• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

water reservoir level sensor

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

trend

Member
Joined
Jun 25, 2002
Location
GA
Hello, I need/want a rig that had a water reservoir with a built in sensor to tell me how much water is in it... or at the minimum, a setup to indicate to low water level or to high... any suggestions?

Also what about a water pump that can hook up to a computer to make it push or pull.. hook up directly or indirectly.

thanks-Lee
 
it could be cool to implement something like this gauge. that is just an example of the product i was thinking of. you could probably get one at a lawn mower place. im not sure its what you were thinking of.
 
You're looking for a thing called a tank level indicator (TLI).

How they work is quite simple. You have a bunch of resistors in series. between each resistor you have a magnetic reed switch. Put all of this in a water tight tube. Around the tube you have a foam dohnut with a ring magnent. As the dohnut moves up and down the tube with water level, it closes a reed switch at that level giving you an analog resistance value proportional to water level.

tli.JPG
 
i'm actually going to need the computer to know how much water i have.. so i think some of these ideas will work.. but anyone know how they will hook up to the computer ? rs232?

thanks-Lee
 
Any reason for this? as long as you seal the res. the water level won't change. Unless there is a leak, but if there is one, your water level in the res will be the least of your concerns.
 
Actually. this project is bigger than a single computer.. So I must do this... any ideas?
 
A level indicator out of a car fuel tank maybe?

They are normaly a long arm with a float on one end. I dont knw how big you res is but they could be shortend to fit.
 
here's my idea on this:
my mobo: asus a7v8x had an additional conector for a powersupply temperature sensor. A temp sensor is nothing more than a variable resistor. So if you can find the right resistance range and connect the switch/float system UberBlue was talking about. You can monitor water level using mbm5, or any other software that reads mobo sensors.
There are scematcis how to make a parallel or serial port I/O boards, but i gota look for a link.
 
so something like this? (the blue part is the backboard)
and the magnetic reed switches are :
Normally Open Without Magnet

and what type of resisters would i use?
and i just slide that board downa tub and seel it and have a magnet around the tub and that is it? (after connecting to my mobo)
pic.jpg



and i was just thinking.. the res will only need to hold 50ml of water.. so maybe have a tub and have the water in the tub and have the mag in the tub, and have the mag reed switches outside...

doe sthat sound like my best option?

thanks_Lee
 
Last edited:
trend said:
so something like this? (the blue part is the backboard)
and the magnetic reed switches are :
Normally Open Without Magnet

and what type of resisters would i use?
and i just slide that board downa tub and seel it and have a magnet around the tub and that is it? (after connecting to my mobo)
pic.jpg

^^^^ Pretty much how it works. Just remember, the magnet needs to float.

My deailed knowledge of Ohm's law has long faded into the backgroung

To make this work:
  • The first step would be to dertermine the resistance range needed. A good starting point would be to measure the resistance of the of the power supply temp sensor at a warm instance and a cool instance. This will define the range you need.
  • Dertermine how many reed switches there will be. I'll use your example with three. Realize that having more reed switches will give you better resolution.
  • Use the resistance range you measure from the PS temp sensor, lets assume 100 ohms, to figure each induvidual resistors value. Take the resistance range value and divide by the total number of reed switches. In this case each resistor would need to be roughly 33 ohms (100/3).

Some things to take note of:

  • if the PS temp sensor is a thermal diode, which I think it is, that part of the plan is shot to hell.
  • The whole unit will need to be mounted with the highest total resistance value at the top to corespond with the thermal sensors resistance value increasing with temperature.
  • the greater the resistance range the better. If the PS temp sensor turns out NOT to be a thermal diode, I would throw it in the coldest freezer I could find, then get it as hot as I dared. or if you are using a monitoring program like MBM5, determine the max and min temps it is capable of displaying and take readings at those temps.
  • You're not going to know what the water level is in absolute liquid units. you will have a high to low scale expressed in degrees. Where at on that scale full and empty are needs to be determined through experimentation.
  • All of this is pointless without software to take advantage of it.
 
Last edited:
Upon further investigation, the PS thermal sensor is in fact not a thermal DIode.

I cut the thermal sensor off a dead Enermax PS I had lying around

In hot tap water a value of 30 K ohm was read.

In ice water a value of 4 K ohm was read.

A range of 26 K ohm that should be within the temp range of most monitoring programs.

So trend,

The question now is how long can this thing be; as that will determine the maximum number of switches that can be packed in, and that will determine the induvidual resistors values.
 
lets see.. my case is the cheap case (~45bucks on
newegg)


and the PSU is antec SL300S
i'm looking right now and seeing what kind of diode it uses



Hmm.. I need a sensor at: ground 0, 14ml,19ml , 25ml, 40ml, and maybe another one higher to make sure i don't get overflow...


thmm
 
Last edited:
trend said:
[B
and the PSU is antec SL300S
i'm looking right now and seeing what kind of diode it uses [/B]

That part shouldn't matter. If the PS conforms to any standard motherboard interface, the thermal sensor readings should be the same regardless of manufacturer.
 
what do you think would be the best design for a small res doing what we are talking about?
just the water in a cylinder and have the magnetic in the cylinder?

also, how durable is this system going to be? I need something VERY durable.. the res. will be fluctuated from 0 to 40ml and inbetween up to 20 times a day..

thanks soo muych for the help!

thanks_LEE
 
trend said:
what do you think would be the best design for a small res doing what we are talking about?
just the water in a cylinder and have the magnetic in the cylinder?

The smaller it is, the harder it's going to be to make.

Ideally the res would be tall and skinny, vice short and squat.

Tall and skinny would be more sensitive to water level fluctuation.

I'd use a hunk of 2" PVC, about 8" long.

Give me a bit to put my mad MS paint skills to use. I'll draw you a detailed diagram.

Need to think of something that is very bouyant to make a float out of...
 
crap, i need 0ml, 14ml, 19ml, 25ml, 40ml and maybe another one to make sure i don't overfill the cylinder

i forgot to mention 19ml
thanks-Lee
 
trend said:
what do you think would be the best design for a small res doing what we are talking about?
just the water in a cylinder and have the magnetic in the cylinder?

also, how durable is this system going to be? I need something VERY durable.. the res. will be fluctuated from 0 to 40ml and inbetween up to 20 times a day..

thanks soo muych for the help!

thanks_LEE

What the HECK are you doing that's going to cause that kind of fluctuation!?

As far as durabilty is concerned:

How I know so much about these things is I've spent the better part of the last five years on a nuclear submarine. Not anymore though (thank god!). Every tank onboard, and there are a ton of them, has TLI's. Part of my job was to keep them in working order. The bulk of that involved keeping all the mineral deposits and organic growth off, to keep the floats from getting stuck. Let me tell you, You don't know the meaning of "suck" untill you've been stuck in a dark, cramped, stanky butt tank cleaning lord knows what off those damn things. Anyways, in that five years I only saw one TLI fail (there are hundreds, some of which are subjected to ungodly pressures.) and that was due to a bad cable connector which allowed the TLI to flood internaly.

So in short, they are very reliable.
 
Back