suspect electrical noise on 555 trigger input

Thread Starter

waltsar

Joined Nov 10, 2012
16
I have a circuit that uses a 555 timer to generate a 500 msec clock pulse to a 4017 decade counter. The trigger input of the 555 comes from 2 reed switches which come on sequentially. When I tested the circuit on the bench using micro switches with one foot long twisted pair leads, the circuit worked fine. Now it's installed on a train layout and the reed switches are located about 12 to 15 feet from the PCB. Whenever either of the 2 switches is tripped (momentarily closes), the 4017 counter produces erratic counts. The reed switches are wired with twisted pair wires (about 4 twists per inch). Also, there is a .01 mfd cap across each switch. Any ideas?

Thanks,
Walt
 

ericgibbs

Joined Jan 29, 2010
9,056
Hi W,
You need to add an extra debounce component close to the PCB end of the cable connection.
A resistor in series with the input close to the capacitor should improve the debounce.
If that is not sufficient consider using a CD40106 Schmitt IC to debounce.
E
 
Last edited:

danadak

Joined Mar 10, 2018
3,682
For future reference a single chip solution for this would be a PSOC 4 low end part.

$ 4 board to do this - http://www.cypress.com/documentation/development-kitsboards/psoc-4-cy8ckit-049-4xxx-prototyping-kits

Has the debounce component and the counter and a ton of other stuff. A "component"
in PSOC is a functional resource on chip. Also has clock component to generate the
pulse, clock for counter.

Here is an example debounce component used to drive an onboard timer/counter.

Google "PSOC 101" to see a series of videos on using PSOC.



If you needed LCD PSOC can also drive that.


Regards, Dana.
 

ElectricSpidey

Joined Dec 2, 2017
770
A 555 in one shot mode “is” a debounce circuit when clocking a 4017. I would suspect either the output pulse is too short, or that 15 foot “antenna” connected to the trigger is causing RFI problems, and will require some isolation.

Of course there could be other issues such as lack of decoupling…etc.
 

Thread Starter

waltsar

Joined Nov 10, 2012
16
A 555 in one shot mode “is” a debounce circuit when clocking a 4017. I would suspect either the output pulse is too short, or that 15 foot “antenna” connected to the trigger is causing RFI problems, and will require some isolation.

Of course there could be other issues such as lack of decoupling…etc.
You're right, the 555 "is" the de-bounce, that's why I put it there, and I agree that there must be some kind of "noise" coming in on the 555 trigger input because of the 15-foot "antenna" (the circuit works fine on the bench with a switch (and no .01mfd cap) connected to the input with only one-foot wires). I was hoping that the twisted pair wires and the .01mfd cap across the reed switches would be enough to suppress any induced noise.

I have .22mfd caps across VCC to ground on each IC chip and a .01mfd cap on term 5 to ground of the 555.
 

ElectricSpidey

Joined Dec 2, 2017
770
Well if you are SURE your debounce pulse is long enough, remember the switch can bounce closing and opening, then you need false triggering components.

Do you have a small cap and resistor connected from pin 2 to VCC?

And why do you think you need the one across the reed? I’m not sure you need that one. You need to shunt any noise to positive...not ground.

Let me know…I do have a radical solution if all the usual stuff fails…
 

Thread Starter

waltsar

Joined Nov 10, 2012
16
Well if you are SURE your debounce pulse is long enough, remember the switch can bounce closing and opening, then you need false triggering components.

Do you have a small cap and resistor connected from pin 2 to VCC?

And why do you think you need the one across the reed? I’m not sure you need that one. You need to shunt any noise to positive...not ground.

Let me know…I do have a radical solution if all the usual stuff fails…
Thanks for the followup - I have a 10K pull resistor on pin 2 to VCC to keep the input high until the negative going leading edge of the contact closure, but no small cap. The cap across the contacts should present a low impedence to induced line to line noise, no?
 

crutschow

Joined Mar 14, 2008
23,741
A capacitor across the contacts will also tend to weld them from the high current spike when the switch closing suddenly discharges the cap.
 

sghioto

Joined Dec 31, 2017
1,038
I have a 10K pull resistor on pin 2 to VCC to keep the input high until the negative going leading edge of the contact closure,
You can also try adding a 100nf cap between pin2 and the reed switch. This is to prevent false triggering on the positive edge of the pulse.
SG
 

Thread Starter

waltsar

Joined Nov 10, 2012
16
You can also try adding a 100nf cap between pin2 and the reed switch. This is to prevent false triggering on the positive edge of the pulse.
SG
Replacing the twisted pair wires and removing the .01 mfd caps seems to have stopped the erratic triggering of the 555.

Thanks for all the replies.

Walt
 
Top