IC input triggered by very short voltage drop

Thread Starter

- Dirk -

Joined Aug 8, 2019
14
Hi everyone,

This started out with a project from the book Make:Electronics, more specifically the one based on the Jeopardy game where only the fastest player (to press his button) his led goes on. There's a master switch to reset the states and the author mentions he included some capacitors to counteract a false trigger most probably due to switch bounce.

There are indeed false triggers when I leave these capacitors out. But after reducing this circuit to something very simple (but similar) to investigate, it starts getting bizarre.

When I build a voltage divider
Vcc (9Volt)---------10K Ohm---------300K Ohm-------GND

and touch the node in between both resistors with a screwdriver I from time to time get very short voltage drops at that point. My oscilloscope gets triggered and I indeed see a very short voltage drop to zero Volt. Note:
my hand doesn't touch the metal part.

Before this simplification I started with this voltage divider and in between those resistors there was a wire to a spst switch. When I connected the switch (in open position) to ground I got those same voltage drops, say once every 10 times I made the connection with ground. It made no sense to me... the wire is interrupted by the switch?! How can it pull down the voltage?! Even if only for some fraction of time.

I had the same with the original circuit. The master reset switch made contact with ground and the 555 timer trigger input was where I put my 300 K Ohm resistor.

Once I'm back home I'll attach some screenshot and a drawing of what I described.

Maybe it is clear to you? I'm puzzled

Many thanks,
Dirk
 

ci139

Joined Jul 11, 2016
1,696
10:300 k divider is "slow" the most insensitive input of the NE555 (a bi-polar version) is the RESET (p4 - see datasheet) -0.4mA/-1.5mA.max (12kΩ average) the most sensitive is THRESH 30nA/500nA.max ( ! 133MΩ ! average) , the TRIG lies in between 0.5µA/2µA.max (9.4MΩ average) ← (( @ 5 to 15V ~ 7.5V average ))

your screw driver acts as a sphere capacitor.. (the static charge travels also over the insulation , so - you may add and your body to this) ..drawing the electrons off or or injecting them to the point of contact ↑with quite high sense trigger levels↑ that may be sufficient to cause switching

9V / 310kΩ = 29µA to be over-ridden . . . ? with the 10k pull-up is suspicious but for a short spike . . . maybe

with op amps (that have their input sensitivity about the same there is no problem driving the inputs through the 100kΩ attenuating resistors)
 

AlbertHall

Joined Jun 4, 2014
10,915
Try connecting a 100nF capacitor from the junction of the two resistors to gnd and see if that fixes the problem. Difficult to say without knowing the whole circuit.
 

Thread Starter

- Dirk -

Joined Aug 8, 2019
14
Hi,

Indeed from the second edition. Here's the drawing:

circuit.jpg

I marked the required capacitors with a red circle. The green circle is the node where I notice voltage drops when the master switch (set, reset) is being moved. And as a result the leds (one at a time or both) go ON which is unwanted. It doesn't make sense to me: the push buttons are depressed so they should not pull low their connected resistor when the master switch is "set". When "reset" is used any bouncing is irrelevant as well as the 555 in bistable mode will register a pull low once => it's output won't change due to bouncing. And the 555 output is connected to different OR IC inputs, not the node where I register a pull down to zero.

Maybe it wasn't very clear from my description but I tried to reproduce with this simple "circuit":

issue.png

Not sure how well made these cheap tiny tactile pushbuttons are but at least I wouldn't expect this to happen?!

button.png
This is the one I have been using but it also happened with others.

Hope this clears out my initial question/description :)

Rgds,
Dirk
 

MrChips

Joined Oct 2, 2009
22,893
The 300KΩ resistor has little effect in the circuit. The only effect would be to reduce the voltage at the node to 8.7V.
I can see no reason for the effect you are observing. It is not caused by the quality of the tactile pushbutton.
I would suggest that you look for a cause some other place in your entire circuit.

What is the input sensitivity of your oscilloscope?
Transients seen on the oscilloscope are very common. For example, just the thermal cycling of a temperature regulated soldering station will cause transients.
 

Thread Starter

- Dirk -

Joined Aug 8, 2019
14
The 300K Ohm resistor was just to mimic the input of an IC. Something with a higher resistance to have high voltage behind the pull up resistor which can be pulled low.

Oscilloscope is Owon 1022i. But even without the oscilloscope connected, I observe those leds start to shine if you switch that master switch a sufficient amount of times. I'm talking about the circuit from the book here... Adding those capacitors solves it (of course) but I want to understand why it happens...
 

Thread Starter

- Dirk -

Joined Aug 8, 2019
14
Hi everyone,

I recently had some time to further investigate. After replacing the sliding switch with a push button to reset the 555 and just permanently connecting those push buttons to ground, I noticed the circuit became stable without those extra capacitors. But of course this modifies the intended functionality of not allowing the players to push their button until allowed.

Then I focused on the push buttons and (un)connecting ground. That's were the issue is. What I added to the circuit is orange on the drawing. Even though the push button should interrupt the connection, I can trigger the OR input by plugging and unplugging the connection to ground. This happens with any kind of switch I had lying around. I also tried to put a diode (in reverse) instead of a switch. In case of the diode, it doesn't trigger. But anytime I touch the OR input with something of metal it triggers as well (due to its high sensitivity I guess)

Any idea?

Thanks,
Dirk
 

Attachments

crutschow

Joined Mar 14, 2008
26,407
For your possible interest, below is a simple game circuit using small SCRs, that is not sensitive to switch bounce.
First one to push his button latches the corresponding LED and locks out all others.

1585581535723.png
 
Last edited:
Top