Seeking advice on electricity

[TonyS]

I am making a simple alarm for the holding tank from a project in the "Boatowners Wiring Manual". This alarm functions when water makes contact with two probes which completes a circuit that drives the input of an op amp positive and therefore gives an output to the amplifier that in turn causes an LED to light. The circuit works when I stick the probes in a glass of water. It also works if I join the probes with a resistance under 100K. However if I put the probes of my digital meter on the resistance scale in the water I get no reading on any scale up to 2 M ohm. On a more expensive meter the resistance appears to be 7 M ohm. Can any of you electronic experts expalin to me what is happening.
The circuit function would indicate that the resistance of water is in fact a few thousand ohms but why can this not be measured on an ohmmeter. This is important for designing the best probes for a holding tank and how far apart the two electrodes should be without doing a lot of trial and error. I am think of drilling the cap and putting through two SS bolts with rubber washers and a nut behind so that when the water level reaches the two bolt ends the circuit is made. Can any of you experts comment or help please.

 

[bdsweeting]

I am no expert here but... there are some instances where a digital multimeter is not the best instrument to use.

If you can get hold of a good old analogue one you might find the answer to your question. I think it is something to do with the way the meter reads its' own leads/probes.

 

[TonyS]

Thanks for that. You are right. I have just been down to the garage. Found the old analogue meter in a cardboard box. Replaced its batteries. Now get a reading of 20 K ohms which is the region I expected. It still surprises me that there is no difference to the resistance whether the probes are close together or far apart. I also don't understand why the digital meter, which measures accurately the resistance of resistors on all ranges will not measure the resistance of water when the analogue one will.

 

[Ardenfour]

It's related to the level of voltage, and hence current, which the different types of meters use to indicate resistance. An analogue meter has to physically move a pointer via a coil to produce a magnetic field. This causes the deflection of the needle. This requires more power than the digital meter, where the signal inputs are via op-amps in the digital meter circuitry, and are of much higher impedance than the analogue

 

[BobOwen]

My creaking brain cell recalls a few pertinent matters from basic electronics.

Measuring resistance with two probes in water is difficult. There are an infinite number of parallel paths between the probes. This parallel circuit confuses matters.

Also pure water is a very good insulator. Water with impurities such as salt will lower the overall resistance. Anything inbetween will give hugely varying readings. I think temperature is a big factor too.

Water is not normally measured directly by resistance though. There are devices with parallel plates at a set distance apart. These pass current via a bridge circuit and measure conductivity (related to resistance, but I can't remember the formula ). It is this fixed set up that allows accurate measurements to be made in any water mixture.

Your circuit is probably capable of reacting to a wide range of varying resistances though. If it has a variable feedback circuit on the op amp you should be able to adjust it to react to almost any change. The circuit needs to be set to just below a trigger point. Air between the probes is a very high resistance. Any liquid contact between your probes will immediately drop the resistance by a high factor and this should easily 'fire' the amp off.

All a bit vague - but it was a long time ago! I'm sure someone a little more "current" will clarify this.

 

[jimduncan1]

Another one here trying to remember the old basic's. When comparing readings you should take into account the sensitivity of the meter which is expressed in ohms/volt which refers to the deflection or reading of the instrument (ie it's sensitivity). When meters are produced for different resistance ranges they use a series of shunt resistors inside so that most of the current runs through them and that a smaller proportion runs through the meter (it doesn't get damaged and readings are proportional to the range setting). These resistors also have a tolerance setting (ie: 1%, 3%, 5% etc 4th colour band on old ones) which can have a significant impact on the reading you get depending on the range setting and of course the voltage source in your meter (ensure batteries are fresh) which drives or deflects the meter. Salt water should get lower resistance readings.

To cut to the chase I think it will work fine. Good luck!

 

[daveyjones]

Some years ago I had a leaky wooden boat, so to leave her in safety I had to rig a float switch to operate the bige pump automatically. I found proprietary switches to be expensive and short-lived, so I arranged my own float switch. A small length of 20mm plastic tube, open at both ends, with a plastic float inside. On the top of the float I glued a small magnet, and an enclosed reed switch (as in door sensors for burglar alarms) was glued in position at the top, with all cable connections enclosed in epoxy resin. If you position and design it properly, the switch can remain out of the water at all times. This arrangement avoids any problems with resistance of the liquid you are monitoring, as well as electrolysis problems. Generally speaking, it is better not to have metal with voltages (no matter how small) in direct contact with water of other liquid. Incidentally, my float switch triggered a very simple timer circuit, with an adjustable delay to avoid "cycling" - the pump would run for a preset time, enabling the bilge to be cleared. It just needed a bit of experimentation to adjust the switch height and timer setting. The whole thing cost a few pounds (all parts cheap from Maplin) a few hours work at most, and it worked without a hitch for 5 years, and was still ok when I sold the boat. If you would like a circuit diagram let me know. It also had the advantages that the electronics could be mounted well out of the wet areas, say behind the switch panel.

 

[TonyS]

Thanks for your advice and proposal. However, when you consider a holding tank and what it contains, any floats of tubes cannot be considered. The point is that the circuit I have works and some of the answers above point to possible reasons why an analogue meter records the right result and a digital one doesn't. I still am not convinced why as both have 9 volt batteries to power their ohmmeter. The question of conduction of water I put on Google and got more than 1000000 responses. It appears that water has many peculiar properties and pure water is an insulator but only needs a few salt ions to make it quite conducting. This is as we find normal tap water. What I will find in the holding take remains to be seen but I think the probe I have should work - I will let you know in a few weeks. I have already built the circuit which fits in a tiny Maplin box on the chart table bulkhead and have used a quad op amp ( 29p) so have three inputs and three LEDs and 3 switches for test and to set the individual circuits. The other two probes will be in the water tanks to indicate before they are overfilled and leak! It is for my brothers older boat that he keeps on Belle Isle and he hopes to fit it next week.
Thanks for all the replies and just one point on the last three posts which is quite interesting. I needed a number of 100k resistors so went to Maplins and the guy gave me a strip with the third ring orange. I told him I didn't want 10 k but 100 k resistors. He said they came out of the 100 k bin! I said they must be in the wrong bin then. He then opened the catalogue and showed me that for certain ranges there is a new colour code and 100 k is now orange!! So I think we should forget the rhymes or develop new ones and check with a digital resistance meter.

 

[halcyon]

Four colour code, they have a extra black ring, so you have one less on the multiplier, i.e. orange not yellow.
You may well find you will need to clean the probes at intervals, as you will get a build up of a white deposite that insulates the electrode. We designed a similar system for Sealine back in the early 90's. But we had a pulsed positive to the electrodes, as a permanent feed could cause electrolytic corrosion, if the pos probe is in the liquid or the tank not emptied.

Brian

 

[TonyS]

Thanks for that. For the water tanks he will only switch it on when filling the tanks - hense the switch. I agree that we may have a build up problem on the holding tank and it will remain to be seen how often the probe needs cleaning - but anything is better than an overflowing holding tank! I would like to have made an ultrasonic depth sounder that sits on top of the holding tank but am not good enough at electronics to design it and there are no articles that I have found. If anyone could provide a circuit I think that would be the best method. It would need a tranducer, a pulse circuit and a receiver and some sort of comparator so the alarm would trigger above a certain depth.

 

[halcyon]

The problems with ultra sonic is minimum range, stick to the porbes.

Brian