EEVblog

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.
Forum: www.eevblog.com/forum/blog/eev...
EEVblog Main Web Site: www.eevblog.com
The 2nd EEVblog Channel: / eevblog2
Support the EEVblog through Patreon!
/ eevblog
EEVblog Amazon Store (Dave gets a cut):
astore.amazon.com/eevblogstore-20
T-Shirts: teespring.com/stores/eevblog
💗 Likecoin - Coins for Likes: likecoin.pro/@eevblog/dil9/hcq3

[@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.

[@Boffin55]

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.

[@TechyBen]

See next video. :D

[@mrlithium69]

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

[@Jefferson-ly5qe]

If only Thunderfoot was as easy to listen to

[@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.

[@andrewwhite1793]

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

[@UpcycleElectronics]

Hey Dave, I like the short content videos for spontaneous watching, like when I'm looking through my new subscribed content. Your my go-to source for in-depth information as well, especially when I want a visual approach to something like some subjects/chapters from "The Art of Electronics." I really hope you can keep up a good mix of short and long uploads. Thanks for sharing.

[@BerndFelsche]

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

[@firefly618]

Lazy video is lazy

[@Kryoclasm]

great video Dave!

[@brendanzotto1239]

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

[@symik3]

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

[@capoman1]

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!

[@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)

[@magnehaneberg8605]

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!

[@MDFRESCUER]

Nicely explained.

[@MrRobbyvent]

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

[@mikeselectricstuff]

I'd use a magnet and hall sensor

[@PBGBen]

Should have used an Arduino too, as many of us have yet to collect lab gear ;( Bonus is that it can be programmed to count one and then sleep to better handle "Bounce/Chatter" - Or, perhaps a rPi with a pythin script to count and post it to a live page along with other stats like the total run time, number of claps, battery voltage etc.

[@TheEricBooth]

Ok contacting PETFA (People for the ethical treatment of fake animals) :-P

[@toddberg3892]

Eric Booth Call Peter Gabriel- Shock the Monkey!

[@SaturnV2000]

All joking aside, another great tutorial by Dave . . . thumbs up!

[@DoRC]

man you're gonna kill the batterizer testing!

[@capoman1]

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.

[@pepe6666]

damn dave this is a fantastic and relevant video. im about to make a mixing desk. 65 buttons! gotta get this debouncing sorted before i make the same mistakes 65 times over :) thanks once again dave!

[@LazorVideosDestruction]

Preparing for the batteriser I see :P

[@shazam6274]

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.

[@alleycat9884]

You should call it DAD... Dave Assisted Design! My 12yo son came up with that...He's a big fan, and loves his tech........... little smart arse!

[@TheDenzel2112]

One issue with counting only cycles and not individual claps is when the batteries run low and his overall speed slows down, it may end up counting the individual claps as a cycle if he slows down enough.

[@ctrlaltdel02]

Whoa, 330uF for debouncing, very generous value Dave, I usualy use around 500nF for debouncing, works great, and working speed is also decent.

[@MOTOSNOWRIDER]

neato video Dave!

[@vinitshandilya]

Thanks a lot, Dave. That MUT would've caused me a lot of trouble if I were doing the tests! :D

[@KevinDay]

In this case the problem with counting cycles instead of claps is that he slows down as the batteries get low, so eventually the capacitor would probably charge above the threshold every clap.

[@ernststavroblofeld1961]

You removed the cymbals …. The curse of Shamilmaa is now upon you!

[@MrAwyork]

Thanks

[@registratoreprimo9778]

i like these more instructional videos, i really think you should make electronics lessons for beginners, you are a great teacher

[@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...

[@maxine_red]

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

[@SkyCharger001]

I think it would need a little too much play at the hands to work. (or you could go overkill on isolation on that part of the cables)

[@andreasdill4329]

If you set up your system to count cycles you get a problem when the batteries are going weak at the end. The time between 2 claps will rise, and you will count claps instead of cycles.

[@Tedybear315]

Looks like this is becoming a rube goldberg type device. Simple task, but explodes to be come complex to take care of all the side effects. I'd go with just a diode, 5v supply and a simple basic electric-mechanical counter. Very simple and does the job. (we use them on coin switch counting inputs.)

[@code123ns]

Add another diode parallel to R1 - has no use, but the middle part of the circuit then reads: R2, D2 8-)

[@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

[@mipmipmipmipmip]

Andrew Sparrow Probably don't even need to hack it! Maybe fix the monkey on one arm and connect it to the other for extra movenent, or bind it at the end of a stick connected to his arm. Do test runs to see if you need to correct for anything. Then just run it until the monkey stops, read out the value, and run additional samples until you get good statistics!

[@tablatronix]

I just realized Dave has never done an episode on frequency counters practical uses and common use of the tool, eg. maths available and what you would use them for, period measurement, difference etc. In fact there does not seem to be anything like that on the tubes.

[@zusurs]

Should mention that when batteries gets low, monkey's movements will get slower and will false triger counter...

[@vladimirmalic]

Great!

[@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.

[@Ybalrid]

you're right, but you're not doing an educational show about electronics on KZbin :p

[@John_Ridley]

Ybalrid Well, I do, but it's a different channel. And I have nearly one subscriber!

[@carlzimmerman8700]

Good video, would've like to have seen the interference from the chirping displayed on the scope.

[@DemetryRomanowski]

Much simpler way to do it, use a Micro. read the input check in the loop versus last state count up delay bout 50ms and write it to some serial interface.

[@zx8401ztv]

Bloody hell dave, im sure i saw sidney operahouse on the rigol lol.

[@deanrubine2955]

I was expecting the end to be: Put in some new batteries and autocount claps. Put on Batterizer and autocount additional claps. Compute percentage, additional/new x 100. Compare to 800% claimed. Compute lie factor = claimed / measured.

[@josephf151]

hmm, I figured you were going to make a logic de-bounce circuit with a 7474 flip flop. although this method is more educational as to what you might run into.

[@44209david]

Fuck Yes finally the EEV blog is back into doing electronics

[@electrodacus]

The monkey may get tired as the batteries get weak and thus your 1.5V trigger level may not be a good idea since it may start counting the individual claps. Better just to reduce the trigger level or capacitor value and count all claps. Still clapping monkey is not a good measure of energy used since while battery voltage will go down without the batteriser and the current used by the monkey will also drop not to mention compared with a fixed 1.5V output using the batteriser. That is not to say that batteriser has any useful application.

[@DennisXiloj]

More videos with probes!

[@EastyUK]

Didn't MJ get in trouble for probing the monkey? I didn't understand everything in the video, though certainly learnt a few things. cheers!

[@pvc988]

Also for FPGAs there is a neat method involving a shift register.

[@power-max]

You could probably program in a very basic FSM (finite state machine) as a counter, or with some T flip flops, it's trivial!

[@zx8401ztv]

An ultrasonic pickup and threshold detector perhaps?

[@hannonm]

i'd go about adding a Diode for boot strapping. help clean things up.

[@johncundiss9098]

I suspect that noise is comming from the incredibly bad dc brushed motor in the monkey. Expeienced similar situation with simple 12 volt brushed fans. You will be looking at your scope and wondering where this noise is comming from everywhere. Ditch anythinhg with brushed motors where posibble. Slap catch doides across motors if got access to them.

[@DaruoshAghajaney]

i was gonna say why not using micro-controller?!!? ahhh.. you mentioned it... still soft-debouncing is a lot less hassle.

[@gl1500ctv]

Agreed! Analog sucks... :)

[@baprof261]

Why the R2 resistor cannot be placed in series with the switch? On the other hand, a diode won't just complicate things? As far as I can understand C1 cap will charged through R1 below (VCC-0.7), and after reaching that voltage, will charged through R2, giving two separate charging profiles.

[@SaturnV2000]

Need to use banana jacks for power and ground connections.

[@power-max]

would it be beneficial to twist up those long wires or use some spare cat5 cabling for connecting to the symbols to eliminate noise?

[@svampebob007]

_9:25__ "so that's actually pretty good, it counts the cycles" counter is at 16.... __9:28__ instantly jumps to 20. well that's not so pretty good after all :)_ well that's what I was going to comment, but being a EEVblog video you noticed it "20"seconds later, and now that I look at the timer we're only half way through. "quick video" Dave Jones™ Never change that :)

[@mikgus]

needs a livestream for the test :)

[@kennethdeitz4561]

I think this would be a great application for an MC14490.

[@SuburbAllied]

Is this the same monkey from Stephen King's short horror novel "The Monkey"? If so, Dave goes hardcore I see! :D

[@EETechs]

Why not just use a shielded cable to the monkey and tie one end of the ground braid that is inside the cable to an earth ground (do not tied both ends to ground. Only one side) ?That will solve 90% of your noise issues. Do it all the time in PLC systems that have to read analog singles from sensors mounted far away from the control panel.....

[@dsmith5167]

More Testing and Test Setup Videos Please..............

[@zynthos9]

12:00 So in this case the current flows from one end of the cap (which is located right at the red and black terminals of the universal counter) all the way through the monkey hands and then to the negative end of the cap?

[@NoxMarcus]

I guess this video was about debouncing. But doesn't that fancy universal counter have a programmable trigger reset delay, or whatever it's called? I remember playing with that in school, which was a while ago.

[@Landrew0]

It would be easier to measure the power-drain per clap, and divide into the available watt-hours of power available.

[@izimsi]

Counting individual claps is better, when battery gets low, the clapping frequency might get so low, that claps might be counted as cycles.

[@Omgalof]

you cold use a mic and sound wave analyses, but that is a totally different ball park

[@SproutyPottedPlant]

I wonder what the most complicated way of doing it would be? Would it involve air tubes and hydraulics?

[@PcandTech]

things i learn here priceless

[@km5405]

maybe probes the test monkey would make a good automatic component tester?

Every once in a while the hands stop when together, you can see that in the video a few times. That is a problem.

[@cjfdaniels]

Never been so mad/frustrated with a toy monkey. What a jerk.

[@r3bers]

No need rc filter when you have prof 20 Hz filter.

[@kewakl8891]

I put your circuit on DaveCAD(TM) Freeware. I get a netnode error: Not Connected at R2/D/C. I think the free version does NOT have enough connector dots. :p

[@teslafreak]

What's the cost on a single node home use license of Dave CAD? It looks pretty legit.

[@CatNolara]

Isn't the capacitor a little bit too big? Would've expected something around 100n or something like that.

[@KeanM]

Yes, but it depends on the resistor value which was relatively low (I think he said 5k6) and so 100n would have given quite a short time constant

[@CatNolara]

Yeah, but in his version you can see the voltage rising in real time, you really don't need such a big time constant.

[@KeanM]

Agreed

[@OrbiterElectronics]

Put a tally counter between his hands :)

[@berenscott8999]

You know even at the end it wasn't correctly counting.

[@bean2594]

Why not put r2 in series with the switch instead of where it is. This would eliminate the need for the diode.

[@zer3vro897]

Can you do an in depth video on debouncing?

[@adambutt3690]

Was anyone else irrationally annoyed that Dave did not put his scope in free run mode.

[@thetaleteller4692]

How about a smaller capacitor to get a matching low-pass filter?

[@HughMungoose]

Oh god that noise reminds me of Asian market stalls. Why the hell do they have these things going continually?

[@yolofapper9668]

please do Owon XDS 12bit scope review & teardown!

[@DjResR]

What about connecting capacitor between switch output and Vcc?

[@tubastuff]

SPDT pushbutton and a latch. Guaranteed to be bounce-free.

[@0MoTheG]

You mean a RS-FF that is set or reset depending on the closed contract?

[@tubastuff]

@@0MoTheG Yup.

[@God-CDXX]

put a 5 ohm resistor on the switch

[@dgtlmoon]

What's your opinion on using a 555 for debouncing?

[@Soundfactory24]

many other "modern peoples" would now program an arduino for debouncing or looking for a smartphone- app ;-)

[@uriituw]

Removing the cymbals might effect the test data.

[@DanielHeppner]

It doesn't matter because he's just going to use it for comparison.

[@glaubhafieber]

I learned DUT from an aussie when i worked for cochlear, now I learned MUT from an aussie 😹😹😂🙈

[@santopino2546]

"This is a traditional arse" .... uhmm..... oh shit..... RC !!!

[@johnbrass7196]

I would use a 555 timer, to debounce

[@skuula]

Why not just put R2 in series with the switch and save the diode?

[@Biela2008]

DaveCad(R) patent pending... :D :D :D

[@yoramstein]

On Min17Sec09 Monostagble vibrator should come : Courtesy "Talkingelectronics.com" Colin Mitchell's site. Could not you lower the grigger point in order to count all klapping?