6x6x5mm 4pin Tactile Push Button Switch

[andering]

I got simple 3.3V circuit
https://goo.gl/q8fEM9

with this tactile push button
http://www.ebay.com/itm/400632384863?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

when i measure by multimeter voltage right after the open switch against the ground i got voltage around 2mV.
Is it failure of the tactile switch or for example 3.3V is too high voltage for such small switch or  something else ?

Thank you.

[rrinker]

 You are measuring across the switch? What settings are you using on your meter? Try it with a jumper wire replacing your LED and see what happens.

[andering]

First of all thank you for your kind response.
I created better picture of what i'm actually doing
and what is my reading. cca 1mV .
With bigger tactile push button i don't have this problem.

[rrinker]

 Do you get 0.0 if you short the meter leads together? Might just be meter error. I'll have to check some of my tact switches and see if they have leakage. Could be they all do that, but work when I hook them to a micro with a pull up resistor because 2-8mv is still well within the logic state threshold. It's just noise relative to a 5V pull up.

[andering]

I get 0.0 if i short the leads. But as you said i'm actually using the switch with Gate of mosfet, so pull down it's necessary anyway. It fix issue itself,
but it would be nice if you would tell me if your tact switch do same. Just to know if i should announce it to the seller or not.

Thank you one more time.

[rrinker]

 I have 2 different sizes of tac switches I tried, using my EEVBlog Bryman 235, and as far as I can tell, any voltage that shows is the result of induced currents caused by equipment and other noise picked up by the meter, not as the result of any leakage past the switch itself. Initially I was seeing as much as .8mv, which is after zeroing the meter with the probes shorted. At one point I actually saw a reading as high as 22mv but that was not repeatable and went away as soon as I rearranged the leads and changed the pins on the breadboard I was using. Just leaving the test leads open, laying on my ESD mat, the display will fluctuate as much as 100mv up and down, but it drops to zero when the leads are shorted. Conclusion: this is background noise. On volts, instead of millivolts, I get a steady 0.000 reading regardless of which tac switch I test. Also, I had my power supply set on 5V, not 3.3. What sort of reading does your meter show if you place the test leads on your bench, not shorted together?

 I thought of an other test I could do on the switch, but I don't have my trusty old FLuke 8012A set up. Then I realized my 8060A handheld ALSO does the conductance measurement, in siemens. Hooked to the switch, not pressing it, in the 200nS range (only one on this meter) it was 0. No conductance. Pressing the button - overload. Worst case was the other way around, resistance measurement, on all ranges it was over range with the button not pressed. On the lowest range, 2000 ohms, I got 1.74 ohms on the one I tested no matter how hard I pressed it, which is probably well within spec for these cheap buttons. I did not test the larger size one like this.

 Conclusion: I doubt there is anything wrong with your switch. At issue is a limitation of the test equipment and setup. Try the ohms measurements, if you don't get open circuit readings on all ranges, then maybe you do have a defective switch, but I'll bet your meter will show infinite resistance (overload condition) no matter what resistance scale you try.  Try the lowest range and press the button, it should be only an ohm or so when pressed firmly.

[sleemanj]

Let's assume that the switch was not working correctly and was passing current when it was supposed to be open - in other words it has some non-infinite resistance when it should be infinite (open).

Following your diagram of your measurement you have placed the meter in series of your circuit in voltage mode, thereby you could say you are measuring across the input impedance of your meter's voltage measurement mode. 

Say your meter has a typical 10M Ohm, we could in a way see this as you have a 3v3 supply, through the resistance of the "faulty switch" and the input impedance of your meter.  You have made a voltage divider and the measured voltage is the mid-point of said divider.

   

Code: [Select]

[+3v3] --- [ ??R switch] -- (0.001V) -- [10M meter] -- [0v]

If you work it out on that basis the "faulty switch" would have a resistance of 32990M Ohm (almost 33 billion ohms) instead of infinite.

33 Billion Ohms, and Infinite Ohms... is there really a difference, let alone a measurable one with a multimeter.

So anyway I think you can see that your measurement as you drew it doesn't give you a very useful data point :-)

[Psi]

Double check you are using the right switch terminals.

The 4 legs are actually 2 long pieces of metal folded over inside in a U shape.
So some combinations are just a dead short.

If you look closely you can see the direction the metal is folded.

Or just test the switch with a multi meter and find the right terminals that way.

[KL27x]

If you want that node to instantly and unequivocally be at ground, you need to give it a reason. When that switch goes open circuit, it will stay at 3.3V until the float charge is disspated thru humidity, dust, etc. Some meters have a very high input impedance which won't fully dissipate the float charge (99.9% of which is in the test leads, rather than the contacts on the switch), instantly.

If you want a clean 0.000V, instantly, you could put a pulldown on that node or use a double pole switch between power and ground.

No current != 0.000 V. An ideal test instrument would be able to actually measure a full 3.3V on that switch contact for at least an instant, even after it is fully open circuit.

[andering]

I'm very pleased by this forum,
thank to all of you for your warm answers.

@rrinker:
thank you for the test,
and also also for tip with resistance,
it's in overload state in any multimeter position
(i get -00.2 when leads just lay)

@sleemanj:
thank you for the calculation,
i learnt to try looking on problem from more perspectives.

@psi:
thanks for the the sentence

direction the metal is folded

it takes me all the time half minute to figure out where is contact and where not,
now with imagining folded metal i will guess it right away.

@KL27x:

you need to give it a reason

You are right, there will be some noise somewhere all the time,
so i should have pull it up or down all the time. Let it be my lesson .

Conclusion:
Indeed it will be some kind of noise (or capacitance),
one more time thank you for all participants.
I will just from curiosity order different same size tactile buttons(6mm) from different vendors and Bryman 235 instead my 5$ multimeter,
and i'll see .

But i see the pull down or pull up resistor as solution to my question.

Thank you.