I had built the circuit from about a year... I designed it to give the number of barrels in a 10-barrels water reservoir. There is a positive voltage line at the bottom of the reservoir, and for each level there is a line (wire)-their terminals, the sensing points, are fixed on a long plastic rod and distributed with equal distances along the rod according to the volume, when water reaches the line, then its number is shown by a 7-segment display... A priority encoder is used. This project works fine but the copper wires (the level sensors) react and oxidize with time... I clean them periodically.
The sensing lines are connected each to the base of npn transistor so that if water reaches the sensing point then Ib flows and Vc becomes 0 logic, and it is 1 when there is no water..
As I noticed, the high humidity between the sensing points on the plastic rod in the reservoir is allowing low current Ib to flow and therefore everything becomes wrong, the 7-segment display shows no clear numbers but all its leds lightens but with different intensity.

The attached circuit could help you understand what I mean specifically, and shows what can I edit in the circuit to solve the problem.

I thought about replacing the collector resistors with 1K ones and placing resistors between the base and the ground.

note that this device is not powered always but only when I press a switch to see the level...

 

where can I get stainless still wires or pieces of rod?

 

stainless still is for making clear unstained moonshine.

if you take two insulated plates that reach from the bottom to the top of the barrel the capacitenance ( sp again ) will vary with water depth. Nothing need touch the metalic parts of the electrodes so they will not corrode. there are many ways to measure the capacitenance.

If you have to stick with conductivity running ac for the measurement might help reduce electrolisis ( sp still again )

 

To ensure that I(base) to each transistor is shared properly, you should have a separate base resistor for each transistor of ~56K. R37 is then no longer required, but can also be reduced to 10K incase something in the tank tries to short VCC to ground.

I once used two nickels as probes to detect "sump pump well over-full"... they cost about 5 cents can be found anywhere (in North America)

Mechanically, you need to place the probes so that condensation is not going to allow a conduction path to VCC. Insulate the wiring well from the water except where the probes are.

Run experiments first before making changes.

Ifixit

 

I can scarcely find stainless still in the stores here in Lebanon, I may cut some stainless kitchenware to pieces. In the reservoir, I used a network wire with 10 lines, each line corresponds to one level (except Vcc) so I have the copper wires inside the water and what I'm thinking is soldering the stainless pieces with the end of the copper wires with solder wire..

 

Try a welding supplies shop, they sell stainless steel wire for welding.

 

hazim,

You will not be able to solder you wires with tin-lead or lead-free electrical solder directly to the stainless steel. You may be able glob on enough to make physical contact, but it will be a poor electrical contact that will be prone to oxidation. Find a welder or jeweler or mechanic or plumber that can coat 5mm of the end of the stainless steel probes with silver solder. You can then solder a copper wire directly to that.

Ken

 

I scratched the stainless and I was able to solder it with a copper wire (though it wasn't that easy)... the attached picture shows.. My multimeter if broken, It was a good idea to measure the resistance between the to solders terminals of the wire as shown in the picture just to check the connectivity of stainless steel and the contact with the solder... What do you think?

 

I would test the resistance. Your soldering looks like it may be OK. There appears to be a little etching on the bowl. Did you use an acid flux containing chloride? If so, be sure it is well cleaned, or you will eventually get corrosion from it. In fact, even if you do clean it, check for corrosion. Another approach is to use the acid flux to tin the bowl. Leave just thin layer on the bowl. Clean, then solder the wires with electrical solder. I have done that to make hollow SS electrodes from thin-wall tubing.

John

 

I didn't used acid or anything else.. I only used grease (I don't know what it is called actually, but it is a grease used in soldering for better contact.)

 

It is called flux, if it was for electronics it will be OK. If it was for plumbing then what jpanhalt said in post #12 applies, strongly. The wires won't last more than a month.

 

But why the wire won't last? I didn't used acid. I could varnish the wire anyway

 

That is a correct flux for electronics. The reason I asked about whether you used an acid flux is that is what is recommended generally for soldering SS. Also, there were some areas on the bowl that looked a little etched. Apparently, you did a very good job of cleaning, so the less active rosin flux worked.

I still suggest measuring the conductivity.

John