T&L Publications

Hello, I'm looking for a way to take the bounce out of a simple circuit. I have a float switch connected in series with the coil of a 12 volt relay. When the float switch closes, the relay coil energizes and pulls in it's contact. The problem is that the float bounces during transition from open to closed and the relay chatters a little bit. I don't want to put a time delay relay in there unless I have to, but what about a capacitor in parallel with the coil? It would be nice if there was about a 5 second delay from float switch closed to relay coil energized. I know that you may ask questions about the minimum current to pull the relay contacts in etc. and I don't know the answers, but how about just a conceptual opinion? Or better yet, does anyone know what kind of circuit would be commonly used in industry for a situation like this? Thanks...

A simple solution would be to install the float in a well. To further clarify, install the float inside of a larger tube/pipe extending into the tank or sump, this will isolate the float from disturbances in the tank or sump. We use this method in our cooling water sumps at my plant.<p>[ April 11, 2003: Message edited by: Crowbar ]</p>

Yes! Crowbars solution would probably work, providing the float/water movement was caused as assumed - You still need considerable construction work to put in the well (Proper name `Stillage Well') so a simple electrical/electronic solution is better.
A time delay might probably do the job but it's not really necessary - Put in a monostable or some other device so that the pump motor starts at the first signal but then latches for 5 seconds or so irrespective of the signal from the float switch. If five seconds were not enough then you need a latch period long enough to hold in the pump circuit until the float is below the level at which it activates. Presumable you have a switch at another level to switch off the pump OR the same float travelling up and down a wire or guide held taut/in place by a weight. I have used this solution in a chemical plant so I know it to be practicable.
Will

Crowbar's solution brought to mind that if the well had a small inlet, its level would lag the outside level. There would be less chance of chatter because the float would be rising.

Well the stillwell solution seems simple and effective, only problem is that this is just not practical to implement in my situation. You'd just have to see what I have. No, this will require some electronics but thanks for that idea, it's a good one.

Hi, mj. I would take a look at your physical setup first, like some of the prior postings said. Most float switches have some hysteresis built in, but if you have agitation on the surface of the liquid, that can cause problems. Also, many float switches aren't meant to drive an inductive load like a relay coil, and you may be seeing a pitted reed switch contact in the magnetic float. You should try ohming out the reed switch -- if it's more than a few tenths of an ohm when closed, or changes with repeated openings/closings, you need to replace the switch. Actually, that will cause the "chatter" you're talking about, too.<p>Here's a solution that's made for those reed switches -- let something else drive the inductive load. Try using the float switch to charge up a 10uF cap thru a 500K resistor (assuming your reed switch is capable of dry circuit switching), and tie the junction of the R and C to 1/6 of a 74C14 hex schmitt trigger inverter. The output can be inverted by another 1/6 of the 74C14, and that output can be used to drive an NPN transistor, which drives the relay. You might have to experiment with the values of R, depending on the manufacturer of the 'C14. I've seen this used in a number of places, and have used it myself successfully a number of times. Assuming the circuit has been off for some time, you should see relay turn-on about 5 seconds after the reed relay turns on, and you will have a couple of seconds of noise immunity on the switching of the reed relay once it does turn on. Take a look at this app note from Fairchild / National Semiconductor on the use of the IC:<p>http://www.fairchildsemi.com/an/AN/AN-140.pdf<p>If you need more help on this, feel free to e-mail me. Happy hunting.<p>[ April 13, 2003: Message edited by: Chris Foley ]</p>

consider using a 555 as a one shot. I would typing 555 into google you will probabally get a circut there. You can hook the output of the 555 to a solid state relay to control the load Than take the float switch and use it in the circut to trigger the 555