Problem with capacitive fuel level sensor..

[chsp129]

HI
i am using an capacitive fuel level sensor which give output as 4 to 20 ma..range signal..i use this sensor for calculate fuel level in an generator.but the problem is when i stop the generator i get some value 'x' ..but when i restart the generator again i get the value more than ('x')previous value..can any body help me what is the problem and how to solve this problem..

Thanks in advance..

 

[keith1200rs]

I guess a couple of possibilities spring to mind:

1. vibration

2. electrical interference

What sensor are you using?

Keith.

 

[chsp129]

HI THANKS for the reply..present am using capacitive fuel level sensor only.can u please tell me what type of hardware circuit needed for read the voltage fromcapacitive fuel level sensor ?.

THANKS IN ADVANCE

 

[keith1200rs]

Could you be more specific about the sensor? Make, part number, data sheet?

Keith

 

[chsp129]

hi ..am using an local made sensor so they r not giving the datasheet like that ,they just give the way it works and the output range.it was just an capacitive based fuel level sensor.and am using pic18f452 for reading the sensor...y am getting wrong values..

THANKS

 

[keith1200rs]

With no data I can only guess. Try some decoupling capacitors on the power supply to the sensor - 100nF and 100uF in parallel. Also try some capacitance on the output signal - maybe 10nF.

Have you an oscilloscope to look at the output signal?

Keith

 

[karesz]

Hi chsp129,
Eventual some vacuum effects in fuel tank?
I think you must control/check-recalibrate it in an external(opticall) controllable tank...
K.

 

[chsp129]

hi Keith and karesz thanks for the replyies..

i will try by using those capacitances,present am using 1000micro and 100 micro in parallel these values are good?
can u tell me what is the accuracy of ADC in PIC. i am using an multimeter to know how much voltage coming from sensor to my board i.e to pic controller,and from the adc value i i am calculate the voltage but both values are slightly different.

 

[keith1200rs]

It sounds like you already have a lot of capacitance. Is that on both the power supply and sensor signal?

The PIC ADC accuracy depends on the part number, you circuit design and possibly how you set up the PIC. Could you post the circuit diagram and PIC part number?

Keith

 

[karesz]

It sounds like you already have a lot of capacitance. Is that on both the power supply and sensor signal?...

Hali,
Yes, its eventual is reason of problems too...
@chsp129,
You can change pls the 100uF & 100uF to smaller values_as Keith told_ for 100nF & 100uF (exact values, as 47n/uF or 100n/uF, are noncritical), but not on the signal_it can maybe drive an 10 nF, eventual less_i.e. only 1nF (you must be experimental).
PIC/uP-ADC accuracies are generell not a problem for your application, then their are usual 10 bit systems, self then when you have only 8 bits as accuracy; its better as 1% , typically you has so at 0.5-0.25 % (from theoretical 10bits/~0.1%)...
K.

 

[chsp129]

hi..am using pic 18f452 ..present am using 100micro,104peco capacitors paralally in output signal..and not using any capacitors in power supply to the sensor,and am taking Vdd and Vss as refrence voltages and using the 10 bit result only..my application needs to show the fuel level in the tank on LCD ..so there must be no up and downs in fuel level ..so please tell me how to achieve this.i got up and down values.and in my code i divided 1to 5 volts into 160 parts..so for each 0.025 voltage change my fuel level changes....help me in this issue below am giving the circuit..

THANKS IN ADVANCE..

 

[karesz]

Aha,
you are writing, that your sensor is a current (4-20mA) type, I cannot see the (for me) proper connection for that, but it can be OK...
K.
Pls refer for data sheets of industrial current output sensors.
Are you sure, that at sensor out a (signal)voltage to measure is?
Has your sensor 2 or 3 or 4 lines pls, has the output connector/part of PCB some markings/characters...?

 

[chsp129]

ya am miss the resistor in that diagram..but am using an 250 ohm resistor to measure voltage from 1 to 5 volt...and am sure about voltage measuring..what are the things effect this voltage value..

 

[karesz]

How To Drive and Read A 4-20mA Sensor:
www.embeddedrelated.com/usenet/embedded/show/102992-1.php
www.datadolphin.com/Support/Tech Notes/DD-128 Datalogger/DD0105.pdf
Sorry; you didnt answered me the Nr of sensor wires & if you see some marks on that?....
If your sensor is similarly realized as these MAX1459 -> your signal-line load-capacitors (1000uF!! is toooo big) makes possible your problem :-(
OpAmps +input is capacitive loaded!
Can you measure some professionally pls; I mean, take an oscilloscope to check how is these sensor output signal? I can think that it has some "weak-up" effects_a kind of oscillating and so on.
K.

 

[karesz]

I forget it to tell; a 22pF on 1000uF at these (maybe) DC like frequencies is not real!
Change pls to i.e. 22nF! The electrolytic can be eventual smaller too...

 

[chsp129]

Thanks karesz ..i will try with the values what u r told and .i will send the result as fast as possoble.and can u please tell me how to avoid the fluctuations in voltage values ...bcs my fuel tank is very near to generator ..so some vibration is there for fuel tank..due to that am getting floated values..help me to avoid these variations..is averaging the values is good or there is any other procedure available..

HELP ME...

 

[karesz]

Maybe you should measure (ask the vendor): is it a simple resistor(potentiometric) sensor or a real 4-20mA current loop output solution?

Dont forget to check pls. for some vacuum effects in fuel tank?, then if your fluid has after some "sleeping" other level as befor_ you can not others measure it as yet: only with variable values...
Good progress!
K.

Added after 17 minutes:

How much is your variation please?
Is your tank to separate from generator-for short time only, for checking only the vibrations influence?
In my opinion its not an vibration influence, but more a vacuum effect, if the electronic is not disturbed over the generators function (sparks / EMC effects)...
You can apply some integratio into signal liine: between sensor connector & C1/C2 a serial resistor of 1-10KOhm? If your sensor signal input is a voltage (picked up from the spoken 250Ohm and GND) it will work.
Apropos working:
your PIC dont have some 100nF to! You must have it at the Vcc pins!!
100nF at the chip direct & an 10-100uF electrolytic somewhere on the Board...

 

[chsp129]

ya am sure about the sensor it was current loop output sensor only..can u please tell me what is the vacuum effect please..bcs i am install the sensor upto 10 tanks but 3 r working fine and remaining 7 gives the problem like this...all the 10 devices are having same hardware..only some giving the problem...ok tell me what is vacuum effect please...

THANKS for u r replies

 

[karesz]

Hi,
Its only one idea from me, maybe is wrong, bu t I can me introduce, that in the problematics tanks is not enough (or practically zero) air over the fuel, is some lower pressure as outside, can not breathe through the cap?
If you have 3 right system(or did I misunderstand & you have installed the same system in 10 generators?), change pls the sensors, than the electronics & even the tank caps (or self the tanks)too_ you must see what is other functioning, maybe it is not an electronics problem...
K.

Added after 1 hours 6 minutes:

Another more idea-question: Are all your tanks fully clean pls?
If their has on the bottom some mud_will cause similar effects :-(...
K.