PDA

View Full Version : HYDROMETER- To Measure Sea Water Salinity Levels



craigwb
- 22nd October 2011, 13:54
Hi All

Can someone please help me with the following:
I am in the process of building a Hydrometer to measure the changes in salinity levels of sea water. I am using Bronze probes due to there corrosive properties. I am using the ADC on the Pic 16F887 with Vref+ and the ADIN command. I get readings on the probes on large changes but, the slight changes in the Salinity levels I cannot seem to read, has anyone got some info on best way of doing this and reading it on the Pic? Should I amplify the signal before it goes to the Pic?

Kind Regards
Craig:)

Jumper
- 22nd October 2011, 14:21
Hi,

What will change with the salinity level? Is it the OHM or the Capacitance or any other physical factor?

What level of accuraccy and resolution are you looking for? Maybe you can start by providing the formulas for how the theory behind salinity and the factor you wish to measure.

How much will oxidation of the probes influence the value?
How will the water level on the probes influence?
Does temperature affect the physical factors?

Interesting project but I am not sure you can find a simple answer here without giving alot more background information about the theories behind this.

But if I have a deep look into the crystal ball I would say that a 12-bit ADC could be worth looking into, the 887 pic only has 10-bit.

amgen
- 23rd October 2011, 00:58
Best Answer - Chosen by AskerThe average conductivity of sea water at 20degC and a salinity of 35g/kg is:
4.788 S/m (Siemens/meter)
http://www.kayelaby.npl.co.uk/general_ph… (http://www.kayelaby.npl.co.uk/general_physics/2_7/2_7_9.html)

The resistance is the reciprocal of the conductance, so:
R = 1/(4.788 S/m) = 0.2089 ohm/m <===ANSWER
at 20 degC and a salinity of 35g salt per kg water.

If you place two probes 1 meter apart in seawater under the stated conditions, the resistance will be 0.2089 ohm, according to the calculation and published conductance data.

10 months ago
Report Abuse (http://www.picbasic.co.uk/answer/report;_ylt=Agdqegi_OMHq__hiSBoDS7UIGAx.;_ylv=3?qi d=20101214192859AAc4eE3&kid=NoZ_GE_NUEgaW8wwnXGV&.crumb=.lq4oMfnpktE)
Asker's Rating:http://l.yimg.com/h/28158/images/all/rating-5.gifAsker's Comment:perfect answer!
--------------------------------------------------
looks like the readings move to the milliOhms once the water has some salt introduced, so you would have to work with op-amps to detect small changes in a certain range. Also mabye a reference to calculate from...such as .35 Ohms(some a/d reading) equals 200 PPM,parts/million.

that link seems broken, just a search for conductivity of salt water.

Jumper
- 23rd October 2011, 08:23
http://www.reefcentral.com/forums/showthread.php?t=1812946

So you might want to look at a AC solution.

craigwb
- 23rd October 2011, 16:04
Hi Guys

Thanks for all of the input so far, I was going to use two Bronze Probes (1.6mm TIG Silica Bronze Welding Rods) as their Corrosive Resistance is very good against salt water. My water tank will hold 200 Liters of salt water. The water level will be monitored and will remain pretty constant and the temperature will me monitored and kept at 25 Degrees Celsius.

I want to measure the PPM which will be between 27 and 33 which has a Specific Gravity of 1.019 - 1.025. (Resistive measurement taken on probes)
After having a look at your replies and then doing a bit of searching on the net I found a very interesting article on this very matter, with an AC circuit etc have a look at: http://www.octiva.net/projects/ppm/

I will build his circuit and give it a bash with bronze probes, I will report back on my finding and thanks again for all of your help!

Craig

amgen
- 23rd October 2011, 22:17
Wondering, will that measuring current, as small as it is, hurt or annoy the fish.
You can turn on to measure for X milliseconds then off for X seconds. Also the touch-sensor ckt on PIC is actually an AC signal.

craigwb
- 24th October 2011, 07:27
To take very short sample measurements will definitely be the way to go, I will have a look at the touch sensor on the Pic that looks like a marvelous idea.
Craig

Demon
- 22nd January 2012, 04:44
Moved from Schematics.

Robert