DIY Janko Piano: 5. Initial Prototype
Вставка
- Опубліковано 10 лют 2025
- And so the time has finally come to start covering the actual design of the Janko digital piano instead of going over the background information about the project.
In this episode I will cover the first prototype that I built after planning everything out in Fusion 360, or at least the first working prototype as I had to redesign and reprint almost its entire weight in plastic parts before I got to the point I was satisfied... or more to the point until I realized that any further changes would require almost a complete redesign (more on that later).
Time wise I spent around 6 months total on this portion of the project, with maybe 4 of those months taking up the initial design in Fusion 360. I actually had to cut back on the scope of the project as my original plan included the damper mechanism and a sostenuto mechanism (complete with mechanical pedals to control them) for no other reason than to 'make the action more realistic'. Thankfully I realized that this added complexity was a bit too much before I started printing the prototype, but it did waste a few weeks of design.
In the end this prototype of the digital Janko piano action had a few issues, both in terms of printability, and the sub-optimal operation of the hickman action... but I will cover those issues and how I solved them in a later episode along with the finalized prototype.
Introduction: same as 0:teaser.
Project discussion forum: pianoclack.com...
Not much in terms of 'further reading' here, though if you are interested (and havent watched the previous episodes) here are the links to the concepts I talked about here covered in depth (or you can just watch the entire playlist from the start)
Introduction (general overview of the project)
• DIY Janko Piano: 1. In...
Janko Layout (going over the key layout as compared to the classical piano, covering isomorphic layouts in general and a simplified explanation of how the Janko layout can be 'crafted' from the classical one)
• DIY Janko Piano: 2. K...
Hickman Action (going over the hammer action that I decided to go with; including its mechanics, history, and comparison to the Erard's piano action used in pretty much all grand pianos today)
• DIY Janko Piano: 3. Hi...
Key Levers and Parallelograms (going over the key levers used to ensure equal key dip for all buttons of the key along with comparing it to the simple lever design used in all classical pianos)
• DIY Janko Piano: 4. Ke...
The sheer level of detail, precision and organization in that CAD file alone is outstanding
Got some professional experience in CAD so that definitely helped; though the switch from AutoCAD with its 'each object is a file' to Fusion 360's 'single project file' took a bit getting used to.
Did end up having to re-make the project from scratch for the next version in order to include the rest of the piano, so if you notice organization differences between this one and the next one thats why.
Man this is so awesome. I hope the rest of your development goes well and you sell these one day lmao I'm a pretty good piano player, but I've wanted to try a janko piano for years!
Well, the development is definitely going well; already working on the assembly (currently sanding keys... sooo much sanding...).
As for selling them; most likely no - the amount of time I expect the assembly to take plus the material costs places it in the $6,000+ range which isnt really viable. I think Rico Raven is putting together a more 'accessible' option for a janko keyboard, so check him out if you are interested.
@@DIYJankoPiano Aww damn :(
For me as a piano player, a huge benefit for the janko would be the ability to play larger chord voicings than would be possible on the standard layout, so sadly doesn't look like that happens with Riko's, still excited to see his finished product though.
Yeah, I guess $6k is pushing it, honestly, I'd 100% buy it at Lumatone's pricing though.
Yea, I went into this project with a 'no-costs-spared' mentality after trying out the 10,000$ digital pianos with their hybrid actions and deciding I want that in my build as well.
I might see about scaling down the build to something more reasonable after I finish this project (simplifying the action, making it be MIDI output only so as to remove the mini PC & DAC, etc - basically more of the lines of what Lippens was supposed to be) that would make it more manageable for pricing, but thats all at the 'cool idea, maybe later' stage of development.
Very impressive!
Gland you liked it!
Cool! I'm excited to start following this project :)
Thanks for joining, glad to have you on board!
Brilliant! I’m looking forward to seeing your next refinements!
Well, the next video is going to cover the electronics (as the next prototype incorporates them), but after that will be the episode about the refinements and changes from this first prototype to the final working model. Should be interesting! ... hopefully :)
Great work, Mike! I think it can be sort of 'optimised' as we discussed it briefly but regardless - you have done tremendous job with this!
Thanks! Already working on the assembly (started the video series late enough that there wont be a 3 month wait time between the last 'design' video and the first 'assembly' one), so its just the matter of time.
The algorithm must know me. I was recently thinking about how to implement a simple velocity and after touch sensing mechanism into a 3d printed keyboard mechanism. Didn't get around to testing yet (also because i dont have an action designed) but i bought some components. Not sure about scaling though because of a round robin approach that my sensor setup would require.
Well, you are (just about) right on time then. The episode next week will cover the sensors and electronics, so that might help you out.
I also have a thread about the project (link below) for discussing the technical stuff - feel free to join the discussion:
pianoclack.com/forum/d/1825-diy-digital-piano-a-janko-variant
BTW, I recently got a Hammond SK1. For non-velocity sensitive sounds like organ and mono-synths it’s excellent, due to its high trigger point. I can simultaneously send different sensitivity curves on two additional MIDI channels.
Unless, this is strictly for piano, it could be interesting to have a sensor on the keys, as well as on the hammer target (the nail? lol.)
The high sensor enables organ-specific articulations, like percussive stabs, smears , and light-quick playing. Modern synth keys with low trigger and velocity simply don’t deliver the same results.
I’m wondering if a key trigger in addition to the hammer trigger isn’t beneficial, if not necessary to control the sustain. There might be additional advantages, such as eliminating false triggers, etc.
Given that you are doing a no-compromise build, having it be able to control piano or organ could be good insurance.
Of course, the SK1 has waterfall keys for comfortable glisses. That might or might not work on the Janko buttons. Smears and stabs can be more percussive noise than tonal, so the lack of a diatomic row might not be a problem
Another advantage of the SK1 over my Arturia KeyLab mkII is that the Hammond has nice, wide black keys. The tops on the Arturia keys is small, so it was easy to slip off. I’d imagine that the Janko layout completely eliminates the problem of slipping off of the black keys. I don’t know about glisses though. Can you slide your hand over the keys?
Well, the next video will cover the electronics and sensors, but until then:
Sensor wise, I have 2 optical sensors for measuring velocity of the hammer right at the end of its flight which should correspond to hammer velocity in an acoustic piano with a similar action. This does (as you mention) not work for sustain / dampers, so there is a separate magnetic sensor that measures the actual position of the key. When combined this means I get a single 'velocity' value when the hammer hits the hammer-hit felt (works the same way as in an acoustic piano hitting the string) along with the exact position of the key every millisecond. The rest is pretty much firmware.
I have a 6 step state machine for each key that combines both the optical (velocity) sensors and the magnetic (key position) sensor in order to prevent ghost repeats as well as send key-on (hammer hit) and key-off (key released - dampers dropping) events. I will likely have multiple methods for triggering both of them (such as using the magnetic sensor only for both as an alternative), which will be compatible with variable trigger points if necessary.
I do however plan for this to be used strictly for piano playing though - never played organ, so...
Still - the hardware is there, so its not like its impossible to add extra features at a later date; even setting up multiple simultaneous sensitivity curves along several MIDI channels simultaneously would just be an issue of writing the firmware for it. The bandwidth for the sensor boards can certainly handle it.
Finally - in terms of glissandos along the keys; yes - it should definitely be possible. I made sure that the keys are strong enough (and have minimal side-to-side sway) that both whole-tone and semi-tone glissandos should be doable. Though as always - I will only truly know after its all assembled...