EEVblog

[@corri303]

I can confirm that this way of sensing velocity works perfectly fine. As mentioned high-end keyboards like the Roland D-50, Alpha Juno 2, JX 10, etc. used the same method as well. No need to make things more complicated than necessary

[@sp10sn]

What's the method called?

[@mikeselectricstuff]

foam on speaker wires is porobably to stop them rattling against the case

[@johndrachenberg2254]

I miss your videos! I hope that you're well and that you upload more teardowns soon. :)

[@SionynJones]

Yep more teardowns please Mike. Feel I learnt so much from your videos.

[@Fifury161]

Of course - that make sense (seems so obvious now that you have pointed it out) - cheers mike!

[@smilingbuddha2836]

I agree.

[@Mtaalas]

That double switch arrangement has been done for velocity sensitive keyboards since forever, the thing is, that this particular model is really really cheap and the piano sound on it is absolutely phenomenal for the price. (I play piano myself) A really high end keyboards have triple contacts and mechanical hammer arrangement for the keys. The hammer mechanism simulates the feel of a real piano and three switch arrangement allows the note to re-trigger even if the key isn't lifted off completely before re-engaged, like in a real piano can be played. Since in two switch arrangement you have to lift both contacts off the board before you can press the key again, the third just allows two of the three to be off the board earlier then full travel of the key. :) It's necessary for very fast and demanding piano pieces where there are repeated notes in quick succession and there's no time to let the mechanism lift the key completely before having to press it again.

[@thetheatreorgan168]

Theater organs from the 1920s had a feature called second touch where if the keys were pressed hard, they would activate a second set of contacts. Most electric organs were inspired by theater organs, and some of them were built theater style.

[@musicaldracula2017]

This is a keyboard with non-weighted synth action and you are talking about digital piano. Even high-end keyboards such as Kronos61, Montage6 or Genos do not have hammer mechanism. It needs playing a most extended range of sounds with these types of arrangers and hammer action would not be suitable for that particular purpose of these keyboards. Manufacturers have been trying to design some lightish hammer actions for these kinds of arrangers or workstations, namely Roland FA08, Yamaha MODX8, Korg Kross 88, etc. However, lightish hammer actions are not suitable for playing piano properly. The bottom line is that the above mentioned triple contact is special for the hammer action mechanism and this Casio CT-X700 is another type of instrument. I have recently had a Casio CT-X5000 and this CT-X series is good for beginners, P/P ratio is excellent. Cheers!

[@8BitParadise]

Brilliant intro, *Sick beats Dave*

[@EEVblog]

I think I'll give up my day job...

[@8BitParadise]

EEVblog Never had one lesson? Pfftt. You're a Professional!

[@TMS5100]

sounds as professional as any modern commercial pop music.

[@tigerstein]

Reminds me of a pianist in a restaurant a few years ago. He went out to the bathroom and the music kept playing...

[@mpoerot8013]

ROFL indeed that's sick psycho beats

[@mindslain6752]

Almost all synth keyboards use this type of velocity sensing. Some have additional pressure sensitive strip to provide shared aftertouch functionality. Some Ensoniq synths used inductive sensing to provide continuous per key position for polyphonic aftertouch in addition to initial velocity.

[@Sharklops]

That's interesting. I figured it would have been hall effect sensors on each key

[@rich1051414]

Nothing wrong with this type of sensing, in spite of it being cheap. 2 digital signals per key is all that is needed for perfect velocity calculations.

[@Jeff-Russ]

Arturia seem to be going with the Force Sensitive Resistor strip approach all the time and just sampling the current when a new note is played and just discarding the rest (when velocity is the desired parameter.) I like this approach; it's cleaner and you get aftertouch if you want it.

[@movax20h]

There are also keyboards that do per-key after-touch pressure and absolute position sensing during all stages of the press, so they can sense more dynamical behavior of the press and depress (providing very good expressiveness and articulation for good piano players), not just average velocity between two points. They can be also useful for emulating other instruments, like pipe organs, and many more. This is called displacement sensing.

[@boowonder888]

Mindslain, You sound like someone who could know. My casio ctk-671 has a nice sound. The b key doesn't work when you just push it in. I f you slap/pound it hard it does work. Is this a problem with the electronic contacts or is it a mechanical problem? I never opened it up to try and fix it.

[@NebukadV]

That blue strip does most likely serve another purpose: The highest notes on a piano are always lighter (easier to press). I bet Casio emulates that by using a slightly softer rubber for the highest octave (12 keys).

[@SionynJones]

So like tactile response.

[@happyrat1garyg.510]

The blue strip is as Dave noted, a visual indicator for the assembly line that it goes on the top octave which has 13 keys instead of the standard 12 of the rest of them. Not all manufacturers are so considerate. :)

[@NebukadV]

It could be both - or do you own one of these to prove, that all keys feel the same? There are easier aka. cheaper ways to make a visual differentiation.

[@happyrat1garyg.510]

Watch this video about an 88 key Kurzweil which I happen to own along with several other synths. kzbin.info/www/bejne/b2KqepuwarmVkLM There is always a High C key above the top octave on virtually every keyboard instrument. Since an Octave is twelve notes if they simply built a 5 Octave or 6 or 7 Octave keyboard with uniform parts they would still have to include a single rubber boot for the top note. It's more practical to have a single strip with 13 notes instead of a series of 12 plus 1. Furthermore on 88 keyboards the lowest notes are only a half an octave or so so there's a special strip for that one too. As a player I can tell you that the top octave on all keyboards except for graded hammer action keyboards have the same weight across the entire board. And on graded hammer action Boards they use variable springs and cam mechanisms to achieve the desired feel. But I can assure you the membrane contacts are just that. Dumb membrane switches with no fancy graded elasticity. If a board is advertised with graded hammer action that is the only time you will find a lighter touch on the high octaves and a heavier touch on the lowers. And typically you will only find these actions on specific electronic pianos and high end synths. Gary ;)

[@WaltonPete]

This keyboard doesn't have weight distributed keys, to mimic an acoustic piano, it only has velocity sensing which correlates to the force used to operate the keys.

[@jimgordley6169]

I'm glad you did this particular tear down. I just ordered a CT-X700, now I don't have to fight the urge to take it apart to see how it works. As far as your piano skills you and I are on par. Thanks for the informative video.

[@TheDefpom]

@ 16.30 those strips would be to hide the joins so you cannot see the pcb between the keys.

[@happyrat1garyg.510]

Just to confirm what others have been saying. The dual membrane contacts are a music industry standard for velocity sensing keyboards and has been since the late 80's. As was also mentioned there is occasionally a pressure sensing strip on higher end boards for channel aftertouch. Basically the rule is, if it's cheap and it works don't screw with it. Generally these sorts of contacts have a functional lifespan of about 20 to 30 years or so. A common point of failure on older boards occurs when the suction of the silicone cups sucks dust and crud into the contacts causing either the key to fail outright or else the velocity sensor goes out of whack and the key always plays at full velocity. Simple fix usually entails removing the strips and cleaning the contacts with denatured alcohol. I'd love to see a board with optronic sensors instead of membrane contacts to address this problem but as you said. If they can save a nickel on a switch the cheaper solution always wins :P

[@happyrat1garyg.510]

Just to drive the point home. Here's the keyboard of a $4000 Current Korg Kronos. Same membrane design. kzbin.info/www/bejne/bKnNi2qHbL-siJI

[@barbudoru]

My girlfriend's PortableGrand NP-30 did exactly that, sucked some tiny amount of gunk under a dome so the same key on all octaves was stuck at full velocity. After removing the gazillion screws (metal reinforcing strut on the back tho, not plastic like the one in Dave's video), cleaned up everything with alcohol and now it's working like new.

[@happyrat1garyg.510]

Yeah, digital pianos and keyboards come in three types. Unweighted synth action (Like the Casio in Dave's Video), weighted action and fully weighted hammer action (Like the Korg in my first reply) . Electronic pianos usually have fully weighted hammer actions with counterweights and a cam type mechanism to simulate the striking action of a real piano. These can be a real bitch to work on but basically the sensing mechanism is the same on all types. I'm glad to hear that your girlfriend's unit was repaired OK.

[@smilingbuddha2836]

It seems to me that this method was pioneered and perfected by the Japanese.

[@happyrat1garyg.510]

I have no idea who invented these contacts but if I had to guess I'd guess Yamaha or more likely Casio was the first to start using them. I'm too lazy to look up the original patent though :)

[@eviloatmeal]

There is indeed a musical term for that - we refer to the black keys on a piano as "black keys". Those carbon pad boards look very similar to the ones in my M-Audio keyboard. I would be shocked if I found a budget instrument with any other technology for the velocity thing.

[@rich1051414]

Velocity only needs two digital signals. It just measures the time delay, and from this, you can get perfectly accurate velocity.

[@nt654321]

This kind of construction with a pair of carbon contacts is actually the most common way of making a velocity sensitive keyboard mechanism, even in very expensive high-end instruments. It was also used widely back in the 80s - I own several different vintage synthesizers which use the exact same kind of mechanism (such as a 1986 Roland JX-10), and apart from a sturdier metal construction and a resistive aftertouch strip, it is exactly the same as this cheap Casio.

[@EEVblog]

Cool, now we know nothing has changed in the latest models.

[@electronash]

EEVblog A far better system (IMHO) is where they have top and bottom "bus bars" on groups of keys, then there is generally a spring (or spring wire) that sits between the two bus bars. The springs are all in contact with top bus bar normally, and the velocity is timed from when the top contact is opened and the lower one is made. The membrane type is used on way too many keyboards now. I have a Roland Juno-D (2004?), which was marketed as a kind of semi-pro stage keyboard, and that still uses a membrane. I had to replace all of the membranes and contact PCBs when I bought it second-hand, as half the keys weren't working. In the 90s, we found a "Cheetah MK7VA" MIDI keyboard at our local dump. It used a Z80 and had full-sized keys with the proper bus bars. It was amazing to play. I tried to relive that recently, by buying a Cheetah MK5 on eBay, thinking it was just a smaller version. Turned out to have a damned membrane. Oops. lol

[@electronash]

EEVblog See this vid around the 4-minute mark. The keyboard has separate loops of wire instead of a full lower bus bar, but yeah, same principle. Obviously the membrane type is just a lot cheaper. Just a real shame they are so commonplace now, even on expensive keyboards. kzbin.info/www/bejne/sGiooIFjh9B9bMU

[@Sumida2]

Not sure I would call it better. New keyboards with bus bars have been virtually nonexistent after the silicone membrane type took over in the 1980's. The metal wire oxidizes and stops working well after some time, but the carbon contacts get dirty too. Nothing lasts forever.

[@chinesemusic8019]

The Yamaha keyboards from the DX7 (1983)to the Motifs (of 2003 vintage) still use the metal mechanical switches. The silicone switches and carbon contacts will degrade over time. The Roland keyboards have the silicone switches. That's why Yamaha keyboards cost more but last . My DX7 bought new from 1986 is almost 32 years old and still working (apart from replacing the power supply). It's the only surviving working electronic device from the 80's I have.

[@w0mblemania]

As someone who didn't know what a scale was (or a musical Key for that matter) until quite a few decades in to his life, can strongly recommend that you take up the piano. It will help your brain as you age... learning new, complex skills that combine motor coordination, memory, etc. It's interesting how many benefits it has. It's given me a new appreciation for musicians and musical engineers. It also imposes a level of discipline and provides a reward system for that discipline that you often don't get in daily life.

[@johndrachenberg2254]

It sounds like you might enjoy the KZbin channel 12tone.

[@stuartmcconnachie]

Masking so you can’t see the green PCB through the gaps between the keys. Should have used a black PCB, but probably costs more?

[@EEVblog]

Perhaps it does on this mass scale. They already have the carbon process so maybe they get it for free effectively.

[@nrdesign1991]

Looks like the top overlay/silkscreen layer to me

[@stuartmcconnachie]

nrdesign1991 Ah yes, it’s the same colour as the silkscreen - so is effectively free. I’d missed that.. I’m sure Dave just mistyped. It’s obviously not the printed in the carbon process layer.

[@Psychlist1972]

Velocity sensitivity is done by timing the interval between the two contacts firing. Even the high-end keybeds do it this way. There are other ways, especially in older stuff, but this is definitely the way the majority are done today. Many older keybeds (80's, 90's) need the carbon replaced on those contacts. The more expensive keyboards let you adjust the timing curve between those to change the response of the keybed. Aftertouch (technically called channel pressure), which this doesn't do, is usually accomplished through a sensing strip that goes across the full length of the keybed. This is why the vast majority of keybeds only do *channel* pressure, meaning they apply the pressure to every key, vs individual or polyphonic expression. That limitation has gotten under enough people's skin that we recently voted to approve the MIDI Polyphonic Expression standard (MPE). Where did they save money? - Main thing is reuse across their product lines. - Keys are in octave units, so they can make the keybed as big or small as they want. Likely reused across many of their consumer keybeds - Printed carbon resistors for the board (because they already had to print carbon). - No springs on the keybed. They just used the plastic hinge effect. - No weights on the keys - No poly pressure on the keybed - No classic DIN MIDI input or output (saves a couple connectors, an opto-isolator, and a serial connection or two) - Surely reuse that logic board across a number of their keyboards - Cheap screen :) Other than any labor, the most expensive thing here that is specific to this one keyboard is the molding for the bottom of the body. I bet even the top is sectional so they can reuse the speaker moldings across bigger/smaller and only need to replace the center molding

[@SkyCharger001]

About your question regarding the labeling inside the keys. I've had some piano-lessons as a kid and I still remember the basic octave-layout: C # D # E F # G # A # B the key-section you first took out provided the D, F and A keys, so the labeling simply identifies the notes they are meant for. (simply a case of better safe than sorry)

[@SpeccyMan]

You left out all the flats, lol

[@moristo]

start on 2:12 .... in one tone key there are two rubber contacts (1st and 2nd contacs). the distance between 1st and 2nd will be a logic (timing), the stronger you press is to faster the timings, and the voice will gets stronger

[@trptmbalmer]

The letters stamped into the backs of the keys you mentioned refer to the pitch name of the key. You were correct. The boards are also labeled to provide a guide for where each octave of keys goes (note the big "C3", "C4", "C5", and "C6" printed on the tops of the boards). As for the odd row of 13 blue keys at the end? That's for the top octave of the keyboard because a standard keyboard includes a final, 13th "C" key at the very end.

[@MrMyneworld]

13:00 All low price keyboards use this kind of keys, I would like them to make the joints with steel strap, so they would never break. When a model becomes old, it may be impossible to get the matching keys.

[@SidneyCritic]

I just realised, you could slide a thin double sided PCB between the battery and it's terminal for a meter to read the current consumption.

[@Graham_Langley]

Made up one some 30 years back and it's still in use.

[@HebaruSan]

The alternating moldings for the white keys: to avoid adjacent keys sticking to one another when they're being molded?

[@fpham8004]

Not for sticking, but the mold would need to have very thin sides between the keys and soon or later it would get damaged or wavy. This is a clever way to have solid mold that can produce tens of thousands of keys.

[@dannyarnold9823]

I think if you look closely the rubber membrane is stepped, with the front contact being shorter than the rear one.

[@movax20h]

Dave, it is not about the angle of the key determining which contact is pressed first. If you look closely into the rubber molding, you will see that there are two pins, and they are of different length / depth. The longer one will make contact first with carbon tracks on a PCB, the shorter one will make a contact next. Casio (and their ASIC) knows geometry, and difference in heights of these pins, so can easily calculate absolute physical velocity of the key, and calibrate the rest, and do processing.

[@EEVblog]

I must have missed that.

[@FindLiberty]

exactly: Those carbon tipped silicone nibs are different lengths. The long one always makes contact first, then time is measured 'till the shorter one also makes contact (if at all) and that result is translated to a velocity value to change the voice from soft to hard.

[@movax20h]

Would be interesting to connect a scope or 2 channel universal frequency counter (possibly with Schmitt trigger in-between, and only trigger on one edge) to one of the keys to determine and do experiment on timing difference between pulses vs speed. It should be a simple (inversely) linear behavior.

[@johnnodge4327]

The resistance will change with the amount of pressure applied to the key. The processor can work out how hard and fast the key has been struck, simply by analysing the change in residence curve. Games console controllers use the same technique to determine how hard the player is pressing the buttons.

[@movax20h]

John Nodge, no, it will not. Not on this keyboard.

[@dentakuweb]

Yup, that's pretty much how they all work. It just measures the difference in time between both carbon contacts being pressed. The most interesting pressure and position sensing devices are on much more expensive instruments like the Haken Continuum or the LinnStrument. Roger Linn talked about designing those pads during his interview on Embedded.fm podcast. www.embedded.fm/episodes/236

[@WaltonPete]

I have opened up a few keyboards in my time and I believe that the sponge strip on the back of the main panel is purely to reduce vibration against the case when it's being played.

[@PrincipalAudio]

Casios are the best! If you run some on low voltage, you can get a kind of "circuit bending" effect on them. Random sounds playing when you press a key. Something to try. :) EDIT: I used to plug the audio output of my casio keyboard into a control for a radio-controlled car to broadcast music on both the AM and FM bands (on several broadcast harmonics). I could broadcast music to the whole house and even managed to pick it up on the television, albeit with very low reception.

[@fpham8004]

Yeah, sadly these new ones are too much computerized for being circuit bend, it either works or not now. Pity, it would be great like in the old days when you would solder pots across some guess points and it would totally freaks.

[@frollard]

17:45 the stepped pads are visible on the carbon side - in the silicone mold one is shorter than the other. The key pressed both at the same time but they are different height. (Visible at 17:25)

[@joacimwennerberg8310]

Thanks Dave. More stuff like this, please!

[@ot4kon]

You are not the 8bit guy!!!

[@beerbaron9802]

Pero 8bit guy esta de vacaciones....asi que puede ser un reemplazo momentáneo.

[@ot4kon]

Dampir Zero ya veo

[@thomasandrews9355]

I love the 8bit guy, but sadly you meant 8bit keys. lol

[@TanjoGalbi]

8-Bit Guy/Keys would not review this keyboard as it's too modern! If you watched his videos you would know this. It's why he uses 8-Bit in his channel names :P

[@michvod]

He will be reviewing this keyboard in 2048 :D

[@TheDefpom]

The plastics being done in sections will also be to reduce costs, they can then use smaller moulds (which are easier to make and tune) and can also run in smaller injection moulding machines. Also some parts could be used across different models, effectively reducing the cost even further.

[@EngrJemarME]

Thanks. Troubleshooting done by myself after seeing how it's safely opened up. Wonderful content mate.

[@digitalhands6304]

Simple, but clever way to create velocity effect. I like it.

[@fpham8004]

I have to say I enjoy your enthusiasm. Makes the video much better. And somehow you are remind me of my london born mailman. Always happy to chat.

[@squeakD]

Glad I saw this vid. I’m a pianist and keyboardist, been playing for over 35 years now. I’ve owned many keyboards over the years. One of my current keyboards is Casio’s highest arranger workstation (MZ-X500). Casio has been using this keybed across their line for years now. It’s considered one of the lowest quality set of keys out there too. I’ve been wanting to take my very expensive Casio apart (which costs $1,100) to see how cheap the keys really are and your vid saved me the hassle. I’m shocked at how low the quality is on those keys! I know a lot of keyboard players that own Casio’s more pro oriented models that would love to see this vid just for the disassembly of the keys. We knew they were cheap, but damn!!!!!

[@PeterCCamilleri]

My guess is that all the bits of foam are needed to keep parts from buzzing when notes near their resonant frequency are played.

[@BlackburnBigdragon]

I love to find old Casio keyboards that people junk at the local consignment shops and circuit bend the hell out of them. The keyboards are junk, but circuit bending them at least makes them into something interesting and fun.

[@BlackburnBigdragon]

I wouldn't know where in my car to find the car horn. Haha!!! As a rule, I don't rewire or short circuit anything that has a power supply greater than plug in batteries. AC current is too damn dangerous for what I do. Usually, I grab battery powered elecronic gadgets that make sounds (radios, kids toys, anything I can find that's battery powered that makes sounds), rip the things open, then I'll poke around on that circuit board until I'll get a some neat strangeness out of that speaker, setting up a switch, button, knob, or touch resistor to trigger it. I'll also rip out components or put new components in. I might rout signals from other cannibalized circuit boards just to see what they do. There's no rules. I end up with some crazy instruments. I have this one radio that plays like a theremin using a glove with a photo cell on it. It also has metal knobs that you can touch and alter pitch with using skin resistance. It also has sound cutoff and vibratto buttons, as well as pitch limiters and all kinds of other effect goodies on it.

[@BlackburnBigdragon]

Do a search on KZbin for "Circuit Bending", and you'll find LOTS of interesting strangeness that people have twisted up. Some really amateur, and some, just absolutely out of control and just head spinning in their complexity. As an electronic musician, I've gotten sounds out of some of these monstrosities that I just can't get out of commercial synthesizers.

[@PrincipalAudio]

13:31 - "D", "F" and "A" refer to the notes they are. The "56" I'm not so sure about.

[@Fifury161]

Nice change of pace (no pun intended!), it's good to see how they keep the cost down, such a simple mechanical solution...

[@PrincipalAudio]

17:25 - These looks like time-delayed carbon pads to detect the "velocity" of the keys. The faster you press the key, the less of a time delay there is between the first and second one making contact, so it produces a higher amplitude note.

[@btizef2008]

Very informative. More videos like this showing how they get the price point down is a thumbs up from me.

[@nrdesign1991]

17:33 Had that same exact kind of construction on a 1993 Roland E-16, just different mouldings.

[@MrKillswitch88]

If you want a real treat to teardown is a Yamaha CS80, analog core with a digital back end and only 800 units were made around the end of the 70s.

[@JMacQ77]

And now, for your listening pleasure: the fantastical musical stylings of Dave!!!

[@ranzee]

Always enjoy electronic musical instrument teardowns. I've done a few myself. The velocity keyboard configuration is very common, and has been used for at least 30 years. I did a video on my channel repairing the keys on a Roland GAIA SH-01 synthesizer, and you'll see the same 2 carbon contact pads under each key. The interesting difference is the spring mechanism on the Casio is just using the plastic, as you point out may not be a long-life solution. The Roland keyboard I repaired had actual springs, and although it is also very "plasticy", it has a little bit more quality to it. Let me know if you want me to link it :)

[@Graham_Wideman]

Interesting: kzbin.info/www/bejne/o3XMqpicmtmneLM

[@smilingbuddha2836]

For not working on key beds, you figured the workings very quick. Almost all keyboards use this kind of arrangement since the mid '80s. For aftertouch they typically use a force sensitive resistor strip that each key can make contact with which allows variable modulation control depending how hard you press down on any key AFTER you initially TOUCH it.

[@virtualjd6824]

Dave, if you look closely at the dual carbon contacts under each key, you should see that one contact is maybe a millimetre closer to the pcb than the other (shown at 17:19 on the video). That’s where the offset is, and the timing between them is how they track velocity. I’ve got an old Rhodes (Roland) 660 that I’m repairing at the moment, and that has an almost identical setup, and also has a membrane that senses all the key’s backstop position for aftertouch.

[@Bluuplanet]

Moulded channels on the undetside: That's a structural consideration. It "corrugates" the case for stiffness and provides two vertical walls to support the piano key fulcrums. If there weren't channels in the bottom of the case, there would have to have been either a single flimsy thin walled fulcrum bar or a separate moulded fulcrum part that may not have been as stiff as using the whole case.

[@sefarkas0]

Foam maybe for buzz rattle control.

[@jamesvalentine925]

You should try and get hold of an M-Audio Axiom series midi controller (such as the Axiom 49), not only do the do velocity with their keys they also allow for after touch for vibrato. When you monitor the MIDI output they are constantly measuring the pressure on each key so you can change the volume of a synth on the fly or add extra effects as you add or remove pressure.

[@hellcoreproductions]

The Axiom uses channel based aftertouch so likely there is a single strip that runs the length of the keyboard that aggregates the pressure over all the keys at once. At least this is the way I've seen other keyboards implement it. Polyphonic aftertouch though, that is more interesting. :)

[@jamesvalentine925]

That's some cool information :D I have an Axiom that I've had apart a few times, so far I've not had to go into the key mechanism, perhaps I should explore a bit further into it and pray that the keys all go back into place again. That said though, it's been a nightmare of a controller to own (the reason it's been apart a few times).

[@Sumida2]

Just like 99.9% of all keyboards with after touch, the Axiom does NOT measure the pressure of each key individually. There is simply one long pressure sensitive strip across all the keys.

[@EternityofNight]

HidekiSamba is correct, it is an aftertouch strip that goes under all the keys.

[@generuffalo4374]

I remember requesting a CASIO KEYBOARD TEARDOWN just a few weeks ago on *EEVblog #1102 - Casio FX730P 1980's BASIC Scientific Computer* and _YOU HAVE DELIVERED!!_ I feel like a celebrity now (well, not really) Any how, YOU ARE THE MAN!!

[@WaltonPete]

The alphanumeric markings on the backs of the white (major) keys denote the note within the diatonic scale (A to G) and presumably the black (minor) keys have similar markings.

[@QrchackOfficial]

The two-contact method has been used ever since velocity sensitivity came along, and it's used in every single keyboard because it simply works. No one bothers with distance sensors, that's just silly and expensive. They work exactly as you guessed, measuring the time difference. Actually, exactly the same thing is happening in electronic organs, just with a single contact because there's no need for velocity sensitivity there. The ASIC is most likely doing what everyone else's doing, especially at the cost, which is playing back old regular WAV files, recorded from a grand piano for each key separately and at 4 different velocities. They're then played back and crossfaded together for the intermediate velocities. There's also likely a DSP there for reverb.

[@danmenes3143]

Would be cool to see some analysis of the design. How is the matrix organized? Why do you need the diodes? (Ans: because you need to separate multiple keys pressed simultaneously). What is the scan speed? What do the waveforms look like on a scope? What is the range of delays between a "soft" sound and a "loud" sound? Can you hack in some relays or opto-isolators to trigger "soft" and "loud" sounds with different pulse timings?

[@danmenes3143]

What is the time delay between the second contact closure and the beginning of the note? Or between the contact closure and the midi waveform?

[@Wander1236]

That's some clever cost-cutting construction there. It won't sound great to a professional or hobbyist, but for someone who would just use it to practice piano (or for a children's piano lesson recital), this is pretty great.

[@ericofcrows]

I have a relatively high end stage piano that uses the same method to detect velocity.

[@p_mouse8676]

That board looks EXACTLY the same as a cheap "non-brand" I repaired a few years ago. Unfortunately I forgot the name of that one, but looks like Casio also rebrand stuff or maybe they all use the same "datasheet schematic"

[@tone2106]

I just picked one up the other day, for the quality of sounds the price is hard to beat.

[@sergelavallee913]

2 contacts for if you hit the note hard you get a lower volume so only the one that is not the same size, longer one touch and if you really hit the note hard than it's the other one that touch, shortest one, and the sound is than louder ( Louder input). Usually it's for that reason, i'm not saying that this is the case for this one but usually it is the reason why there are 2 of them different longer. Thank's for your very instructive videos Dave.

[@danielsargent5695]

I enjoyed this simple teardown! Hopefully you get some more interesting keyboards to look into!

[@scottfirman]

My first teardown of something like that was a Concertina I found at the local dump. My dad actually let me BRING something home from the dump. It had some sticky keys and I decided to try and fix it. The bellows were made of cardboard, tissue paper and cloth coverings. They used beeswax to hold things together. Sadly, I forgot and left it outside over night and it got rained on. I decided to leave it out all day in the hot summer sun to dry it. Well instead, the wax melted and everything fell apart. Apparently, the glue holding it together was water based becase it immediately began to fall to pieces as soon As I picked it up. Then the wax got hot and melted, causing it to stick together. The reeds inside fell out, the bellows un wrapped and the keys all fell off the board,and low and behold, all these tiny springs began jumping out from under the keys. This explained why the keys would stick and as soon as you worked the bellows. My parents actually bought the octagon shaped one for my sister for christmas a. I actually learned how to play it as it was much like playing my harmonica, without blowing into it. I got pretty good at playing it too. I sold it years ago now. I still have my harmonica and once in a while, haul it out and play it.

[@StevenOBrien]

Didn't know you could play Schoenberg so well, really impressed.

[@MSP_TechLab]

I think that white silicone thing under the key is used to slow down key's up movement. Similarly as it is done in cassette deck.

[@brainfornothing]

Very interesting how they measure the pressure (=velocity) on the keys. Thanks for sharing !

[@Tedd755]

Did a quick browse of the comments, so couldn't find an explanation for the carbon resistors underneath the keys. I'm willing to bet that they are used for multi-level logic. You can push down as many keys as you like and the keyboard registers all of them, like a computer keyboard with n-key rollover. And yet there aren't 61 (with velocity, 122!) wires going from the keyboard to the processor, so resistors are used to divide the logic into multiple levels. I've seen this before myself, disassembling my car's steering wheel. All the buttons on the wheel are going back to the computer through one wire, with resistors on each to allow the computer to deduce which buttons are pressed.

[@dosgos]

can you explain a bit more here?

[@Tedd755]

Not very well, but I'll try. At a basic level, a switch connected to a processor (and the other side grounded) would either be open or short, so the voltage seen by the processor would be (for example) 8V or 0V. A system with multi-level logic would have multiple switches, each with a different resistor in series to put its unique load on the input line. Maybe in a binary arrangement so the logic levels could be 0V, 1V, 2V, 4V and 8V. That example would encode the state for 4 switches on one line. Any permutation of those four switch states would be a unique voltage that the processor sees, and can subsequently decode. To double the capacity of the single input, you double the maximum voltage of the system, though that gets impractical pretty quickly, but that's the basic idea. You could likely use 0V to 5V in a more efficient way to cram more inputs on the same line but I'll leave that to others. ;) The keyboard wouldn't have all the keys on one line. They'd divide it down to maybe an octave shares an input or even less (remembering that each key has two switches, for velocity detection). Incidentally, a similar idea is used for solid state memory (flash). Most drives are using MLC (Multi-Level Cells). Each cell of a drive stores a charge that corresponds to two bits of data. 00 could be 0V, 01 could be 1V, 10 could be 2V and 11 could be 4V.

[@Sumida2]

Nice thinking, but no... the carbon resistors are simply used as jumpers. This is a classic diode matrix like used in the 1980's. The difference being that they have used SMD dual diodes instead of through-hole, so there is no room to run traces between the diodes.

[@daryltownsend]

Hey Dave ... not sure if you spotted it but the PCB of the volume pot looks like it was manufactured as a whole with the board; routed out and rat bites put in so they could easily snap it off and flip it over.

[@MatthewSuffidy]

I heard the keys had this time sense based on the 2 sensors. I think by definition, there must be a time where the first one is being still pushed in as the 2nd is moving. So I guess at least one of the switches has an elastic on state, that has to be durable. I've got an M-Audio 88es, a Casio LK230 66 key with an octave less, but it lights up when you play midi through it or play it, and a Casio CZ-101 with small keys. The CZ-101 is pretty useless but you can play a lot at the same time because the keys are so small. It has some nostalgia value because some early DM used it.

[@ChurchOfTheHolyMho]

I'll agree that a surprising number of bands used the cz-101 back in the day.

[@GadgetBoy]

The keys are manufactured in a chromatic octave. C, C# (D♭), D, D#(E♭), E, F, F#(G♭), G, G#(A♭), A, A#(B♭), B, then it rolls onto the next section of keys back to C.

[@GadgetBoy]

P.s. The letters on the bottom of the keys corresponded to the note. As well, the C3, C4, etc on the PCBs denote (no pun intended) the octave.

[@frollard]

8:50 I would bet without that foam the plastic would rattle against the outer casing adding buzz whenever the natural frequency is hit.

[@Lalo-rh7dp]

Excellent channel! Greetings from México

[@lasersbee]

18:13 It would push on the hole closest to the pivot point of the key first.

[@electroumit]

Thank you again. You are awesome.

[@Ghozer]

The bottom 3 sections are removable for cleaning, the keys will get sticky and dirt etc will get stuck under during use, take those off, clean the hair and dirt, just like new again! :D

[@jabelsjabels]

Same style of mechanism is used on the Yamaha DX7, one of the most popular professional synths of all time, even!

[@PlasmaHH]

Maybe they use the change in resistance from pressure on the pads too, at least as a hint

[@spahndirge]

From memory, the Sequential Prophet T8 has a fairly convoluted optical sensor setup for velocity sensing and polyphonic aftertouch. Quite a bit less attainable than a Casio consumer keyboard unfortunately.

[@chetancrasta]

Very informative and interesting. Thanks for doing this teardown.

[@EpicLPer]

Ohey, EEVKeys uploaded!

[@athf226]

Probably been said a lot, but that's a pretty bog standard velocity detection system, and it tends to get the job done pretty well. Just as you had said, you just use the timing of the two presses to get a delta for your velocity.

[@boggisthecat]

High expectations for US$280, Dave. These things are amazing examples of cost-reduction through industrial design. I didn’t realise that the two ‘sensors’ ;) were quite that close together, though. I would have thought that the variability from key to key would be quite high by having them so close together - my expectation was at least 15 mm between each. But I guess it is ‘good enough’, and reduces the cost a bit. The considerably more expensive ‘digital pianos’ have clever key designs, but they’re a huge PITA to disassemble. If this one gave you trouble and you decide to pull apart one of those, then maybe hand it off to David - ‘student job’. :D

[@Sumida2]

No spacing is required, since the height of the carbon pads (inside the silicone) is different for the two contacts. In rare cases, the two "sensors" are even combined into one with separate contacts inside, like on this (1996) Yamaha audiodestrukt.wordpress.com/2012/05/19/yamaha-cs1x-teardown-and-repair/

[@JerryWalker001]

The two pads are probably not the same height rather than relying on the angle of the key. Then as you said it is just a start / stop timer.

[@ThePoxun]

I Have an EZ220 Casio which is a couple of year old model which i picked up broken for next to nothing and repaired by re-soldering a couple of bad joints. The construction of the key mechanism/velocity control etc is identical. The only real difference is the pannel section and the processor board. I guess they are making a huge saving by reusing already designed (and paid for) moulds etc. Actually not quite identical... in the EZ220 they still used connectors on the ends of the ribbon cables

[@TimSavage-drummer]

The D on that key moulding is likely because the first key was a D, I noticed on the PCB itself C4 and C3 this refers scientific pitch notation or how to describe a particular note. C4 is known as Middle C (~261.63 Hz), the 4 is the octave number eg C3 is an octave lower that C4.

[@JohnBurgessMusic]

"Good for the price" - well depends on what you think they're good for. I wouldn't give them to a student unless they were kids wanting to try piano. These are very common in schools where they get ruined quickly. I've repaired many of them all with the same type of failure mode - the connectors have no strain relief, no connection to the chassis, and all the mechanical stress goes straight to the PCB, which in the best case cracks the solder, in the worst case cracks the board. Its such an annoying cost saving feature, since its not like consumer goods like a TV where you might only occasionally change plugs. On musical equipment you're changing them constantly.

[@AndreasA.S.]

you can get velocity with this keypad, but i dont see any way to get aftertouch (the pressure on the key currently being played) typically used as an per key effect in electronic keyboards. most classic piano players dont like the plastic keys, they push back as basic plastic lumps, thats why when you see "weighted keyboards" a specific amount of metal weight is added to each key for a more realistic feel of key play, it effects the velocity and does provide a more realistic play. the D labeling was most likely for the starting key. D, im sure the one under it would have been labeled C (corresponded to its first key in the scale).

[@murilocesar5080]

I have a ctx700 and always wandered how it was inside. Thnks for your video.

[@stonent]

So my prediction was a rubber pad striking a piezo strip, then the shape of the strike signal was then applied to the sample. I suppose this required infinitely less cost though.

[@orondf343]

Reminds me of the Casio CTK-4200 that I used to have a few years ago, it had similar features for the price

[@zusurs]

I'm wondering why they don't make something like infrared LED and photo receptor next to it - so that the closer the key is to sensor, the more voltage is sensed. Then just use timing difference between intervals while measuring voltage, and you have precise position detection of each key.

[@stephenwise9131]

Way too expensive!

[@bigblackbadger1]

9:24 I think 8 keys together (from C to the next C) are called an octave, so from C to B is probably called a heptave lol

[@penguineer]

Looks quite like what I see on my Yamaha Clavinova (CLP-150), except for a more complex mechanic to mimit a piano key (they basically build the hammer mechanic but ignore the hammer and take the measurement as shown in the video). You mentioned the problem of getting the plastic right: Yamaha failed on the Clavinova in this regard, i.e. keys tend to break after a while and become sticky. You can buy replacements for 15€ each plus assembly, if you cannot to this yourself. So the plastic is a weak point, if not done right. However, after almost 20 years I only had to replace one key.

[@zxchienowhere]

Is it possible to bring the place the keyboard motherboard in another piano

[@hikaru-live]

The keys are manufacturers in molds sized in octaves, and the keys are divided into thirds.

[@griftereck]

I had a Casio CT650 that I bought at a local charity shop. cost £7.50. was a good keyboard.

[@BogdanSerban]

I once opened a Korg Microkey midi keyboard and it had the same design on the keys

[@DJignyte]

That was interesting, cheers Dave!

[@ojustaboo]

As others have said, the use of the twin rubber membrane has been in use since forever and is still used today, even on really expensive models. The better ones now have triple membranes (sensors) although the three aren't always connected together, so that you don't need to fully release your finger before playing the note again (needed for playing the same note over and over really quickly). A new £3k digital piano will have all sorts of things done to the way the keys press to get as close as possible to a real piano feel, but the triple membrane will still be the final bit pressed. The reason the PCB is split into sections is because manufacturers often release the same model in different sizes, the most popular being 49key, 61 key and 88 key (same size as standard piano) The second paragraph here explains the triple sensor www.kawaimp.com/mp7se/detail/touch/ other examples casio kzbin.info/www/bejne/gp6nY5JvdrV1l6c kzbin.info/www/bejne/aKO9coOmfaidiNE

[@tarheels100]

Is it just me or are the switch contacts different lengths? That would make sense for offsetting the contact timing.

[@finntv4280]

Hello. How to fix a noisy keys? I have a brand new CT-X3000 and some keys are noisy. Others are not. Please help me. 😩

[@Robonza]

It looks to me like the dimples are longer on the back row than the other which makes them contact first.

[@nrdesign1991]

10:00 And it's even part of the big PCB, just snapped off and flipped around