶d̶ ̶a̶ ̶f̶u̶l̶l̶ ̶f̶u̶e̶l̶ ̶t̶a̶n̶k&#822̶ ̶a̶n̶d̶ ̶y̶o̶u̶'̶r̶e̶ ̶u̶s̶i̶n̶g̶ ̶a̶ ̶b̶i̶a̶s̶ ̶r̶e̶s̶i̶s̶t̶a̶n̶c̶e̶ ̶o̶f̶ ̶2̶7̶0̶ ̶o̶h̶m̶ ̶a̶n̶d̶ ̶a̶ ̶7̶.̶4̶v̶ ̶p̶a̶c̶k̶ ̶.̶ ̶.̶ ̶.̶ ̶a̶l̶l̶ ̶c̶o̶n̶n̶e̶c̶t̶e̶d̶ ̶i̶n̶ ̶s̶e̶r̶i̶e̶s̶ ̶w̶i̶t̶h̶ ̶a̶ ̶r̶e̶d̶ ̶L̶E̶D̶,̶ ̶a̶n̶d̶ ̶I̶'̶m̶ ̶a̶s̶s̶u̶m̶i̶n̶g̶ ̶y̶o̶u̶'̶r̶e̶ ̶c̶o̶n̶n̶e̶c̶t̶i̶n̶g̶ ̶y̶o̶u̶r̶ ̶D̶-̶s̶e̶r̶i̶e̶s̶ ̶v̶o̶l̶t̶a̶g̶e̶ ̶a̶c̶r̶o̶s̶s̶ ̶t̶h̶e̶ ̶f̶u̶e̶l̶ ̶t̶a̶n̶k̶.̶ ̶ ̶ ̶
̶
̶R̶u̶n̶n̶i̶n̶g̶ ̶t̶h̶e̶ ̶n̶u̶m̶b̶e̶r̶s̶,̶ ̶t̶y̶p̶i̶c̶a̶l̶ ̶v̶o̶l̶t̶a̶g̶e̶ ̶d̶r̶o̶p̶ ̶a̶c̶r̶o̶s̶s̶ ̶a̶ ̶r̶e̶d̶ ̶L̶E̶D̶ ̶i̶s̶ ̶a̶b̶o̶u̶t̶ ̶1̶.̶8̶v̶,̶ ̶s̶o̶ ̶t̶h̶a̶t̶ ̶l̶e̶a̶v̶e̶s̶ ̶5̶.̶6̶v̶ ̶(̶7̶.̶4̶-̶1̶.̶8̶=̶5̶.̶6&#822̶ ̶t̶o̶ ̶d̶r̶o̶p̶ ̶a̶c̶r̶o̶s̶s̶ ̶y̶o̶u̶r̶ ̶b̶i̶a̶s̶ ̶a̶n̶d̶ ̶f̶u̶e̶l̶.̶ ̶ ̶U̶s̶i̶n̶g̶ ̶v̶o̶l̶t̶a̶g̶e̶ ̶d̶i̶v̶i̶s̶i̶o̶n̶,̶ ̶a̶ ̶2̶0̶0̶ ̶o̶h̶m̶ ̶r̶e̶s̶i̶s̶t̶a̶n̶c̶e̶ ̶w̶i̶t̶h̶ ̶a̶ ̶2̶8̶0̶ ̶o̶h̶m̶ ̶b̶i̶a̶s̶ ̶w̶i̶l̶l̶ ̶d̶r̶o̶p̶ ̶4̶2̶%̶ ̶o̶f̶ ̶t̶h̶e̶ ̶v̶o̶l̶t̶a̶g̶e̶,̶ ̶o̶r̶ ̶2̶.̶3̶v̶o̶l̶t̶s̶ ̶-̶-̶ ̶t̶h̶a̶t̶'̶s̶ ̶n̶o̶w̶h̶e̶r̶e̶ ̶n̶e̶a̶r̶ ̶t̶h̶e̶ ̶0̶.̶5̶v̶ ̶y̶o̶u̶ ̶r̶e̶p̶o̶r̶t̶.̶ ̶ ̶
̶
̶I̶f̶ ̶t̶h̶e̶ ̶f̶u̶e̶l̶ ̶t̶r̶u̶l̶y̶ ̶i̶s̶ ̶a̶r̶o̶u̶n̶d̶ ̶2̶0̶0̶ ̶o̶h̶m̶ ̶w̶h̶e̶n̶ ̶f̶u̶l̶l̶,̶ ̶t̶h̶e̶ ̶5̶.̶6̶v̶ ̶l̶e̶f̶t̶ ̶o̶f̶ ̶t̶h̶e̶ ̶7̶.̶4̶v̶ ̶b̶a̶t̶t̶e̶r̶y̶ ̶s̶h̶o̶u̶l̶d̶ ̶b̶e̶ ̶m̶o̶r̶e̶ ̶t̶h̶a̶n̶ ̶e̶n̶o̶u̶g̶h̶ ̶t̶o̶ ̶t̶r̶i̶p̶ ̶t̶h̶e̶ ̶2̶0̶m̶A̶ ̶c̶u̶r̶r̶e̶n̶t̶ ̶t̶h̶e̶ ̶L̶E̶D̶ ̶n̶e̶e̶d̶s̶ ̶t̶o̶ ̶l̶i̶g̶h̶t̶ ̶(̶w̶i̶t̶h̶o̶u̶t̶ ̶b̶i̶a̶s̶,̶ ̶2̶0̶0̶ ̶o̶h̶m̶ ̶w̶i̶l̶l̶ ̶d̶r̶a̶w̶ ̶r̶o̶u̶g̶h̶l̶y̶ ̶2̶8̶m̶A&#822̶.̶ ̶ ̶
̶
̶P̶e̶r̶h̶a̶p̶s̶ ̶a̶ ̶q̶u̶i̶c̶k̶ ̶s̶k̶e̶t̶c̶h̶ ̶i̶s̶ ̶i̶n̶ ̶o̶r̶d̶e̶r̶?̶ ̶ ̶P̶e̶r̶h̶a̶p̶s̶ ̶a̶ ̶f̶e̶w̶ ̶p̶i̶c̶s̶?̶ ̶I̶ ̶c̶a̶n̶ ̶s̶e̶e̶ ̶w̶i̶t̶h̶ ̶t̶h̶o̶s̶e̶ ̶n̶u̶m̶b̶e̶r̶s̶ ̶w̶h̶y̶ ̶t̶h̶e̶ ̶L̶E̶D̶ ̶w̶o̶u̶l̶d̶n̶'̶t̶ ̶l̶i̶g̶h̶t̶,̶ ̶b̶u̶t̶ ̶n̶o̶t̶ ̶w̶h̶y̶ ̶t̶h̶e̶ ̶t̶e̶l̶e̶m̶e̶t̶r̶y̶ ̶i̶s̶ ̶r̶e̶a̶d̶i̶n̶g̶ ̶s̶o̶ ̶v̶e̶r̶y̶ ̶l̶o̶w̶.̶ ̶ ̶s̶t̶r̶a̶i̶g̶h̶t̶e̶n̶i̶n̶g̶ ̶o̶u̶t̶ ̶t̶h̶a̶t̶ ̶m̶a̶y̶ ̶t̶e̶l̶l̶ ̶w̶h̶e̶t̶h̶e̶r̶ ̶y̶o̶u̶ ̶j̶u̶s̶t̶ ̶n̶e̶e̶d̶ ̶t̶o̶ ̶r̶e̶d̶u̶c̶e̶ ̶y̶o̶u̶r̶ ̶b̶i̶a̶s̶ ̶c̶i̶r̶c̶u̶i̶t̶ ̶.̶ ̶.̶ ̶.̶ ̶o̶r̶ ̶a̶t̶t̶a̶c̶h̶ ̶a̶n̶o̶t̶h̶e̶r̶ ̶c̶i̶r̶c̶u̶i̶t̶ ̶p̶a̶r̶a̶l̶l̶e̶l̶ ̶t̶o̶ ̶y̶o̶u̶r̶ ̶t̶e̶l̶e̶m̶e̶t̶r̶y̶ ̶t̶o̶ ̶t̶u̶r̶n̶ ̶o̶n̶ ̶a̶n̶ ̶L̶E̶D̶.̶


Edit:

OK, Nevermind. Thanks for the drawing.

remove the 270 ohm resistor and replace with a 100 Ohm Potentiometer. connect the wiper to the "full" fuel sensor and one leg to the LED. fill the tank and turn the pot until the LED lights.


*** KEEP IN MIND ***

As your battery drains, the voltage drops and the LED might no longer light because the battery voltage has dropped.

 

[erwabo]

Well that was my original design, once I realized the fuel had so much resistance I bypassed the 270ohm resistor and just connect directly to the leads. That's when I was only seeing .5v coming through, so forgive my ignorance, but what would the 100ohm potentiometer bring into the equation? Also I looked up how to use my multimeter ohm setting....and if understand correctly, the reading said 200 with a 1k multiplier so that's 200,000k.....not sure I can get another volt through there haha

I guess my question would be then...if I used an 11.1v 2000ma Lipo or even a 9volt @400ma would that get me an extra volt through there? My voltage sensor is looking for a "lipo" supposedly and I have to solder the hot wire to 1s, 2s, or 3s...but Im thinking that may just be for voltage detection...since it doesn't read the balance wire...so its POSSIBLE a 9v battery could be used

 

[Craftydan]

Ok, 200kohm makes MUCH more sense.

For that kind of resistance to sink 20mA, you'll need 800,000,000v.

Wanna see what sparks do to fuel inside a tank?

Yeah, that's a no-go. the 100ohm potentiometer would allow you to put a "less than 100 ohm" bias resistor in place of the 280 ohm resistor, but that was when I believed your measured resistance was in the "workable" region. Not the case, so not gonna work.

you're back to needing something to read the value and respond (which is what Mike was proposing) as opposed to using the fuel itself as the switch. Looking at Circuit A from the link he provided, the lower of the two touch pads in the diagram connects to your "full" sensor contact. Your battery voltage (top touch pad in the diagram) is already connected at the bottom of the tank. Use the parts he's speced and it should work fine on your 2 cell lipo.

 

[erwabo]

Well I have a nice digital ohm meter here at work that im gonna take home today to get the ACTUAL readings, but based on the numbers you are seeing does 200k ohms sound about right and if that's the case, how is .6 volts making it through?

 

[erwabo]

So, yeah im sure this isn't right, but if its not please tell me the correct placement and wiring. I don't know what a mpsa29 transistor looks like and if it has 2 or 3 leads =) In this diagram I used the one that kept popping up the most with 2 leads though its more fit for a PCB

 

[Craftydan]

*cue hand waving explanation*

The voltage sensor you've mounted probably has a high impedance (doesn't sink much current to take the measurement), but it's in the same neighborhood as the fuel's resistance. Since you have the fuel connected to power on one side, and your sensor on the other, the "ground" to complete the circuit is on the other side of the sensor -- it flows through the input impedance. There's another current loop in parallel between the fuel and the LED, but since that loop has fuel between it and the voltage sensor it's probably not impacting the local voltage around the your voltage probe strongly.

so . . . with the circuit of battery -> fuel -> voltage sensor input impedance, dropping 7.2v to 0.6v over a 200kohm fuel resistance . . . run the numbers and it looks like 15k ohm input impedance. That seems a bit low for a high quality sensor, but not outrageously so. There's also a fair chance the resistance you measured at the LED connection isn't the same as the voltage sensor's connection . . . which would change those numbers.

 

[erwabo]

So does that new diagram with the transistor and 100k pot look correct or is that totally wrong haha?

Also I was thinking there may be some resistance on the material is I used for leads. They are made from control surface push rods...which are about 1-2mils thick...they are pretty short but im sure they are not helping. I ll have to measure that resistance as well.

 

[Craftydan]

So, yeah im sure this isn't right, but if its not please tell me the correct placement and wiring. I don't know what a mpsa29 transistor looks like and if it has 2 or 3 leads =) In this diagram I used the one that kept popping up the most with 2 leads though its more fit for a PCB

Ok . . . sorry for the butchery . . . try this diagram:



The three connections on the mpsa29 are called Base, Collector and Emitter (abbreviated B,C, and E on the diagram). in the metal TO-3 cases, the casing itself becomes the collector:

 

[erwabo]

So does it matter which transistor I use since they have different wattages? I prefer the 3 prong one over the case. Also If I get a more accurate resistance reading, will that make a difference in what transistor is used and if not should a POT be put in line as suggested in that other post? I just cant believe there is 200,000k in the fuel, gotta be something else. Water is only. .6 and and Vodka is 1.2 and this fuel is 15% nitromethane and the rest is methanol and a neutral lubricant. Perhaps I need copper posts, but I was just worried about corrosion ( though I do a fuel filter in there)

 

[Craftydan]

Actually it doesn't matter. This circuit just needs the resistance between the touch pads to drop significantly below the resistance of air (which is A LOT). 200kOhm is way below that, and should trip the circuit nicely.

As for comparative resistances . . . in bulk fluids it's really over the volume (change the tank or probe positions, you change the resistances) and it greatly depends on purity. DI water clocks in around 18 million ohm-cm, where typical clean drinking water clocks in at 2kOhm-CM and sea water is around 20 ohms-cm . . . pretty strong extremes. 200k for a refined fuel doesn't seem outrageous, but it also depends on how much fuel is between the probes.

As for power, 20mA over 7.4v is 15mW. You can pick the smallest transistor of that type you can find . . . but I'd stay away form surface mount unless your soldering kung foo is strong.


By the way . . . the unlabeled resistor in my butchered diagram should be roughly 100kOhm (close will do here), and I forgot to add the bias resistor for the LED -- you can put the 180 ohm resistor Mike spec'ed in line with either side of the LED, but the LED will need a resistor in series with it.

 

[erwabo]

Okay, so how does this look? If this is right, I will try this

 

[Craftydan]

Double check your pinouts on the transistor as you wire it in, but yeah, looks good.

 

[erwabo]

Thank a ton for the help! I really appreciate it. Im still going to try a couple more things for fun while waiting on those parts. Radio Shack as limited components...so have to order online.

SO...HOW does that little tiny transistor negate all of this resistance or amplify the voltage? Im not understanding the science of that little tiny part.

 

[Craftydan]

At it's heart, a transistor is an amplifier -- a little current flows into the base, then a LOT of current will be allowed to flow into the collector (all are "emitted" out of the emitter) . . . but there's a point the physics of it becomes "saturated".

If you have a transistor with a current gain of 100 and give it 1mA, it will allow 100mA to flow into the collector.

What if you gave it 1A? It will allow 100A to flow . . .

. . . but what if at the same time, the voltage at the source will only allow for 5A? The extra 95A can't appear out of nowhere . . . The transistor has entered it's "saturation range". So long as I give it *enough* current to allow the current the load demands, it will allow that much to flow. In this notional case, any base current above 50mA will open the transistor fully open like a switch and sink 5A through the load.

In this case, the MPSA29 suggested is a darlington pair -- two transistors in one package, where the first transistor amplifies the tiny base current, then uses that current to drive the base current of an even bigger transistor . . . the total DC gain is rated at 10,000. You need a mere 2uA to saturate transistor with a the 20mA LED load. over 7.4v, the 300kohm base resistance (100k in bias + 200k in fluid) will happily generate over 20uA, so you're solidly in the saturation region. That's why the exact fluid resistance doesn't really matter that much.

 

[erwabo]

So with regards to my voltage sensor...im not going to get more than .6v reporting to that correct? Right now Im using that as ANY amount of voltage vs ZERO voltage... to trip my alarm. However as the battery loses capacity, from say 7.4volts to 7.1 volts, Im wondering if .6v will still pass through because if it doesn't it will trigger a false alarm and I could very well have half a tank left

 

[Craftydan]

sounds like a good test -- drain the tank with a nearly empty battery plugged in.