EEVblog #961 - Monkey Debouncing

How do you debounce a clapping monkey?
Probes The Monkey makes his return.
All about contact debouncing, setting up a universal counter as an event counter, and other issues that can arise on a seemingly simple test setup.

Support the EEVblog through Patreon!

EEVblog Amazon Store (Dave gets a cut):

[dentakuweb]

I love videos like this about practical stuff. They might not get the most views but they will still be relevant and educational years from now.

[Boffin]

RC = time. If you don't want anything faster than say 1/20 second, R1 = 5k,
hence C = 0.05s / 5k = 10uF.
Use a smaller cap.

[John_Ridley]

Dave, I once built a "number of seconds to graduation" counter with a calculator. I bought a 1hz timebase from Radio Shack, hooked it to a 4066, and hooked that to the = key on a calculator. Poked in the starting value and "-1" and started the circuit. This was back in the late 80s. Mounted the whole thing on my dorm room door.

[mrlithium]

Special thank you to Thunderfoot for appearing in this video as a test subject.

[magnus-j]

The technological knowledge, passion and time put into this video (a setup meant to debunk the Batteriser claims, I would presume) far exceeds the technical knowledge, R&D resources put in to developing Batteriser.
Love it.

[andrewwhite]

This is a good video, Dave. It really shows why simple stuff is not so simple in analogue electronics..

[BerndFelsche]

You had a scope. You could MEASURE the contact bounce interval and calculate the RC values accordingly.

[magnehaneberg]

Hi, Dave! The sensitivity / hysteresis is in fact specified in the operating manual on page 3-2. But to be fair, not everyone has time to search through the manual to find it. First you set the trigger level to the midpoint (or slightly under in your application using a capacitor). Then you look at the section "Voltage range and sensitivity (single shot pulse)". We find the sensitivity is 50mVpp for this application. Then we see LO sensitivity is roughly 2x HI sensitivity, so our LO sensitivity is around 100mVpp in this application. Easy. :-) Anyway, I like your channel. You're doing some great work!

[brendanzotto]

AH Ganssle's debounce circuit, we used this in my micro electronics course. Great video!

[MrRobbyvent]

This troubleshooting approach is very instructive whatever is the interest in the subject

[DoRC]

man you're gonna kill the batterizer testing!

[capoman]

Sweet little tutorial based on a simple project. Love it. People must watch the calculator counter hacking video. Shows how you can get the same result with nothing but a dirt cheap calculator!

[shazam]

Use a '4538 one-shot IC. It has edge triggering and hysteresis built in. Set the RC time constant for the longest contact noise duration and wire it as "non-re-trigger-able". Need it faster than the noise duration? Use the 2nd half of the IC to "kill" the input noise.

[GenerationXerography]

The cymbals had just the right resistivity and distributed capacitance to begin with, and their mass provided a built in debounce in the added inertia! :) (just joshing ya Dave)

[maxine_red]

And I see twisted pair for the input to battle noise pickup some!

[John_Ridley]

Honestly, I'd just hook it to an Arduino, and put a 200ms delay after contact (require a 2ms solid pulse to trigger to eliminate the noise problems) before starting to look for the next one. It could be the counter as well. Enable the internal 10K pullup and it probably won't even need any external components. This would take about 10 minutes total to implement, at an entire cost of $5. This is the same as the 555 one shot plus the counter really.

[vilts]

I'd rig a tally counter on the MUT arms. Though it might not be able to hit the counter when battery is low...

[Andrew_Sparrow]

Would still try hacking a cheap $1 step counter from the dollar store, they must have debouncing built-in. When they count the steps it's a simple weighted switch

[symik]

Excellent video as usual nice job Dave, i have learnt a lot from this video, as an EE student. And i have thought that i wont have anything to watch in the time of xmas and new year :D

[capoman]

You could have picked up faster triggers by setting the threshold nearer to zero volts instead of 1 volt like you did. So even if the cap was only slightly charged, you catch those falling edges at the bottom of the graph.