There needs to be a load resistor between Q1's collector and ground. You could stick a 1K in there.

Ken

 

put them in. The one thing that is interesting is that I read 3.7vac on the short lead terminals(right hand side terminal) when they are connected.

The long lead terminal is showing contact(it is not) but when I read voltage I get an open signal and I read 1.53vac. So just me touching the wire with my meter makes it change states. On that note, even if the meter is off it does it. It does it with anything metallic I put on terminal. Only right hand side, and it happens 100% when I use multimeter, but not 100% with other things.

 

I'm gonna try larger wire tomorrow.

 

Tried a couple things this morning. I think first I should clarify my vocabulary for how I see the circuit working. I thought about it and it may be confusing to those reading. I'll stick with "open signal/LED on" when the probes/leads are not touching, and a "closed signal/LED off" for when they are touching.

First I tried a 4 conductor, twisted pair, shielded cable. Hooked up a twisted pair, and tried it with two wires from the opposite twisted pairs. Both times I get a closed signal whether or not the leads are touching. Soon as I take the shield wire and touch it to the 0vdc terminal, I get an open signal, however, touching the ends of the wires together does not give me a closed signal, it maintains an open signal no matter what I do. The shield is for sure not touching the wires at any point in the cable. Also, the cable is about 45' long.

Second I tried some 16awg wire. Cut two separate 45' long lengths and hooked them up. It kept an open signal until I started moving the wires closer together not far from the board and got the closed signal, even though they were separated at the far end. I separated them near the board, went to the other end of the wires, and when I would bring them close they would give me a closed signal. If I tried real hard to keep them as far apart as possible, and only bring the two ends together, sometimes it would work correctly, other times it would give me the false closed signal.

So there it is. I guess the next step is to see just how long of leads I can actually use before getting this erroneous indication. I would hope/think that there is another way around it, but I am not sure what is happening exactly. This gets way deep into electrical engineering and past generic circuitry I believe. However, I would think there might be something that can be done to the circuit to alleviate the issue.

Correct me if I am wrong, but the pin 5&6 are getting a high signal/voltage which is why pin 4 goes low, creating the closed signal when it should be an open signal.

I measured, and compared, voltage on 5&6 circuit #1 with 5&6 circuit #6(which just has short jumper wire for leads). #1 with leads in terminal open circuit I get 7.2vdc - #6 is .369vdc. #1 with leads in terminal but now closed circuit I get 11.5vdc - #6 is 11.82vdc.

It should be noted that I have my meter connected to 0vdc and reading dc voltage at the top of the IC. When I touch the meter probe to pin 5 or 6 the circuit immediately becomes open. And, in that open state I can touch the wire ends together and circuit works fine. I assume this has something to do with the resistance inside my meter, or is it voltage?

Any help here would be appreciated. I'm willing to try different things to see what might work.

 

Got it.

Turned down the frequency, sitting at around 36hz right now. I deducted that from all the testing that I was simply getting Parasitic Capacitance, if I am correct? I've dealt with things causing issue with 4-20mA signals before, running them to AC conductors, but I just knew not to do it and to use shielded wire. Well, shielded wire, I think, only worsened my issue. In fact, I did not try it, but I bet I could go back to 60hz(that's where I first turned it down to). It works with shielded wire, but you can see the LED pulsing slightly, even more so at 60hz. With standard 16awg(not sure of the insulation) it is a steady open state when the leads are hooked up and not touching, and a steady closed state when they are touching or in a bucket of water.

I'm gonna leave the leads in water over night, but I see no sign of electrolysis thus far which you can almost immediately see with DC. I assume there would be a threshold frequency where it would start with AC?

Big thanks to Ken. I would have been stuck a lot farther back had it not been for you, chased many rabbits down needless paths, and who knows when I would have ended up with a final project(I don't give up unless it just seems impossible).

I'm not done yet though. All I did was prove I could use 1 set of long leads. Lol. I am pretty sure that multiple sets wont effect anything if the frequency is knocked down, but who knows. As before, I'll keep things up do date as I go. At this point, I know better than to think that I have it completely whooped.

 

Frequency! Long parallel leads! Yes that happened to me too! I'm a little familiar with this type of water detectorhttp://www.picaxeforum.co.uk/showthread.php?12266-08M-Christmas-tree-watering-controller
I had cross coupling between the drive and the detector wires in the cable on mine. Had to turn down the frequency. But that was on a 4' cable. You might want to run the drive and detector wires as two separate shielded cables if you have to go very far.

But, are you running the drive and detect leads 45'. NO...no...no! Not a good idea with this type of AC drive. Put your circuit as near to the probes as possible. Run your final output back to your PLC. I must have missed a description of the layout of you plant.

Ken

 

Frequency! Long parallel leads! Yes that happened to me too! I'm a little familiar with this type of water detectorhttp://www.picaxeforum.co.uk/showthread.php?12266-08M-Christmas-tree-watering-controller
I had cross coupling between the drive and the detector wires in the cable on mine. Had to turn down the frequency. But that was on a 4' cable. You might want to run the drive and detector wires as two separate shielded cables if you have to go very far.

But, are you running the drive and detect leads 45'. NO...no...no! Not a good idea with this type of AC drive. Put your circuit as near to the probes as possible. Run your final output back to your PLC. I must have missed a description of the layout of you plant.

Ken

Yep, I am going to be putting them as close as possible, but the filters themselves are 20' deep. So the leads that go to the bottom point will be 21'-22' long. I was going to originally put the box with the circuit board inside the filter house which sits in the middle of all four filters, but I think any length of wire I can save is better. I never thought that the length of wire would be this big of an issue.

Interesting thing is I just did some more testing and there is for sure things I have to tweak. I opened the other circuits and #1 died(the only one that has the frequency turned down). So I turned it back up to 60hz, and it works with three additional circuits open. So I am going to turn it up a tad bit more and see if that fixes things. Of course, at this point I will also cut the wires to 25' and use that for my testing.

 

Ok, need help. I am not thinking I am doing my calculations wrong for frequency. It should be

Correct? I've been changing C2 and R2. C2 to 3.3nf and R2 to 10M.

I am out of caps for the day, will scrounge some tomorrow. Can I use polarized radial caps in C2? I have a lot of them laying around. #1 circuit runs fine, but until I change the others I will not know if everything will be good to go.

I wonder what the best combination would be to have in a capacitor for C2 and a variable resistor for R2 in order to be able to fine tune the frequency once in place. I am thinking I need to stay well below 100hz.

 

As you lower the frequency, C2/3/4 will also have increased. This lower frequency may also may make it more susceptible to environmental power line noise I'm starting to question the appropriateness of this design for you particular environment. Not knowing all of the physical installation limitations makes me question this approach. Sorry.

I feel like I might be leading you to a dead end.

Ken

 

Don't be sorry. I have gone way far above anything I dreamed of doing this. I know it's not a "big" project/circuit, but it's big compared to my level.

I understand the limitations here, as far as your understanding of exactly the application and all. It truly is one of a kind even though there are similar things out there. Some would say there may be a reason why no device like this specifically exists, I like to look at it from the perspective of no one has wanted to figure it out bad enough yet.

I have one thought. Not that it's something I have really considered best, but it's a thought. What if I was to put the drive and detect part of the circuit closer to the probes, if not even within a few inches of them. Then just run a wire coming from pin 3 on the IC, and one too pin 5&6 on the IC.

They would have to be put in waterproof potting, but they would also be inside a pipe and surrounded by waterproof foam. The leads of course epoxied through the pipe wall.

Do you think the output of pin 3 coming down and the running next too the input for pin 5&6 would induce voltage like what's happening now? It would be DC at that point. I would have a common ground wire running the length for all the circuits.

Thoughts? I'm not giving up on this anytime soon, if at all. This is one of those gut feeling things that I know I can make happen, just need to have correct line of sight.

 

"I have one thought. Not that it's something I have really considered best, but it's a thought. What if I was to put the drive and detect part of the circuit closer to the probes, if not even within a few inches of them. Then just run a wire coming from pin 3 on the IC, and one too pin 5&6 on the IC.
They would have to be put in waterproof potting, but they would also be inside a pipe and surrounded by waterproof foam. The leads of course epoxied through the pipe wall."

That might work. Remember that the probes have to be "dry" to cause a change of voltage when the water drops. I can't visualize how you intend to have them mounted. Submerged, but then no water film between them when exposed.

"Do you think the output of pin 3 coming down and the running next too the input for pin 5&6 would induce voltage like what's happening now? It would be DC at that point. I would have a common ground wire running the length for all the circuits."

I would put the oscillator and detector at the probe site. Then just run 3 wires, +12V, common, and an output line, up to the surface/PLC. That is more like I had envisioned. The delay circuit could be at the bottom or the top.

"I'm not giving up on this anytime soon, if at all."

Hang in there.

Ken

 

The main body of the probe is going to be schedule 80 pvc. Not sure of the size just yet, but it will be mounted to the filter wall on existing unistrut, so I'll use stand off pipe clamps.

So I guess what I'll do is build 6 individual boards, coat them good I'm thinking with liquid electrical tape, then epoxy coat them. I will put them inside of a section of pipe, maybe 6" long at most, have the probe wires attached to stainless steel probes - one on each side opposite each other, the 12vdc and 0vdc coming in and running out(except the last one in the chain, very bottom probe), and a signal wire running straight through which each section/probe will contain as many as needed for the probes below it.

Here is a very basic idea of it.



I'll put the following part of the circuit in the pipe, and the rest up top. It will mean that I will use 6 extra IC's and socket's, but I really need to have the ability to control the time delay portion of the circuit.



Do you see any issues with this?

 

What is the gray wire on the left in yours?

Mine:


The open bottom, sealed chamber would keep the water from reaching the top of the sealed chamber and allowing conduction between the probes where they pass through the seal, after the water drops. Not sure if condensation would defeat this.

Ken

 

Grey wire is the wire(s) that will carry the signal from the probes below it. Red 12vdc, green 0vdc. The blue wires are going to the screw/probes.

Each probe 'section' will be sealed with waterproof foam. The probes themselves will be stainless steel pan head screws which will be drilled and taped into the pipe wall, threads sealed with epoxy. The entire thing should be water proof inside, but I still will seal all the electronics. The top of the pipe will extend well above the surface of the water, and be capped to not allow water to get down in it.

I just am curious if the 12vdc/0vdc wires would cause any issues with the signal wires coming out going to the rest of the circuit? Being DC instead of AC it should not correct?

 

Totally different arrangement than I had pictured. Will a film of water still connect the screw heads after the water level drops below them?

Ken

 

Totally different arrangement than I had pictured. Will a film of water still connect the screw heads after the water level drops below them?

Ken

I figured we were thinking two separate arrangements.

I do not believe they will. At least I cannot see how. It will be hung vertically, and the water is relatively clean(not oily or greasy or anything like that). The only thing will be algae that could bridge the probe heads if allowed to go/grow that long, but we generally keep things washed down inside the filter with big sprayers, and it would be a general rule that the probes would need to be sprayed off regularly.

 

Do you foresee an issue with signal wires run with the 12vdc/0vdc wires? I'll do a test circuit to see, but I suspect there will be no issue since its done with other types of sensors.

 

Do you foresee an issue with signal wires run with the 12vdc/0vdc wires? I'll do a test circuit to see, but I suspect there will be no issue since its done with other types of sensors.

No, as the signal wire(s) will just be on or off.

Ken

 

Just to throw out another multilevel detector concept that I did a while ago.

Ken

 

Interesting. I looked and looked, but I am not sure if I ever saw that one. How does it differ in how it works vs the one I am building?

It was easy to weed out 95% of what I did not want/could not use. It has to use AC for the contacts, but be powered by DC, and it needs to be able to control relay(ability to switch on/off mA current) for the input to the PLC - which now I am just using the transistor itself. And of course I needed to be able to control a time delay for when the contacts are made.

The circuit I am using now fits the bill to a "T" but the one little issue with the long leads

I think its all for the better though now. I started out thinking of putting individual circuits(even though I had not found what I wanted to use yet) in the probe pipe. I was just concerned with them being in the water so I immediately moved on to ideas that put the above the water. But now that I was made to actually contemplate it as the only way of making this work, them being under water does not bother me one bit. I'll just need to pay close attention to the building of the probe pipe - not try to rush things like I am so inclined to do......