<< mtl.asm >>

[multi_s]

Hi,

duty cycle changes the shape of the wave form in some sense.

so if you had a square wave, but could change the proportion of high to low time for example.... or with a triangle wave it changes say from a saw tooth to a symmetrical triangle etc . you can change the range of the lfo as well but duty cycle isn't quite that.

and yes thanks Tom for moving us here! we shoudl ahve a give away or qqc. hmmm.

[Strange Tales]

I wish I could, not sure if I'm smart enough to understand the math behind PLLs though. The math behind it is fucking intense and might've killed my idea.

http://liquidsdr.org/blog/pll-howto/

I think most of the math revolves around the phase comparator section though.

What's your idea? Does it have to do with the PLL itself or something in Pure Data?

Well my thought was to do sort of a "Wavetable" PLL where instead of just having a square wave be the tracker you could have a wavetable of various forms track your signal. Not sure how well others would track (like a stepped waveform and what not), but that's the fun of experimenting and trying shit out! Second idea was that I would pass the PLL output into a vocoder because vocoder's are fun. The one thing being I can't do any analog signalling here so it'd be all Pure Data using the mxmxmx Terminal Tedium (basically just a Raspberry Pi breakout board for Euro).

Most of the math on the page is talking about how to transfer an analog filter (paramaterized a certain way) into a typical digital filter structure (which has different parameters, but teh same response) i believe. the concept of how the coded pll works is the same as the analog one. im sure you could easily do it with experimentation as well.

Is there a certain filter structure to LPFs on PLLs?

[multi_s]

I wish I could, not sure if I'm smart enough to understand the math behind PLLs though. The math behind it is fucking intense and might've killed my idea.

http://liquidsdr.org/blog/pll-howto/

I think most of the math revolves around the phase comparator section though.

What's your idea? Does it have to do with the PLL itself or something in Pure Data?

Well my thought was to do sort of a "Wavetable" PLL where instead of just having a square wave be the tracker you could have a wavetable of various forms track your signal. Not sure how well others would track (like a stepped waveform and what not), but that's the fun of experimenting and trying shit out! Second idea was that I would pass the PLL output into a vocoder because vocoder's are fun. The one thing being I can't do any analog signalling here so it'd be all Pure Data using the mxmxmx Terminal Tedium (basically just a Raspberry Pi breakout board for Euro).

Most of the math on the page is talking about how to transfer an analog filter (paramaterized a certain way) into a typical digital filter structure (which has different parameters, but teh same response) i believe. the concept of how the coded pll works is the same as the analog one. im sure you could easily do it with experimentation as well.

Is there a certain filter structure to LPFs on PLLs?

yes i had a similar idea before to the wavetble thing. i don;t know much about puredata but it is definitely codeable. also i thought about using a dsp front end to control many pll voices using some sort of specrral analysis, so you could maybe get chords, or at least polyphony. into a vocoder would also be awesome.

i don;t think it matter much what is the structure of the lpf. just think about why you need one and tune to taste. the block diagrams in the article you linked are very instructive.

[Strange Tales]

Hmmm, I like that polyphony idea. Was your idea to split the voices into multiple lines and go from there? No idea how that stuff works.

I wonder if you could take that idea (splitting the voices) and then control polyphony by having a knob that would create frequency divisions on the PLL in musical scales. Or you could take the outputs of having various frequency divisions and quantitizer it to musical scales but it'll be fucked up to varying frequency divisions.

[multi_s]

well just look at why say the normal pll can't track multiple notes-> because it is a phase comparitor , so if there are many frequencies in the input phase has no meaning really right?. what should it track? so yes if you could like prefilter the input and seect certain frequencies that are present, seperate them and feed them to separate pll voices it could work?


anyway. this is beyond what this one does so maybe a bit ot for the thread. but making musical division only is sort of trivial extension of pll, and also from designing this i also figured out a way to actually go one step further and have it more or less auto tune so you could say add midi and play a chord, then have it arpeggiate the input on the actual notes fed into it which might be fun. im imagining more a microphone being used ass that input though like singing and autotuning it with a keyboard, sot of a synth effect. again, this does not do it now. maybe we will make a super pll in teh future but it woudl probably be quite large form factor.

but ya for this one the main thing i wanted to add that i didnt see on any pll incarnation yet(although maybe it iexists somewhere) was CV control of the multiplier and divider, and then add an LFO on board. certainly not a monumental achievement but it does make the pll circuit a lot more fun.

iw ill try to make some clips on Thursday or the weekend.

Last edited by multi_s on Wed Feb 22, 2017 4:09 pm, edited 1 time in total.

[DRodriguez]

Whoa, dig the new digs. That autotuning super pll plus sounds like it could be fun

[Seance]

congrats on the forum move scott!

[molokaio]

Yeah. You guys had me scared for a minute. I was in the shark tank and couldn't find mtl.asm any more and was like: wtf happened to scott?!?! Why did they kick him? Ahah

[lumena]

The suspense is killing me.

[Inconuucl]

The suspense is killing me.

I was almost expecting a grindcore song to start after that.

[lumena]

[LudicrouSpeed]

Well, after reading through this thread I now get ads for rapecicle paint. Could be worse. The upcoming PLL looks amazing.

[multi_s]

Here 's a pic with the proper drill pattern.

we will be at thw stompbox/synth exhibit at SXSW in austin in a few weeks so we will debut it there. we will have most of our pedals there to try.

Attachments

2017-02-27 23-56-51.639.jpg (163.41 KiB) Viewed 4680 times

[oldangelmidnight]

!

[chrisdermo]