Electronic Organ

Sinewaves or complex waveforms?

Attack-Decay-Sustain-Release envelope?

Monophonic or Polyphonic (how many notes)?

Luhan

Hello Luhan,

I just need to produce simple tone, so I assume sinewave would be enough... as for the envelope, you will have to excuse my musical knowledge... I'm not sure what to answer you...

As for the number of notes, we would like to be able to play several note at the same time.

Does it help you?

Thanks

Eric

Luhan wrote:

(please bottom post)

The simplest method may be a microcontroller making square waves. I have gotten as many a 4 simultaneous notes out of a PIC - but you have to do the programming.

Note producing systems are generally rated by the degree of polyphony - how many notes you can have at the same time. There is usually no real limit on the number of notes available.

Luhan

Look in any toy store. There's no need to re-invent the wheel for the billionth time.

School assignment?

Pick up a cheap synthesizer like a Casio. They have many voices and most operate over four octaves or more. As was said above, don't re-invent the wheel unless you want the exercise in programming a micro or something. BTW, the eight note octave Do to Do is incomplete. A chromatic octave has 13 notes not eight. C, C#, D, D#, E, F, F#, G, G#, A, A#, B, C. If your octave is not complete, you are very limited in key. Bob

Strictly speaking, the "well-tempered" scale is one that's not an exact chromatic octave in any key, but "close enough" in _every_ key to get away with it. Each freq. is the lower one times the twelfth root of 2.

Cheers! Rich

Rich Grise wrote in sci.electronics.design:

Err... the octaves in a well-tempred scale *are* exact, it's the other intervals that are approximations.

Anno

Even temperament means that the ratio between successive semitones is the twelfth root of 2, 0r about 1.0595. Only electronic instruments really get tuned that way though. Piano tuners bend the tuning quite a bit, particularly at the extreme ends of the keyboard.

Paul Burke

Many years back pop-tronics (I think it was) had a design that used NE-2s. It only needed one RC and NE-2 per note. IIRC the signal was taken from the common point of all the capacitors.

OK, I misspoke. The _octave_ is a perfect octave, as you say, and it's the other notes - maybe I should have said, "chromatic scale" instead of "octave".

Thanks, Rich

Thank you very much for helping me. You are right...I do not want to re-invent the wheel!

I also found an IC that is specialized for this application: MOSTEK MK50240 Top-Octave Frequency Generator. I just need to find a supplier for this...

Thanks!

Eric

Rich Grise wrote:

Have you got a time machine?

The MK50240 is outrageously and horrendously obsolete.

Check the Ensoniq chips instead.

But the magic numbers appear in my CMOS and TTL cookbooks. There is only ONE eight bit sequence that gets all of the notes right.

Those things were so impressive when they were introduced, doing away with a bunch of oscillators or at the very least an oscillator and a bunch of small count digital counters. All you needed was an oscillator of the right frequency, and you'd get an octave of notes all with the exact ratio to the others that was required. More octaves were easy, since it just required a string of flip flops per note (or a more integrated string of them). Get that needed sawtooth to boot, with summing resistors of the outputs of the flip flops.

But yes, things have leaped further since then. Last week, I found a small Casio "synthesizer" on a pile of garbage, sadly one black key is broken, which I intend to give to a one year old when he's a tad older. This is twenty or so years old, likely cost a fair amount when new, but they were doing polyphonic in that thing with a single IC (so I hear, I've not opened up), I think a CPU.

It's gotten so easy that every computer basically has a full synthesizer built in, and top octave dividers are long in the past.

Michael

Don't forget all your articles in Popular Electronics where you explained the insides of a synthesizer, about when top-octave dividers were the cat's meow, a big step forward from analog synthesizers, but only a small step in terms of what came later.

Michael

The wave shaping is actually done on the far side of the key contacts.

(This requires only 1 key contact per rank per key.)

Michael Black wrote: > "Michael A. Terrell" ( snipped-for-privacy@earthlink.net) writes: > > Rickydou wrote: > >>

Musically, the 50240 was an 'improvement' over 12 discrete out-of tune oscillators but it was not good. The intervals were all over the place, some notes sharp and others flat -- musically unpleasant. I replaced one with a raft of 74F chips running at 32 MHz instead of 2 MHz which managed to cross the boundary between 'appliance' and 'instrument'.

I bought some organ samples from Milan Digital audio and re-sampled (tuned) using CoolEdit. You can hear an example here.

Not any more.

All the keys are scanned into a standard xy keyboard matrice.

Obviously it's all done differently now.

But even in your CMOS Cookbook, and your article in Popular Electronics that showed the basics of a polyphonic synthesizer using the top octave divider, you show ramps going to the VCAs, and those ramps were generated by summing the outputs of the ripple counters at the outputs of the top octave divider.

And, a page of my copy of the CMOS Cookbook just fell out, 29 years after I got the book. That's the first of your books that have come apart, though I admit I likely have taken more care of them than some books. (On the other hand, the various TAB books I have have never fallen apart, but they've seen far less use.)

Michael