Good Ideas from talented indivudals

I have to make a new version of an existing sensing circuit to make sure a cap is installed before an air filled vessel goes into the water.

We currently have a piece of all thread that gets pushed down by installing the cap. The all thread pushes a button on a microswitch opening the circuit which tells us that that cap is installed.

This works ok but if the all thread moves down a bit it gets pressed too hard and causes the switch to crack.

Different ideas have been thrown around... proximity switches, hall effect sensors, rf id (to make sure it is there, will not know if the cap is properly seated), optical sensing, etc.

So, i need suggestions from all you individuals as to what you think would be a possible solution.

No idea is a bad idea..... How can i make sure a water tight cap is in place? At first it will just be if the cap is there or not and maybe progress into something that will be accurate enough to sense if the cap has been properly tightened.

Thanks for any ideas. If you need an illustration of exactly what i mean let me know and i will make some quick visio diagrams.

If you need an illustration of exactly what i mean let me know and i will make some quick visio diagrams.

yes please...

How can i make sure a water tight cap is in place

step 1: Rewrite the problem as the solution
==> fill the system with water and if it leaks it is not in place.

Step 2: make it realistic
==> fill it with air and if the pressure goes up the cap is in place, if not its not.

Check - www.triz40.com - for pattern based problem solving (might be not the best site to start but OK)

See attached pictures.

Effectively the two spade terminals on the bottom are shorted when the cap is no on the assembly. When the cap is places on the assembly the rod is pushed down, causing the switch to open.

Looking for a way to have two wires connected when no cap is present and open when the cap is present.

robtillaart:

How can i make sure a water tight cap is in place

step 1: Rewrite the problem as the solution
==> fill the system with water and if it leaks it is not in place.

Step 2: make it realistic
==> fill it with air and if the pressure goes up the cap is in place, if not its not.

Check - www.triz40.com - for pattern based problem solving (might be not the best site to start but OK)

Fair suggestions.

1: The air filled can is full of electronics, filling it with water each time i place the cap kind of defeats the purpose of needing to keep the water out.

2: We do a vacuum test to check the rest of the seals, but having a port that i can use to induce air means i have another hole to check and effectively doubles my problem. Then i would have to make sure two caps are in place.

Replace the microswitch with a magnetic reed switch and place a small permanent magnet on the end of the rod. This would prevent any physical contact from the rod to the switch.

Any reason why the rod couldn't be shortened (perhaps by filing a bit off the end)?

Dropdeadkick: I was thinking the same thing. I am going to order some reed switchs and some hall effect sensors to play with. My concern is that I have some 110vac in that general area that might cause some issues. But I do intend to try.

Cr0sh: as the cap is put on and off it changes the travel slightly. You won't know its going too deep until the switch stops working..... but if you make it any shorter intially it won't trigger the switch.

Why not orient the switch so that it does not present an end stop to the cap?

PeterH:
Why not orient the switch so that it does not present an end stop to the cap?

Interesting. How so?

1: The air filled can is full of electronics, filling it with water each time i place the cap kind of defeats the purpose of needing to keep the water out.

2: We do a vacuum test to check the rest of the seals, but having a port that i can use to induce air means i have another hole to check and effectively doubles my problem. Then i would have to make sure two caps are in place.

@1: this first reversal step was just to generate ideas :slight_smile:
It brought up the idea of instead letting something flow in (which was the problem) to detect a leak to let something flow out (solution) ==> pattern solution = reverse(problem) - ok not allways but it helps to generate ideas.

@2: you don't need an extra hole, on contrary I would propose a small compressed C02/air cylinder, a valve and a pressure sensor and some bits to control it. Yes it means some additional electronics. After the cap is closed the valve opens to raise the inner pressure by X%. If there is any leakage wherever it will be detected in contrast of just detecting a switch has been pressed the right way.
Additional, if there is a very small leakage some overpressure might prevent water leaking in (keep it local)

Goofballtech:

PeterH:
Why not orient the switch so that it does not present an end stop to the cap?

Interesting. How so?

Hard to know what PeterH had in mind but a lever microswitch held such that the rod passing it deflected the lever would be a very basic way:
____ | |
Sw |\ | |
___| | |
| |

An optical path broken by the rod would be another way.

robtillaart:
@2: you don't need an extra hole, on contrary I would propose a small compressed C02/air cylinder, a valve and a pressure sensor and some bits to control it. Yes it means some additional electronics. After the cap is closed the valve opens to raise the inner pressure by X%. If there is any leakage wherever it will be detected in contrast of just detecting a switch has been pressed the right way.
Additional, if there is a very small leakage some overpressure might prevent water leaking in (keep it local)

I like it, the cylinder would have to be fairly large though, not sure how many trips it would take before i would have to swap it out. It is most certainly something to ponder. I am not sure about any issues that could come about with submerging a compressed air vessel though. I suspect it wouldn't do anything really because it's a maintained 1 atmosphere of pressure inside the container regardless of depth.

robtillaart:
Hard to know what PeterH had in mind but a lever microswitch held such that the rod passing it deflected the lever would be a very basic way:
____ | |
Sw |\ | |
___| | |
| |

An optical path broken by the rod would be another way.

Also a good idea. would probably use some of the same hardware but still serve the purpose. will add it to the possible list as well.

This is good stuff guys. Thanks a ton. I tend to over think things..... elegant solutions are the best but i am always severely lacking in those.

Keep em coming.

I suspect it wouldn't do anything really because it's a maintained 1 atmosphere of pressure inside the container regardless of depth.

It could even compensate for the pressure from outside. At 10 M depth the pressure is 2 Atm, 1ATm for every 10 meter IIRC..

Looking for a way to have two wires connected when no cap is present and open when the cap is present.

So, can you have the wires connect to terminals on either side of that plunger-thing at the "top", with the plunger under spring tension pushing it into place to keep the connection. Then a smaller radius pin in the cap can push down the plunger when the cap is place, breaking the connection.

This would be a simple fix let a spring hit the microswitch The spring would let you over tighten the cap and not break the microswitch.

And could easily be setup to insure you have tighten the cap

pressure sensor inside compared to one outside if they differ cap is not on

thebfs:
pressure sensor inside compared to one outside if they differ cap is not on

Fair idea. The issues i would have would be that the pressure could also potentially be the same if the cap is on... and the fact that the sensor would have to be rated for the depth of water in order for it to be outside of the housing.

robtillaart:
It could even compensate for the pressure from outside. At 10 M depth the pressure is 2 Atm, 1ATm for every 10 meter IIRC..

we have already build the walls of the house to withstand the pressure. Would take a little too much air to compensate for 4,000 meters ( ~17,500 feet )....... :slight_smile:

T_Lex:
So, can you have the wires connect to terminals on either side of that plunger-thing at the "top", with the plunger under spring tension pushing it into place to keep the connection. Then a smaller radius pin in the cap can push down the plunger when the cap is place, breaking the connection.

If i understand you properly this would be similar to a standard hood switch on most alarm systems. It was also discussed at one point and we are looking at how to get it into the housing in the most efficient way possible.

be80be:
This would be a simple fix let a spring hit the microswitch The spring would let you over tighten the cap and not break the microswitch.

And could easily be setup to insure you have tighten the cap

Also a very good idea, we would have to get the spring sized just right and make sure to keep it locked into the bottom so it wouldn't fall out of the housing when the cap was removed but that shouldn't be too bad.