Water tank sensor

[varistors]

Sensors output is 0-190 ohms (2 wire output). Output needs to activate 4 (empty -- -- full)
red LEDs sequentially as the tank fills. What component will activate as a SW by
0, 63, 126, 190 ohms? And how to wire the four SWs from a two wire input. Source 12v.
Thanks for looking Ken

 

[Nanren888]

SW?
LM3914? Data sheet has several example circuits.

 

[Bluejets]

Arduino promini would measure input through analog as a voltage.
A couple of if statements to recognise the analog input levels and output to the appropriate LED's through 4 outputs via series resistors..

I would tend to reduce the 12v down to say 9v through a cheap ($2.00) converter into the RAW of the Arduino as a power supply as in most instances this can go over the 12v limit.
Besides, it is much easier on the onboard regulator.

What is the sensor..??
Photos, details etc.

 

[varistors]

The four LEDs are on a boat panel which cannot be modified to accommodate the LM3914. Possibly four of the dot bar displays LEDs can be paralleled with the four boat LEDs?
The original sensor tube contained four reed switches (SW) evenly spaced in the tube (empty - - full) as the donut shaped magnet rose around the tube each reed SW Closed in turn turning a LED on. (5wires a com. )
The new sensor acts like a pot as the magnet moves up/down tube. It has just two output wires. Four SW need to close at 63 ohms intervals.
Being an older gentleman microprocessors are off the table.
Thanks for your replyies. Ken

 

[Nanren888]

I'm not sure that you can do this without adding other components, in a circuit. That is something to transform a resistance change into LED switching.
The chip I suggested would lead to a small circuit.
As with Bluejets' suggestion and mine, turning your resistance into a voltage & then some sort of circuit to turn voltage into a choice of LED to light seemed the simplest way.
BTW, some of us older guys have managed micros. The single chip solution I suggested was an alternative, just in case uPs were not your preference.
.
Maybe you can give us an idea of what sort of thing you have in mind if it can't include a chip.
Possible, maybe to do it with a few transistors & resistors. Would that work any better for you?

 

[varistors]

Older gentlemen learn all manner of things, sadly microprocessors/programming I have not tackled, my apology. I can put
ICs and components on a PCB with the help of a good diagram.
Transistors & resistors would work best.

 

[Bluejets]

Older gentlemen learn all manner of things, sadly microprocessors/programming I have not tackled, my apology. I can put
ICs and components on a PCB with the help of a good diagram.
Transistors & resistors would work best.

Do you have some specs on the new sensor.
Link to details...??

 

[AnalogKid]

The restrictions in your question are not clear. It sounds like there are 4 LEDs already existing in the boat, and you want them to perform the same function. With the new sensor, that **can not** happen without additional components. They can be purchased modules, or a circuit you whip up from scratch, but somehow you have to synthesize electronically four voltage comparisons and LED drives.

To start: Can you verify that in the old system, each LED was wired directly to a switch and the 12 V boat power? Any sign of external resistors, like current limiting resistors for the LEDs? If not, then the LEDs are 12 V industrial indicator types.

Next, can you measure the LED current?

Next, do you still want a moving dot display (only one LED on at a time), or can you live with a bar graph display. A discrete bar graph display circuit takes considerably fewer parts. With the LM3914 you can select between bar and dot modes with no parts count penalty.

ak

 

[Harald Kapp]

The four LEDs are on a boat panel which cannot be modified to accommodate the LM3914.

You can use the LM3914 anyway by thinking out of the box:

• Use the chip in dot mode. Configure it such that only the first 4 LEDs are used by
• Instead of the LEDs connect a small relay to each of the four relevant LED outputs.
• Connect the Relay output to where formerly the reed switches were connected.

The LM3914 will now drive the LEDs in the panel in the same way the reed switches did before.

 

[Alec_t]

Did the old reed switch connect between its LED and 12V (+) or between the LED and ground (-) ?

 

[ratstar]

What kind of sensor are u using to detect the water level? I couldnt read it here.

 

[Harald Kapp]

I couldnt read it here.

See post #1, line 1:

Sensors output is 0-190 ohms (2 wire output).

 

[varistors]

Each of the four LEDs is protected by a 500 resistor (12v input) output 24mA. The four small reed switches are in the LEDs ground line.
Sensor behaves like a pot empty is 0 ohms full is 190 ohms. The blue wire is ground the brown wire goes to the gauges sensor input. The sensor wire is labeled not to go to battery positive. The rated current for the alarm switch is 500mA, there is no alarm switch in this problem. But as there is no other mention of current & the wires are 16 ga, probably the sensor can handle 500mA. No definitive knowledge of what's in the tube, ( e mailed company twice) .

 

[ratstar]

So you just want to light 1 led at 200 ohms, to 4 leds at 4 ohms right? And everything is working great other than just that?

Just grade the resistance off the output of the sensor- then the leds with the higher resistance will turn on later than the ones with less resistance.

If your getting complex and have an oscillator, you can actually put them through a capacitor instead of a resistor, and its probably safer for the led's, as well as wasting less energy.
Also just got another idea, if you make the output of the sensor control the frequency of the oscillator, that could work in ways too.


Dont blow up your water tank... but ive got a rough draft out for how id tackle the situation. If there isnt enough raw power out of the sensor you could add 1 transistor to the start of it. Note this circuit probably doesnt work its not tested at all, it would be a fluke in hell if this was right first time, its just a rough idea.

 

[Martaine2005]

If your getting complex and have an oscillator,

It should be a criminal offence not having an oscillator and capacitor.

Martin

 

[ratstar]

I'll have a go at anything, works or not thats the other question , but running out of energy to get more experiments going, I still havent implemented bugger all, but I dont blame myself I'm so fatigued every day.

 

[Bluejets]

Sensor behaves like a pot empty is 0 ohms full is 190 ohms

Strange, even for a basic milliamp guage this would be back to front.
Is there some reason any link to the unit is a secret..?