The Curve Steepenates

Despite a few good days of working in the shop, progress has all but ground to a halt.  I had a minor triumph with the following video: actual noise from an actual bell when you press an actual key!

Sadly, the excitement was short-lived.  All that wobbly misguidedness just wouldn’t do, so it was back to the drawing board.  I spent this morning trying to figure out how to add more stability without throwing the basic alignment out of whack, but couldn’t come up with anything that works.  This afternoon I spent many minutes downstairs at the piano, flicking the hammers and wondering at the perfect alignment they achieve.  Effortless motion… no wobble at all.

I feel like I’m heading back to the drawing board.  Wiser, yes, but still: drawing board.  I’m very much looking forward to the day(s) when I can finally (confidently) lock in a design and start cranking out 40 identical pieces of piano over and over and over and over again.

A Ding And A Thud (and a Ding!)

I did some brief tinkering in the workshop today while Ruby played in the backyard with a friend.

In order to amplify and reinforce the sound coming from each bell (the metal plates that make the noise are called bells, even though they aren’t bell shapes), one puts it over top of a precisely dimensioned box (or tube).  A phenomenon known as Helmholtz Resonance takes over and amplifies the specific frequencies you want to hear.  You can see these resonators hanging down below large concert xylophones and marimbas.

Anyway, I’d done some quick calculations today to figure out if I was in the ballpark for the dimensions I’d need for my own resonators, and things were close, I figured I’d turn my attention to the part of the celesta that would hold the bells.

Each bell would be attached to a thin plywood “card”.  Instead of putting the bells above the resonators, I’d put the bells inside the resonators.  The cards would then form the side walls of the resonators and everything would be fantastic, right?  Well, I never got to the resonating part.

Turns out that mounting the bells is a very finicky business.  You need just the right amount of pressure at just the right contact points or you get dramatically reduced sustain — or no sound at all.  Thud.

Complicating matters, my design gives me just 3/8″ in width for the space each bell needs to live in — mounting hardware and all.

So this evening I’ve been playing screws and rubber washers and that kind of stuff.  I’m getting closer , but I haven’t found anything yet which sounds as nice as the original rubber supports the xylophone came with.  You can see the black rectangle of rubber with the points in the photo on the right; it’s perfect except that it’s already a 1/4″ thick, which doesn’t leave enough room for the bell itself in that little 3/8″ cavity.

So I’ve been hunting around the internet for the part I need.  Turns out a “bumper grommet” would be perfect but I don’t think they make them with a small enough point.  Self-stick dome-shaped bumpers might work — like the kind you stick onto the bottom of furniture — but I haven’t found any lying around the house (yet).

Like everything else, this is a matter of finding a simple, cheap, repeatable, safe solution.  I could hand-carve a slot into the plywood for the stock bell support, but I’m not about to do that 64 times.  Something else will work!  But what…?

UPDATE (ten minutes later)

I think I’ve got it!  Or something.  I tried sticking toothpicks into some holes left behind by tiny screws, and … amazingly, I think the sustain is even better than the rubber that the xylophone came with.  I suspect that the toothpicks will deteriorate over time, though, but they’re giving me a very good workable solution for the first time so that’s good news.  Now I’m thinking maybe some old-fashioned thumbtacks will do the trick…

Wobbly

Well the whole notion of a “balance” pin (or two) just wasn’t working out so I went with a flange-and-pin joint like everywhere else.  Unfortunately, because of all the drilling and holes I’d put into things, and due to a bit of short-sightedness, this flange joint is currently only one-sided.  And so the key lever (that long part sticking out at the bottom of the picture below) has a lot of side-to-side play.  Not good — don’t want the lever to drift off-line and get two notes for the price of one!

Wobbly Assembly

There are quite a few flaws in this design that I’m still ironing out, but I think one good principle I’ve centered on is this: any time there’s a hinge, construct it solely out of 1/8″ plywood.  This stuff is pre-milled and consistent and I can plan for that consistency.  I’ve been cutting tenons into thicker (3/8″) pieces to create 1/8-inch “tongues”, but they’ve ended up too inconsistent in size and I’ve had to spend time grinding them down to get them to fit.  The result is that they’re then a little too small, or wedge-shaped, and just have too much play.

I’m already looking forward to a complete do-over starting with fresh wood, but I do need to see this prototype through so I can iron out any subsequent problems in the parts I’ve haven’t even installed yet…

Balance Pins

I thought I had this figured out, but… nope.  Still don’t know how to do the balance pins.

What’s a balance pin?  Well, it’s the main pivot point for the part of the piano that the key is directly attached to.  In modern pianos, it’s a pin that goes vertically through the key, and the hole it sits in is elongated, front-to-back, wider at the top and narrower at the bottom, to accomodate the motion of the key.

A hole which is elongated and wedge shaped just isn’t going to fly in my shop, so I’ve been trying to figure out an alternative.  Here’s where we’re at right now:

As you can see: I’ve got all the pieces for a full octave with a few extra on each side, and they sit nicely in the little text frame I’ve made.  But those long levers just won’t sit still on the nails I’ve driving through the bottom frames.  I think the answer might be to create a flanged hinge (like all the other moving parts have) but now that I’ve drilled holes into the bottoms of all those frames (and glued all those pieces together!) I don’t know if a flange is possible. Back to the drawing board…

Celesta Progress

Things are progressing in celestaville. Not as quickly as I want, of course, but time in the workshop can be scarce. This weekend I made some good progress, though!

A few weeks ago I drew up all the plans I needed to build the action for each piece. I sketched each piece of wood individually, noting its dimensions and where I’d need to cut grooves and dadoes and rabbets and all those fun things.

I also noted some jigs I’d need to make… including a kind of a sled that would slide on my table saw and hold pieces upright. I decided to make this as a kind of a box. And I figured, while I was making a box, I’d use finger joints to hold the joints together!

That led me on quite a lengthy detour. Finger joints are one of the simplest joints to make, but getting everything consistent was a big challenge and led me to refine several of my techniques. In fact, I built a different sled first — one that would make consistent finger joints — so that I could then make the box that would let me do nice tenons without losing any fingers.

I’m currently working on a whole octave’s worth of notes. We’re still in the prototyping stage, but when this is all done and assembled there should be an ugly thing that can make some music.

 

Celesta Workshop Day #2

I spent a few hours in the shop today with the intention of busting out prototype #2, but things didn’t go quite as planned. I made some modifications to the design with the intent of simplifying things, but unfortunately progress went into the other direction. I’m dealing with small pieces of wood (typically 1″ x 6″ x 3/8″ or so) and handling these on a table saw is a tricky, and occasionally dangerous, business. I need to come up with a different design and/or different tools that let me carve the necessary grooves and slots and tabs onto the pieces without putting my fingers into the danger zone.

To that end, my Dad will be coming down next week to give me an introductory lesson to the router he gave me for Christmas. Also, on my drive home tonight I came up with an alternate design for some of the pieces that just might make the grooves and lots and tabs unnecessary…

It’s an interesting challenge to keep in mind that whatever I do, I have to do 32 times (actually more like 40, so I have spare parts) in a consistent, repeatable, safe fashion. It’s a challenge I’m enjoying, because I’m hoping to go make more than one of these.

Celesta Action Prototype #1

I finally got the opportunity to get out into the garage today and bring some of my sketches into reality.  Here’s what about 3 hours of work got me:

Piano action prototype #1

It’s a prototype of a piano action — the part that hits the strings (or, in my case, bell) when you press a key.  It’s currently missing the mechanism for a damper (the part that silences the note when you release the key), the actual bell, and there are a few other things I’m still futzing with.

I need to figure out a better bearing for the movable joints (wooden dowels are too stiff).  Keeping things aligned on the balance pin is tricky, but once everything is lined up with the rest of the keys that might work itself out.  Figuring out where to put the balance pin is also tricky.  Right now, things are a little too heavy feeling.  And after whacking this one rapidly a few times I can see why so much effort went into developing repetition mechanisms a few hundred years ago.

Everything is build from 1/8″ and 3/8″ baltic birch plywood, because in order to fit all the mechanisms for all the keys the action can only be 3/8″ wide.

One of my goals for this project is to make it with simple, consistent, repeatable pieces, so I’ll continue to work with this design to get rid of some of the weird pieces that are harder to make.

Celesta

I spent a lovely evening at one of my favorite bars doing more design work on the piano action last night:

Designing at the bar

Well, actually what’s pictured there is a completely different mechanism I came up with while running yesterday afternoon. It’s a reverse-left-off jack for a new type of instrument I’d like to build that I’m tentatively naming the mateallica. But that’s another story. First, I need to build that toy piano…

Speaking of names, yesterday I learned that what I’m building is actually called a celesta. They’re rare, but Yamaha still offers them for sale. The instrument I’m designing is essentially the same thing, although after reading through their specs I’m starting to think I need to add sustain and soften pedals…

Building A Boat

A wise friend once said to me: “Steve, you should build a boat.” She offered the advice as a suggestion for healing; it would keep my occupied, give me a purpose and a distraction.

So I did: I made a tiny boat for her and her husband as a wedding present:

Never Stop Bailing (a boat for Scooter)

…and then continued on making other things.

But now, I feel like it’s time for something intricate and ambitious and exciting. It’s time to build my boat…  except instead of a boat, I’m building a piano.

Ok, technically it’s a toy piano: instead of using strings (and requiring a giant cast-iron frame to hold them), the hammers will be hitting the bells from a glockenspiel. It will only have 32 keys. But everything else will be as piano-like as possible, from the keys to the action to (hopefully) the way it feels when you play it. I’ll make just about everything out of wood.

Right now I’m in the design phase: scribbling levers and blocks in a notebook and staring off into space on a regular basis. I’m hoping to start working on a prototype action (the action is the name for the mechanism that sounds a note when you press a key) this weekend. Stay tuned for updates!

Designing pianos at the bar