Optical contacts wooden keyboard

Greetings!

I'd like to use optical sensors/shutters as contacts on one of my manuals (short keylength with springs at the back) and I can't seem to find any reference telling me where to purchase these, or even how to make an assembly myself.

I wondered if anyone has purchased a set already configured and ready to install or could give me some pointers about assembling my own.

My other manuals are 4 pin wire with shorting plates attached to the keys. They involved a lot of soldering and adjusting and I'd like to see how an optical set works before I go the reed switch route.

Many thanks,
Jim

Hello Jim, We live in Florida, but last week we visited friends in Toronto, and returned during the Snow last Sunday !

Three years ago I built and installed my own Optical switching system.
My Keyboards and pedals are all optically switched, eliminating all of the original contact blocks, and all of their associated problems !

I have not had one problem with this system in years, and a very obvious, and a most noticeable positiveness to the keying, that never changes !

You also mentioned, Magnetic reed switching,
they will function well but are Very difficult to work with from personal experience, as I rebuilt a three manual Pipe years ago utilizing by necessity, very small reeds , successfully but with much difficulty.

BTW, I am a retired Organ Tech, with a background that includes much electronic experience.

http://s383.photobucket.com/user/TheOrganDoc/profile/

RSVP,, Sincerely Mel

I agree with Mel that reede switches are NOT the way to go for manuals. My optical approach is described here:

viewtopic.php?f=21&t=10628&p=89281#p89281

along with some other console related ideas. I have had some PCB's made for the sensors (they are the dark green ones in the bottom picture). I can post details of these if you are interested.

Nick

I did some initial experiments with a Kingbright KTIR0611S Photointerrupter and managed to get it working with a Midiboutique encoder. However, the KTIR0611S is physically a bit too wide to use in a manual keyboard scenario, so I bought ten HY301-21 slotted opto-switches off Amazon and managed to get eight of these working in an arrangement similar to that shown in Nick's 2nd picture.
On the strength of this "proof of concept" exercise I've just ordered 100 of them from eBay where they work out at just under 20 pence each - which is pretty good, but the down-side is that you have to wait for them to ship from Hong Kong!
http://www.ebay.co.uk/itm/PK-10X-10-Pcs ... SwMsVXjlAk

I'm planning to use them on a keyboard that I originally fitted with reed switches a few years ago. It's an ex-Compton keyboard that has long, centre-pivoted wood-core keys that lend themselves nicely to having the key-contacts at the rear. With this geometry it's easy to achieve {key depressed = flag lifted from gap = IR beam un-interrupted = Note-on}.
However, I still haven't figured out how I would use slotted opto-switches with short, rear-hinged keys where you are more or less obliged to position the opto-switches underneath the keys - unless there's a way of wiring them such that {key depressed = flag blocking gap = IR beam interrupted = Note-on}...

I concur with what Nick said about exposing the photo-interrupters to bright direct sunlight. In fact, experimenting under conditions where they might be affected by not-particularly-bright indirect light (reflected off your hands, for example) can lead to considerable confusion until you realise what's happening!

Graham

Hi Graham,
please post updates (pictures?) of how you get on. I have a pair of Compton keyboards I will (eventually!!) get around to midifying, so I would like success/failure stories to plan the approach to the project.

Cheers,
Tony

However, I still haven't figured out how I would use slotted opto-switches with short, rear-hinged keys where you are more or less obliged to position the opto-switches underneath the keys - unless there's a way of wiring them such that {key depressed = flag blocking gap = IR beam interrupted = Note-on

Hi Graham,

This is exactly how some of my keyboards are wired. My own encoders can be configured for either polariy of scan or (independently) note-on/off polarity. The only down side is that, at rest so to speak, all the opto switches are 'on' which means that the maximum current draw is when no key is depressed.

viewtopic.php?f=15&t=12849&p=94757

Nick

Hello Murph, Nick, and all,

Please click the link, in my post below,
for a few photos of what I put together years ago,
All playing reliably for several years now.
Both Keyboards, and the 32 Pedals are all Opto keyed !

I have recently designed a new PC Board, for use either on top of the keys inverted,
or inverted under the keys, but with all shutters having two, 90 degree bends,
so that when you depress the keys the shutter "UN-blocks" the IR beams as the keys are depressed,
therefore the opto-transistors only conduct when the keys are depressed, like open contacts do !

This newer design also utilizes two opto's for each key,
spaced carefully, in order to be used for touch control !

Mel

Hi Nick

This is exactly how some of my keyboards are wired. My own encoders can be configured for either polariy of scan or (independently) note-on/off polarity. The only down side is that, at rest so to speak, all the opto switches are 'on' which means that the maximum current draw is when no key is depressed.

That's what I was beginning to suspect - that I would need a different kind of encoder in order to achieve it... And I wasn't keen on the idea of the transistors being "on" most of the time.


Hi Mel

I have recently designed a new PC Board, for use either on top of the keys inverted,
or inverted under the keys, but with all shutters having two, 90 degree bends,
so that when you depress the keys the shutter "UN-blocks" the IR beams as the keys are depressed,
therefore the opto-transistors only conduct when the keys are depressed, like open contacts do !

I'd come round to thinking that an appropriate shutter design would probably be the answer - and you have confirmed it. However I'm not sure if I could engineer it in practice, in the limited space underneath the keys and given that I'm limited to working with 1/10" pitch stripboard rather than custom-designed PCB's.
Fortunately it's not an issue at the moment, since I'm usually dealing with keyboards where I can mount the contacts at the rear.


Hi Tony

please post updates (pictures?) of how you get on. I have a pair of Compton keyboards I will (eventually!!) get around to midifying, so I would like success/failure stories to plan the approach to the project.

I've successfully midified a number of Compton keyboards using reed switches, but this will be my first attempt at doing one with opto-switches. Opto-switches pose a different set of challenges from those posed by reed switches, so it will take a little while - but once the challenges have been overcome I think opto-switches will prove to be an easier solution than reed switches for less than half the cost. I'll post an update in due course!

Graham

Hi Graham,

You can of course change the sensing polarity by deciding, at the design stage, whether to put the load resistors for the photo transistors in the collector or emitter circuits.

If in the emitter, the output signal will go high when the shutter is out of the gap and the transistor is on, if in the collector it will go low.

One thing to bear in mind is that phototransistors tend to have large junctions which usually leads to quite high capacitance unles they are photo-darlington transistors. For this reason I find that using rather low load resistance (say 1k) may be needed to get them to switch fast enough that the responses don't overlap with the next bank being scanned. This leads to about 5mA draw for each transistor when on. In an 8x8 matrix scan, only eigth transistors can be on at any one time so the maximum current requirement would be about 40mA (not including the LED IR emitters) which isn't actually too bad. I just resent wasting energy on a matter of principle. For a non scanned system of course you would be looking at 300mA with all transistors conducting, which is getting a bit much.

Also, I expect that you have realised that it is good to get the sensors as far from the pivot point as possible since this maximises the movement of the shutter and makes setting the speaking point more precise and consistent. If not at the tails, I find it much easier installing the sensors above the keys rather than below, if there is room.

Nick

I designed all of my PC boards in the free download, http://www.Pad2Pad.com,
the boards they produced have been working beautifully for years now !

(PS, Nick, all of my devices are Photo-Darlingtons .)

Best wishes, Mel

Hi Graham,
what I was thinking in terms of optical was using parallel scan with something like an arduino mega doing the encoding. No worry about led rise times, no positive/negative worries (just change one line of the code). De-bounce with an rc network local to the optos, if necessary. No shift-registers or any superfluous circuitry. If using the new usb hid library, I would suggest telling each mega to transmit on a different midi channel, which should keep HW happy if windows goes re-numbering........ (Unless someone feels like making the hid library unit description/interface name easily changeable???)

It's the mechanics I hate.....


Happy Turkey day all!!

Tony

Thank you very much everyone for your kind responses. I have to admit that I am a bit daunted with the photos and descriptions involving diodes and resistors and the like. It is clear that I have a lot to learn about circuitry and sourcing the circuit boards and parts needed, never mind assembling them.

Does it make any difference that I already have a midi control unit with a spare board in it? i.e. would I still need additional circuit boards? For my pedalboard I used reed switches from Peterson and just ran the wires directly from the switches to the pins on the midi control unit. I perhaps naively thought I could do the same with the optical switches, since that's what I'd do if I were using reed switches for the manual too.

Mel - I'm going to be in Ft. Lauderdale in late February - are you close to there? It would be great to have a face to face conversation if possible.

Again, thanks for your input everyone.
Jim

Jim,
PSE have a look at this ready-to-install system:
https://www.qrsmusic.com/PNOScan.asp
Not the cheapest, but the best.
Rdgs,
amun.

Hi Jim

[quote]For my pedalboard I used reed switches from Peterson and just ran the wires directly from the switches to the pins on the midi control unit. I perhaps naively thought I could do the same with the optical switches, since that's what I'd do if I were using reed switches for the manual too.
[/quote]

That is exactly the position I was in when I first started considering opto-switches - and the more I read, the more confused I got.
However, I am now thinking that the "typical" circuits described in opto-switch manufacturers' data sheets, the circuits described in Arduino enthusiast's forums and the circuits described by guys like Nick who design their own MIDI encoders are not appropriate to the sort of situation that you and I are in.
I usually work with the Midiboutique HWce or HWce2x encoder, although I have also worked with Roman Sowa's (midi-hardware.com) kit. In either case all that is necessary to get Note-on/off messages is for the key-switches (in my case, reed switches) to be connected across the appropriate pins of the MIDI encoder board.
So talk of pull-up or pull-down resistors, output voltages going high or low and so on had me baffled. Yes - I sort-of understood what they were going-on about, but I just couldn't relate it to my set-up.

However, I think I have now devised a circuit that will work and in which the photo-interrupter acts simply as a switch rather than as some kind of pulse generator.
[img]https://www.dropbox.com/s/fsac2am0ucw4izk/Opto-switch.jpg?dl=0[/img]
To prove the point, first off, I got a single photo-interrupter working, connected across pins 1 and 9 of my MIDI encoder. When the "switch" was closed (slot unblocked and photo-transistor "on") I got the 036-C Note-on message. When the "switch" was open (slot blocked and photo-transistor "off") I got the 036-C Note-off message.
Next, I set up a group of eight photo-interrupters, with their collectors individually connected (in series with a resistor each) to pins 1-8 of the encoder, and their emitters connected all together to pin 9. And with this set-up I was able to get Note-on/off messages for notes 036-C to 043-G.
I see no reason why I shouldn't be able to extend this set-up to the full keyboard compass of notes 036-C to 096-C. And that's what I intend to prove when my photo-interrupters finally arrive on the slow boat from China.

Now, my Midiboutique or midi-hardware.com encoders are of the scan-matrix variety designed for 8x8 matrices in common-cathode configuration - which is why I did the proof-of-concept experiment with a single group of eight opto-switches. When I MIDI-fied my first console more than twelve years ago, I didn't want to know about diodes and I was scared of matrices. But nowadays I can almost wire a reed switch and diode matrix with my eyes shut, and I wouldn't dream of going back to a scan-row set up because a 16-conductor flat cable with 16-way IDC connectors is so much easier and neater than 61 individual wires plus a common! And - take note of this, Jim - the photo-transistor part of the slotted opto-switch acts as a substitute for the old reed switch and diode combination, so half the job is done for you already!

The only downsides that I can see to all this (bearing in mind that I've yet to prove it with a full set of 61 switches) are (a) the infrared LED units of the opto-switches have to be powered, so each opto-switch has four tails that need locating in close proximity and soldering, as opposed to the two tails at opposite ends of a reed switch and (b) opto-switches are a bit bulkier than reed switches, so there is the whole challenge of how to mount them together with their circuitry, plus the shutters attached to the keys, in the space available.
I'll let you know how I get on ...

Graham

Hi all,
I found a good discription of the QRS optical sensing system, although it is quiite complicated in the maner it is utilized with Pianos, and controlled by a smartphone, It is intreresting anyway !
]https://youtu.be/cACZqIlGx_Y Mel