NEW OSCILLATOR

[BobF]

BobsRingOscillator.fsm

(165.8 KiB) Downloaded 1276 times

Hello gang,

Ok, yes I know we have many nice oscillators, but this is NEW and shows off again just what you can do with Flowstone.

This is called a "RING OSCILLATOR" and comes from the TTL and CMOS world, just do a Google search for ring oscillators.

It is controllable over a wide range, but is only a pulse, then square waves as you can see. Anyone want to do some wave sharpers for it? Also I have made a few other ways of controlling it that I will submit later, depending what the response is on this version. I have even made it midi controllable, but that still needs some work, and I may need some help with it. Any volunteers?

Anyway enjoy and PLEASE submit uses, mods, and so on, if you come up with any!

Later then, BobF.....

[BobF]

BobsRingOscillators.fsm

(194.93 KiB) Downloaded 1170 times

Hello gang,

For those who downloaded the above "Ring Oscillator", here is a follow up.

Just some other methods of controlling it.

PLEASE comment with mods, uses, and so on.

Thanks, BobF.....

[Spogg]

Hi Bob

I think I'll need you to educate me on this a bit

AFAIK a ring oscillator relies on gate delays for each of the NOT stages but in Flowstone the whole schematic is evaluated once each sample (at sample rate), so there is no delay like you would see in any solid state component.
Also a ring oscillator should self-oscillate, at a rate determined by the NOT stage delays, and not need a clock.

I would appreciate it very much if you could clarify what's going on in the schematic.

Thanks for another stimulating idea Bob!

Cheers

Spogg

[adamszabo]

I thought ring modulation is simply multiplying two oscs together?

[Spogg]

I thought ring modulation is simply multiplying two oscs together?

A ring oscillator and a Ring Modulator are different beasts. The Ring modulator is what you say, 2 signals multiplied together.

Cheers

Spogg

[BobF]

BobsRingOscTest.fsm

(220.09 KiB) Downloaded 1212 times

Hello Spogg and gang,

Ok, here is another ring oscillator with just three not gates(inverters) and you can see it does oscillate. You can see this does meet the criteria of a ring oscillator as it has an odd number of gates. Also the frequency is very high as it would be in a hardware ring oscillator, but must be aliased down here. Even though I can NOT see it with our Flowstone scopes there must be a small amount of delay in each gate. My guess, if someone has a better answer PLEASE do comment. As for controlling the frequency I am just using a shift register or a all pass filter to introduce more delay and add a controllable frequency range. Of course because of the high frequency out put I am putting it through a divider. Strange stuff I just thought I would try it and it worked. I have even used an odd number of and, nand, or, and nor gates and it still works just fine. Now if we can only come up with some cool uses for it. Anybody want to play with this that likes digital logic it's a lot of fun.

Cheers, BobF

[adamszabo]

These are pretty cool, but its much easier to build them from a standard ramp. All these comparisons (==, >) cost quite come CPU, and generating this triangle with a ramp is so much easier. Maybe this has some other uses though.

[BobF]

Hello adamszabo and gang,

Yes, true it is eaiser to use stock oscillators, but the idea here is that this is something new. Now I hope someone can find some cool uses for them but if anything it is educational and also so shows others just what you can do with Flowstone. As you look around the forum you often come across fsm's that seem poor, have no use, take up cpu cycles, but again they can educate ideas for others, and new fsm's can come from them and so on. A lot of the fsm's I put up are often to pass on an idea, or show a technic. Sometimes they are not even complete or working 100%, my gui's are terrible . I (and others) are often looking for help and comments. Not always forth comming for some reason, but at least the fsm is there for future others. Often like this new oscillator they just lead to some good discussion. Hope you get the idea.

Later then, BobF.....

[Spogg]

Fascinating!

I think this works because for each sample pass of the schematic the NOT gate outputs in the chain will flip. This is not down to a delay, like in analogue or solid state digital circuits, but rather down to the sample-based system. This is also indicated by the frequency readout which is half the sample rate. I put 7 NOT gates in series and got the same frequency but 8 in series gave nothing because, as you say, it needs to be an odd number.
With real-world electronics more gates would mean a lower frequency, due to the limited slew rate of each stage.

Cheers

Spogg

[BobF]

Hi Spogg,

Thinking about this some more you are probably correct. Any Flowstone guru's here that would like to add some comments letting use know what is happening here would really be much appreciated.

Thanks all, BobF.....