T&L Publications

well i have sort of got my circuit going, but i have a slight problem, if the float is bobbing up/down slightly at the trigger point of the low fuel light coming on, it will blink on / off to the point its annoying. is there a way to attenuate the on/off of the led..
i'd like if it comes on to stay on or delay its going off for some time...
ya know what i am asking ? i guess something like a hystress maybe or something similar.

other slight problem is being able to find trim pots to match the resistance of the sending unit..
33 ohm - 240 ohm range most schematics i reference from are in the 10k - 100K range..

Perhaps just a low pass filter? Resistor in series followed by a cap to ground. Pick the values so that RC is say 10 seconds. 10Meg resistor and a 1uF cap seems about right. Take the input to the alarm's comparator from between the resistor to cap connection. This won't give any hysteresis but it will keep the LED from blinking.

i am having no luck..using a 741 op amp. perhaps this is not a suitable IC to use.
i may try my luck with a comparator..
it sux sometimes being a novice hobbiest..lol

Hi dac,

The average height of the fluid will be a DC value, while the ripples will be of various shapes not exactly like a full wave rectifier but something like that. I'm pretty sure you know how a full wave rectifier filter capacitor works so you should be able to understand that is about the same thing you have to do here. You have to integrate the ups and downs and the average will be the average DC value (which is what you want to measure or use as an indicator). Doing this is quite simple, and only requires a resistor and capacitor. The resistor goes in series with the cap and the open end of the resistor goes to the signal to be integrated (smoothed out). The new output comes from the center tap of the resistor and capacitor. The other end of the cap goes to ground.
Unlike a full wave rectifier that has constant harmonic content, the harmonic content of the ripples of the fluid will vary, but there will be some minimum frequency harmonic in there and that is of the most interest because filtering that harmonic is the goal, and the other harmonics will be filtered out even better because they are higher in frequency.
So the key then is to choose a nice long time constant for the resistor and capacitor. A resistor of say 220k and a capacitor of say 100uf to start. See how that works and if it still doesnt work well enough go to 1000uf which will give you a filter time constant of 220 seconds which should get all of the harmonics.

I dont know what your circuit looks like exactly so i cant say how much your circuit will load the capacitor and that is important too because that decreases the time constant which degrades the filtering action. If you post your circuit i can take a look for you if you like, or you can just try larger cap values and see how it goes.
Sometimes a cap across the op amp feedback resistor can be used too, but how well that works depends on other parts of the circuit too like the input resistance.

MrAl,Dac
"I don't know what your circuit looks like exactly so i cant say how much your circuit will load the capacitor and that is important too because that decreases the time constant which degrades the filtering action. If you post your circuit i can take a look for you if you like, or you can just try larger cap values and see how it goes."

When coupling to these types of circuits, I always feed their outputs into a Fet Op-Amp in a follower configuration. The input Z will approach near a giga ohm and loading is so minimal its not even worth calculating.

Hi Robert,

Yes that's a good idea in many cases. There are times though when the appropriate power supplies are not available for running an FET type op amp, and then perhaps almost any type of amplification might work. I thought it would be best to find out what he already had so we could work from there. If he already has an op amp in the circuit there is probably a way to easily mod it to work as a filter combined with what it is already doing. If he can supply some sort of schematic it would make this easier to figure out

Right, and as Confusious say-"A picture (AKA Print) is worth a thousand words! I think it was Confusious

i stumbled across this today..

http://members.shaw.ca/efog/diagrams/lowfuel01.jpg

but i was told to avoid the 741 IC, not sure why, so i was thinking to use a comparator instead.
i think i will try to breadboard this one this weekend and see how it goes..
any input on this will be appreciated.
it looks like it has delay on the input as well...

The 1741 is a rather ancient and maybe harder to obtain chip. If you want to go with a more battery "friendly" circuit, I would use something similar to a TLO81 amp. It may not have nulling adjustment, but for comparator design it not an important parameter. This is a single 8 DIP style and I think some may have nulling features if desired , but the single Op-Amp draws only 2.5 ma as compared to 10 ma for the quad versions. I notice the circuit shown has no hysteresis and this should be a mandatory addition. The 'lytic shown in conjunction with the senders resistance would not give much low frequency rejection. If sloshing gasoline is the problem you experience it would seem that the level change rate would be at a very low frequency possibly a few cycles/second and would require a rather huge filter to average out that that situation.Long term banks may alter the alarm reliability in those turns depending on the senders mechanics by holding a false reading for the duration of the turn. Also it looks like the trip point setting is about 2-4 vdc. above ground. You may want the trip point to be related to EMPTY TANK voltage rather than ground. Dedicated comparators excel where speed of operation is important. Op-Amp comparators should not degrade your purpose in this design. Not trying to change your preferences here, just offering my opinion.

well this is why i am here to get opinions on what is best.
that chip i thought crossed to a 741..
so what would be better the cap or a hysteresis. need schematic help on that..
i cannot believe that there is not more of this on the net...

dacflyer wrote:well this is why i am here to get opinions on what is best.
that chip i thought crossed to a 741..
so what would be better the cap or a hysteresis. need schematic help on that..
i cannot believe that there is not more of this on the net...

Hi dac,

That is why i was suggesting that you post something about what you already have built up, unless of course you dont have anything yet. The simplest solution could be a single resistor and single cap arranged as a low pass filter (which im sure is on the net). This means you may not even need an op amp at all. It depends what you have to feed that into (what is making the measurements).
I also find that some things are on the net and others are not, and the more detail you want about something the less likely you will find it on the net unless you want to pay for it. Some documents they want to sell for as much as 30 dollars each which is just a rip off.

well i started... with a 741.. didn't work like i wanted on a bread board,, so i scrapped it, and looking to make another as soon as i can figure something out.. will experiment some more as i find the info and time