Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards DIY real grand action controller?

Hopefully that's the case! But 80/20 dictates that this is where various unforeseen difficulties really start to come into focus

We're all rooting for you CyberGene. You might make it such that the hardest part of doing this is sourcing the action.

When still looking for an action, I contacted a Chinese piano manufacturer on alibaba and they refused to sell me one but said an order of 20 or more is feasible.

What kind of action are you seeking?
Maybe you can get a used digital piano that has a suitable action?

Did you read any of the thread? The whole thing is about building with a real grand action, which he has.

Prototype real grand piano action controller from Robert Estrin (Living Pianos in California).

https://youtu.be/3KlxlgUepOo?t=55

It's kind of cool looking to have a naked grand piano action in a studio. No housing at all. To see the hammers moving is fun, Might even help with playing to sense what the action is actually doing. And regulation would be a breeze.

It's very mesmerizing looking at the action while playing, I've already realized that with my last prototype (10 keys working) and wondering whether I should try to wire it as neatly as possible and just keep it this way

Currently struggling with coming up with ultra low ripple power supply. Since I rely on voltage comparisons with a 1 mV precision and most power supplies are in the range of 300-700mV ripple. On the other hand my prototype works pretty well and is so dynamic with my iPad charger, maybe ripple isn't such a big problem or the iPad charger is very stable. And it's 2.1 watts which is exactly how much I need, I will probably just keep using it...

Wow, 300-700mV ripple is too much even for poorly designed PS. Are you sure this measurement was taken correctly? Anyway, depending on the current consumption of the circuit, you could use an LDO (low drop-out linear regulator) with an LC input filter to achieve higher precision, lower voltage noise. I have done so in some designs where noise had to be low.

I was looking at the specs of 5V wallwart adapter I have and it states 200mV ripple but I haven’t measured it since I don’t have an oscilloscope. I also have a 24V linear power supply which I measured with my multimeter (in an AC mode) and it showed 700mV and 50Hz but I guess it’s OK. From what I read the best way to obtain clean and low ripple power is to first use a switching regulator followed by a linear LDO and then a filter. I’m not very good at electronics, I learn what’s needed when needed so still scratching my head around that
BTW, I need 5V at 2A, so that’s 10W, not 2W as I said above, it was a typo. EVC2017, do you have a ready schematic that might work for that purpose?

CyberGene, I am not sure measuring like that (i.e. multimeter in AC) is reliable but that may depend on the multimeter brand and model. In my experience multimeters show odd readings used this way. I usually use a oscilloscope to measure ripple (there are some tricks too). 700mV on a linear power supply is a lot, unless it is not regulated whatsoever.

Texas Instruments have some awesome switched DC-DC converters and most of them have an evaluation board. They have an online tool that helps you finding an appropriate part based on your requirements and then you may simulate the design. With the proposed design you may then change the evaluation board to perform tests on it. I have designed some supplies using their converters with an LC filter on the output that achieved a pretty low ripple (on simulation). An most of them are very accurate.

If you send me your requirements, i.e. input voltage (from the AC-DC converter), output voltage (5V?) and output current (2A?) I may generate the design for you. You may PM me.

Alternatively you may simply use a supply such as the iPad charger (wallwarts of good provenance are usually very good) and maybe a small LC filter to remove high frequencies.

You get decent results from the Apple and Samsung switched mode wall warts because they are well designed, well built, and use decent parts. They are also expensive.

Measuring and evaluating noise, ripple, noise supression, power regulation, protections schemes, etc. gets difficult.

So I would bypass mains conversion work and just buy a high-quality linear power supply that is rated well over 2A. That might be a bit too big for "wall wart" form.

"Power bricks" have some disadvantages but can help you to avoid dangerous mains voltage "entirely" and bypass (some) endless analysis. Industrial engineers may use commercial bricks hoping to: (protect young children), simplify in-house design, shift liability to the power supply maker, reduce burden for local health & safety ratings requirements, etc.

Mean Well has a wide variety of decent power supplies at popular prices but you will probably find much higher quality options from some local Bulgarian specialists.

Regardless, consumer and commercial electronics generally don't rely just on the main power supply for all noise filtering, regulation, etc., so you would consider some simple power schemes adjacent to your systems. For now, I would assume the Arduino has sufficient on-board power management. But your other DIY boards might benefit. I do some LED systems design but I fear that doesn't help you much here. EVC2017 and some of the people here probably have some good ideas. As you know the deskthority folks specialise in similar sensors and controllers so it might be worth pinging them.

Apple wall warts are well designed, well built, and use decent parts???
I've not seen or heard that. Rather, it seems that they are notoriously crappy.

Here’s a comparison of various phone and tablet chargers.
A dozen USB chargers in the lab: Apple is very good, but not quite the best

The iPad one is rated highly.

hahaha - I knew you would jump on that comment.

In my experience, Apple and Samsung wall warts are: reliable, meet local safety standards, and provide good quality power for charging phones. I've taken a few apart. They aren't fantastic but fit for purpose IMO.

Some off-brand USB power plugs won't meet standards in any country. No fuse protection. Overdriven components. Poor isolation from mains. Lousy power conversion. YouTuber Big Clive has several videos on a few eBay units that can run full mains on USB out!

Thanks for the interesting blog link CyberGene.

PianpTeq forums referenced a thread where the Russians hacked some of the StudioLogic slabs.

They add a few washers to physically adjust the sensors. Then upgraded MCU to STM32F407VG, allowing faster scanning, HighResolution MIDI, etc.

https://rmmedia.ru/threads/117290/

English translation for those not fluent in Russian

https://translate.google.com/translate?hl=en&sl=ru&u=http://rmmedia.ru/threads/117290/

Thanks newer_player, fortunately I read Russian fluently so this will be an interesting read.

Today, Robert Estrin of www.LivingPianos.com posted a new video with more clues about his California "DIY real grand action controller".

In the video, we real detail of hammer stop-rail, which seems to be a few aluminum bars with rubber at the bottom. The bars are at different heights. We see his nifty wooden DIY support & height adjustment system. Maybe he uses different types of rubber for the higher registers, although I'm not sure that would matter.

I can't see any sensors or wires on the controller (well there are some unrelated fat cables to the huge TV screen on the wall). So the sensor system is a mystery, although fact that everything is hidden provides some clues I suppose. Maybe the bar is hiding something?

https://youtu.be/ostbE81LFwo

This is last month's video showing the piano playing live via PianoTeq so this definitely is working. . .

https://www.youtube.com/watch?v=3KlxlgUepOo&feature=youtu.be&t=55

I think they’ve probably installed a system like PNOScan under the keys. If so, that’s not very good Might be wrong but I can’t believe there would be hammer sensors that are not visible in the video.

I bumped into some interesting diy keyboard project apparently from California. This site has some interesting pages on optical switches, key scanning, etc.

http://diykeyboard.wikidot.com/key-scanning-circuit

Site linked an interesting scanner chip from a UK shop:

http://www.hinton-instruments.co.uk/components/kv64/index.htm
http://www.hinton-instruments.co.uk/images/components/kv64_kbd88.gif

Below is a video of the Yamaha AvantGrand action in action. It is really quite interesting:

https://youtu.be/FG5EVGf98lU?t=155

- Only one demo action unit in the world

- Rep says that Yamaha MIGHT be using over 1023 midi levels for internal sounds, noting Yamaha did not release detail publicly at about 4:10 (John Ebata of Cosmo Music Canada)

- Note large bunch of fibre optics (?) in aluminum cable tray above keys to local board at about 4:02

- Maybe local board runs 12+1 black copper wires towards main board