Looking for a homebrew salinity test circuit...

deno

I have installed a RO watermaker on my boat, and have been looking around for a "inline salinity tester" sensor and· homebrew circuit that will either produce a variable resistance, capacitance, or· pulses in which the Basic Stamp could measure.· The "inline probe would go in the product line of the watermaker and monitor the salinity so as to divert any "bad water" overboard and not to the water tank.

I have tried making different probes that measure the resistance of the water flowing "across" to probes measuring the resistance(units of time)·using RCTIME, with a fixed capacitance, ·however this does not produce consist results.

Varying pulse count using COUNT would be a nice way to go...

Does anyone have a good idea....other then buy a already built device.· The object here for me is to use a Basic Stamp.

Thanks for your reading time and consideration...

deno

Tracy Allen

Maybe chapters 5 and 6 of the Applied Sensors workbook would have something relevant for your project...

www.parallax.com/dl/docs/books/edu/WebAppliedSensors-v1.4.pdf

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

stamptrol

deno,

I expect you're using a couple of electrodes in a piece of plastic pipe to measure conductivity.

In doing a similar thing awhile back, I got more consistent results if the polarity of the electrodes was reversed periodically to keep the electro-chemical build up on each electrode similar. You could even take a reading with polarity one way for few seconds then flip polarity and take a second reading and average them to give better stabilty.

Goes without saying the pipe should be straight, vertical and have the liquid moving at normal speed before the reading is taken.

Cheers

piscis

Get a low cost In-Line blind transmitter with 4-20ma signal and a sensor.



http://www.eutechinst.com/



·

deno

Thanks to all that responded.· I have decided to go with Mr. Allans idea in the Applied Sensors book.· Using a 555 to develop a pulse output which COUNT will count very nicely.

Mr. Allan, one question please.· Does the 100K ohm resistor between pin 2 and 3 on the 555 stay in place with the electrode leads in parallel with·it, or is the resistor to be removed with only the electrode leads to pin 2 and 3 respectively?

Deno

Tracy Allen

It is best to leave the resistor in place. The purpose of the resistor is to set a lower bound of frequency, even when the sensor is dry or slightly wet. That keeps up the AC excitation and forestalls polarization of the electrodes. When your system is operating, there will always be water in the tube, so that will drive up the frequency.

There are a couple of ways the circuit can be improved. One of those is to put capacitors in series with your electrodes on both sides. The purpose of that is to isolate your electrodes from even the leakage currents of the LMC555. Please look at the circuit diagram here: www.emesys.com/smx.htm

I have made good use of that circuit with a variety of low resolution conductivity sensors, and have it available for sale as a surface mount PCB if you are interested. (no obligation!).

Take care for the electrodes, that they are in a cell of constant geometry, covered in a consistent manner in still or at least non-turbulent flow. And the electrode material can make a difference. Stainless steel is okay, but a carbon rod is probably better for low surface energy.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

deno

Mr. Allen...thank you for all the information, and I have built your 555 circuit and included the input capacitors.· The BS2 reads the pulses very nicely. I can calibrate the stamp/probe using the duration argument in COUNT.

I have hacked a test probe from a dual TDS tester that I found on line, and am able to use that probe with the stamp, and inline (running water) use the other to the store bought device at the same time.· I am running well water for testing thru both probes to compare and calibrate.· I noticed that on the probe connected to the store bought device, there is an additional set of wires, and doing some reverse engineering, have determined that this is a temperature sensor.· Which I am not using on the· probe that I hacked from this same device for the stamp.· (It is a dual monitoring device with a slide switch to select either probe.)

Anyway, after Googling around the internet, it seems that temperature plays a part in computing the TDS reading.· How much, I am unable to figure out.· I think it is a positive coefficient, but how much per degree?

Thanks again...deno