I'm having problems testing a 7490 decade counter, the switch sometimes toggles 2 or 3 counts (it's a pushbutton). This is already using this debouncing circuit I found:

(well great, I can't post images yet. It's an inverter that's connected to another inverter that feeds back the first inverter in a loop and is also the output. The push-button for the zero output is connected between ground and output and the one output push-button is connected between the two inverters and ground.)

Does anyone know any ways I can improve my mechanical switch toggling? An easy fix to the circuit I'm using would be ideal, but if that's not possible, what's a decent debouncing circuit I can use that works 100%? I just need it to be negative-edge triggered.

Help much appreciated

EDIT: Here's the image:

enter image description here

  • 1
    \$\begingroup\$ I just gave you an upvote, which should remove the new-user restrictions on your account. See if you can post the images now. \$\endgroup\$
    – apnorton
    May 21, 2013 at 0:35
  • \$\begingroup\$ Post a link to the image and one of us will edit it in for you. \$\endgroup\$ May 21, 2013 at 0:41
  • \$\begingroup\$ Smart circuit you show there, however I don't like it as it (ever so briefly) shorts the inverter's output to ground. What you probably want is a flip flop like the one Michael Karas shows. \$\endgroup\$
    – jippie
    May 21, 2013 at 6:33
  • \$\begingroup\$ If you can translate a level-triggered to edge, adding a large enough capacitor might be an option. It's very simple and does solve de-bouncing issues. \$\endgroup\$
    – MandoMando
    May 21, 2013 at 20:59
  • \$\begingroup\$ You are fortunate to have a double-throw switch. That is the secret to true debouncing, and most of the time, that's not what we're given! I agree with @jippie , this circuit uses the same principle as the one with the flip-flop and it is potentially clever. I think with some series resistance we can avoid shorting the outputs...I can't prove it yet...let's just say all the simulators in town now hate me :) \$\endgroup\$
    – gbarry
    May 21, 2013 at 21:43

1 Answer 1


Implement your debounce like this to fully eliminate the effect of the switch contact clatter.

enter image description here

Captured courtesy of this site.


After discussion in the comments section I am adding the below picture to help visualize the "NAND latch" in the DeMorgan equivalent of its logic to show how the latch really is a low true input latch.

enter image description here

  • \$\begingroup\$ Logic seems wrong in this pic. Switch at 1 is Set high, reset low \$\endgroup\$ May 21, 2013 at 11:13
  • \$\begingroup\$ The logic of the picture is correct. With the switch shown in position 1 the OUT terminal will be low. When the switch transitions back to position 2 the OUT terminal will go back to a high level. This schematic is misleading because the switch applies GND to the inputs of the cross coupled latch. The gates should actually be drawn as their demorgan equivalent as an OR symbol with the bubbles on the input terminals instead of as shown. \$\endgroup\$ May 21, 2013 at 16:32
  • \$\begingroup\$ Thanks Michael -- so the inputs are actually S-not and R-not \$\endgroup\$ May 21, 2013 at 19:18
  • \$\begingroup\$ @ScottSeidman - I've added an update section to the answer to show the DeMorgan gate equivalent that makes it easier to see and understand how the debouncer latch works. \$\endgroup\$ May 21, 2013 at 20:39

Not the answer you're looking for? Browse other questions tagged or ask your own question.